2 Mar
2014
2 Mar
'14
12:49 p.m.
Hello, all,
This morning I did some looking at my RK8E boards, and it appears that
it has been ECO'd to make a change to the clock generation logic for the
Current Address Counter.
This ECO supposedly helped with the issue, but apparently it isn't good
enough for this RK8E to work properly in my 8/e.
I tacked in an extra probe wire to the least significant bit of the
Current Address Counter on the RK8E.
I put things back together, and toggled in my little test program that
loads the Current Address Counter with 0, then loops on itself to
generate scope traces.
When I ran the program, it was instantly obvious that the counter is
indeed incrementing due to the transient in the clock signal.
The LSB of the counter goes to logic 0 for a short time (as a result of
the load operation), then pops up to logic 1 due to the transient
causing a count cycle on the 74161 counter.
You can view the traces here:
http://pail.bensene.com/RK8ECounter.jpg
The top trace is the LOAD signal into the counter, the middle trace is
the CLK signal into the counter, and the bottom trace is the LSB of the
counter.
It's clear that when the CLK signal goes high (the clock triggers on the
positive edge) after the transient, the counter counts up one, causing
the LSB to go to 1.
As David Humphries mentioned, it appears to be a design flaw in this
clock generation circuitry that, depending on small timing variances in
different machines, causes this transient to be generated, messing up
the operation of the RK8E.
A couple of different signals from the master timing of the PDP 8/e are
used to generate this clock signal, and it could be that one of these
signals on my machine is slightly off due to variances in propagation
times in gates and flip flop timing chains that manifests this weakness
in the RK8E Current Address Counter clock circuitry.
The machine where the RK8E was tested as good may have had slightly
different timing that either rendered the transient too short to trigger
a count, or eliminated it completely.
The next step is to try to figure out what I need to modify to make this
transient either go away or such that it is no longer effective in
triggering the Current Address Counter to count up before DMA operations
begin.
Suggestions as to how to clear this up are warmly welcomed.
Thanks to all,
Rick Bensene
3 Mar
3 Mar
8:37 a.m.
Another quick update:
I got a personal Email from David Humphries, where he explained the
modification to eliminate the transient that causes a false increment of
the Current Address Counter in the RK8E disk controller.
The modification is quite simple - piggybacking a 7474 (dual D
edge-triggered flip flops) onto one of the chips that is involved in the
generation of the clock pulses for the Current Address Counter
(piggyback is only for power), and then patching one of the flip flops
into the circuit to latch the TP3 signal to prevent it from going low
before the signal that is the decode of the DLCA (Load Current Address
Counter) IOT instruction de-asserts. On David's RK8E with similar
symptoms, this modification prevents the spike from occurring, and will
hopefully serve as well to fix the issue I've had with my RK8E.
I am going to make this modification to the M7105 board today, and will
post the results as soon as I can.
Thanks to David for digging through his notes to recover the information
on the modification that cured the problem he was having with his RK8E
and sending it to me.
If anyone wants the procedure for the modification, I can Email it to
you. Just drop me an Email and I'll send it along.
Best to all,
-Rick Bensene
4 Mar
4 Mar
5:50 a.m.
New subject: Interest in a PDP-8/e/a/f/m 32K RAM kit?
I've heard from a fellow, Vince S., who has created a PCB layout for the PDP-8/e 32K
RAM design presented in my article on Tronola.com. He's considering a Kickstarter
launch to offer a $95 kit of parts and PCB but is wondering if there is sufficient
interest. If so, please let it be known.
Features include:
- Battery backup with estimated 10-year data retention.
- Inhibit on power fail to prevent memory corruption.
- Most of the full size Omnibus board offers a prototyping layout.
- Cooperation with any mix of core memory boards. Switches disable 4K fields.
- Compatibility with Omnibus specs.
- All through-hole parts, for easy assembly.
I'm not associated with the project, other than having published the article and
encouraging his efforts. Vince has published lots of PDP-8-related and other layouts on
his website and will make this one available as well:
http://so-much-stuff.com/pdp8/cad/cad.php
The article is at:
http://www.tronola.com/html/ram_for_pdp-8e.html
Thanks,
Steve Lafferty
http://www.tronola.com/
10:36 a.m.
New subject: Interest in a PDP-8/e/a/f/m 32K RAM kit?
On 03/04/2014 02:50 PM, Steve Lafferty wrote:
I've heard from a fellow, Vince S., who has
created a PCB layout for the PDP-8/e 32K RAM design presented in my article on
Tronola.com. He's considering a Kickstarter launch to offer a $95 kit of parts and PCB
but is wondering if there is sufficient interest. If so, please let it be known.
Count me in....
Jos
10:52 a.m.
New subject: Interest in a PDP-8/e/a/f/m 32K RAM kit?
Sign me up... --Bill
________________________________
From: Steve Lafferty <steve at tronola.com>
To: General Discussion: On-Topic Posts Only <cctech at classiccmp.org>
Sent: Tuesday, March 4, 2014 5:50 AM
Subject: Interest in a PDP-8/e/a/f/m 32K RAM kit?
I've heard from a fellow, Vince S., who has created a PCB layout for the PDP-8/e 32K
RAM design presented in my article on Tronola.com. He's considering a Kickstarter
launch to offer a $95 kit of parts and PCB but is wondering if there is sufficient
interest. If so, please let it be known.
Features include:
? - Battery backup with estimated 10-year data retention.
? - Inhibit on power fail to prevent memory corruption.
? - Most of the full size Omnibus board offers a prototyping layout.
? - Cooperation with any mix of core memory boards. Switches disable 4K fields.
? - Compatibility with Omnibus specs.
? - All through-hole parts, for easy assembly.
I'm not associated with the project, other than having published the article and
encouraging his efforts. Vince has published lots of PDP-8-related and other layouts on
his website and will make this one available as well:
http://so-much-stuff.com/pdp8/cad/cad.php
The article is at:
http://www.tronola.com/html/ram_for_pdp-8e.html
Thanks,
Steve Lafferty
http://www.tronola.com/
11:09 a.m.
New subject: Interest in a PDP-8/e/a/f/m 32K RAM kit?
It is tempting at that price point.
With risk of increasing the price, wouldn't be nice to include Philipps
"OmniUSB" in the design? There is plenty of room for it.
Perhaps room for a daughterboard with an FPGA?
What I'm thinking is that if you are doing the effort to manufacture
the PCB and source the bus drivers you might as well plan ahead for
other goodies that can share the bus. Perhaps that is it, a "local"
Omnibus if possible?
/P
On Tue, Mar 04, 2014 at 08:50:24AM -0500, Steve Lafferty wrote:
I've heard from a fellow, Vince S., who has created a PCB layout for the PDP-8/e 32K
RAM design presented in my article on Tronola.com. He's considering a Kickstarter
launch to offer a $95 kit of parts and PCB but is wondering if there is sufficient
interest. If so, please let it be known.
Features include:
- Battery backup with estimated 10-year data retention.
- Inhibit on power fail to prevent memory corruption.
- Most of the full size Omnibus board offers a prototyping layout.
- Cooperation with any mix of core memory boards. Switches disable 4K fields.
- Compatibility with Omnibus specs.
- All through-hole parts, for easy assembly.
I'm not associated with the project, other than having published the article and
encouraging his efforts. Vince has published lots of PDP-8-related and other layouts on
his website and will make this one available as well:
http://so-much-stuff.com/pdp8/cad/cad.php
The article is at:
http://www.tronola.com/html/ram_for_pdp-8e.html
Thanks,
Steve Lafferty
http://www.tronola.com/
1:22 p.m.
New subject: Interest in a PDP-8/e/a/f/m 32K RAM kit?
On Tue, Mar 4, 2014 at 1:09 PM, Al Kossow <aek at bitsavers.org> wrote:
On 3/4/14 11:09 AM, Pontus Pihlgren wrote:
Unfortunately the stated input threshold calculations for the CY62256 and
74ABT245 are entirely meaningless. Having a chip with Vil and Vih spec of
0.8V and 2.2V does NOT imply that the threshold is 1.5V. I don't know
where to expect the theshold on those particular chips, but in general the
threshold for CMOS tends toward 1/2 Vcc. Obviously to guarantee Vih at
2.2V, they've biased it to move the threshold, but that doesn't necessarily
make it close to (Vil+Vih)/2.
The same reasoning applies to the 74ABT245 with Vil and Vih of 0.8V and
2.0V.
The input thresholds also vary quite a bit with temperature and supply
voltage, which is another reason why (Vil+Vih)/2 is not a reasonable
assumption.
There's a reason DEC specified the Vih of 1.4V, which is that on a heavily
loaded bus it's entirely possible for a valid logic 1 to be that low. In
1983 I tried to debug a large 8/e system which had been donated to a
community college after years of service controlling the scoreboard at
Kauffman Stadium. I often saw the logic 1 on the bus signals lower than
1.6V.
Eric
source the bus drivers
look closely at the design. 
1:03 p.m.
New subject: Interest in a PDP-8/e/a/f/m 32K RAM kit?
On Mar 4, 2014, at 2:09 PM, Pontus Pihlgren <pontus at Update.UU.SE> wrote:
It is tempting at that price point.
With risk of increasing the price, wouldn't be nice to include Philipps
"OmniUSB" in the design? There is plenty of room for it.
Perhaps room for a daughterboard with an FPGA?
What I'm thinking is that if you are doing the effort to manufacture
the PCB and source the bus drivers you might as well plan ahead for
other goodies that can share the bus. Perhaps that is it, a "local"
Omnibus if possible?
Beware of premature optimization, it is the root of all evil. And
so is trying to start with a product that's something for everyone.
Get the thing working, THEN pile features onto the next spin or a
subsequent version so that people can at least have something to
play with while you work on it. Otherwise, in my experience, you
wind up an in infinite loop of "well, what if I add THIS?" and
never actually ship anything. It's my personal Achilles' heel,
and an urge that I have to constantly suppress (with varying
degrees of success).
As far as the bus transceivers, it looks like he's using 26S10
ICs, which are at least still actively sourced by TI. I don't have
any experience with Omnibus, but DEC generally had the same specs
electrically for most of their buses, and the 26S10 violates a few
of the Qbus rules, at least (threshold should be ~1.5v, transitions
are way too fast, leakage is pretty high), but as Philipp's board
has shown (at least anecdotally), you can violate the specs by a
fair margin and have it work in a lot of cases.
No one seems to manufacture a single transceiver IC these days that
complies with the DEC specs; the main tripping point seems to be
the threshold, followed closely by the transition times. You can
fix the former with some expensive ($1/gate or more) analog
comparators, and you can fix the latter by driving output with
selected transistors, but most people aren't going to invest the
money or board layout effort for that.
- Dave
7:21 p.m.
New subject: Interest in a PDP-8/e/a/f/m 32K RAM kit?
Various replies to various comments...
From: Pontus Pihlgren: Tuesday, March 04, 2014 11:09 AM
It is tempting at that price point.
With risk of increasing the price, wouldn't be nice to include Philipps
"OmniUSB" in the design? There is plenty of room for it.
One of the constraints of a hardware Kickstarter, which I can
agree with in principle, is that there must exist a working
prototype. So, my PCB will be based on a drawing of Steve's
prototype, without substantive change. That also effectively
rules out FRAM, etc. My drawing does include lots of
prototype area, though, just like Steve's prototype.
On Mar 4, 2014, at 2:20 PM, Alan Hightower wrote:
Most people I know who restore old machines, including
me, have
an irrational disdain for batteries.
As Tony has pointed out, one needn't, and probably shouldn't,
put the batteries near anything important. I've laid the board
out like Steve's prototype, with the circuitry near the edge
connectors, and the battery near the handles. If that's not
far enough away, then there is Tony's suggestion of a cable
and a box :-).
From: David Riley: Tuesday, March 04, 2014 1:03 PM
As far as the bus transceivers, it looks like he's
using 26S10
ICs, which are at least still actively sourced by TI.
I want to be clear that this is Steve's design, not one of mine.
While mine typically do use the 26S10, Steve's gone a different
direction, which he nicely describes on his page. His design
has the distinct advantage of a working prototype.
From: Eric Smith: Tuesday, March 04, 2014 1:22 PM
There's a reason DEC specified the Vih of 1.4V,
which is that on a heavily
loaded bus it's entirely possible for a valid logic 1 to be that low. In
1983 I tried to debug a large 8/e system which had been donated to a
community college after years of service controlling the scoreboard at
Kauffman Stadium. I often saw the logic 1 on the bus signals lower than
1.6V.
Without getting into the whole can of worms about bus driver
choices, the prototype works, and that's a big deal. Maybe it
doesn't work if you move your DMA device to the other end
of a 20 foot ribbon cable -- I doubt that's been tested. On
the other hand, you shouldn't need to put it there, as this
thing potentially frees up a lot of slots. I should say that
only the memory pins, not the whole Omnibus, are playing
fast and loose here, and the only devices connected to those
should be CPU, memory, and maybe a DMA device. (Does
anyone actually put more than one data break device on
their Omnibus?)
From: Alan Hightower: Tuesday, March 04, 2014 5:05 PM
FYI, a normal ENIG finish is very thin. Most board
houses will charge
extra for hard gold fingers which are designed for insertion durability.
These boards would have proper gold edge connectors, silk screen,
and solder mask, plated through holes, etc. I've had some experience
with the board house, and their stuff has always been great.
4 layers is not necessary as most signals on the bus
transition or are
registered sub 1 MHz.
Agreed. These would be 2 layer boards, with 16 mil signal traces, 32 mil
power and ground traces (well, battery power is 16 mil), and a fair amount
of polygon area for power and ground as well.
Vince
8:06 p.m.
New subject: Interest in a PDP-8/e/a/f/m 32K RAM kit?
I will commit for for three.
Dave
On Tuesday, March 4, 2014 9:28 PM, Vincent Slyngstad <vrs at msn.com> wrote:
Various replies to various comments...
From: Pontus Pihlgren: Tuesday, March 04, 2014 11:09 AM
It is tempting at that price point.
With risk of increasing the price, wouldn't be nice to include Philipps
"OmniUSB" in the design? There is plenty of room for it.
One of the constraints of a hardware Kickstarter, which I can
agree with in principle, is that there must exist a working
prototype.? So, my PCB will be based on a drawing of Steve's
prototype, without substantive change.? That also effectively
rules out FRAM, etc.? My drawing does include lots of
prototype area, though, just like Steve's prototype.
On Mar 4, 2014, at 2:20 PM, Alan Hightower wrote:
Most people I know who restore old machines,
including me, have
an irrational disdain for batteries.
As Tony has pointed out, one needn't, and probably shouldn't,
put the batteries near anything important.? I've laid the board
out like Steve's prototype, with the circuitry near the edge
connectors, and the battery near the handles.? If that's not
far enough away, then there is Tony's suggestion of a cable
and a box :-).
From: David Riley: Tuesday, March 04, 2014 1:03 PM
As far as the bus transceivers, it looks like
he's using 26S10
ICs, which are at least still actively sourced by TI.
I want to be clear that this is Steve's design, not one of mine.
While mine typically do use the 26S10, Steve's gone a different
direction, which he nicely describes on his page.? His design
has the distinct advantage of a working prototype.
From: Eric Smith: Tuesday, March 04, 2014 1:22 PM
There's a reason DEC specified the Vih of
1.4V, which is that on a heavily
loaded bus it's entirely possible for a valid logic 1 to be that low.? In
1983 I tried to debug a large 8/e system which had been donated to a
community college after years of service controlling the scoreboard at
Kauffman Stadium. I often saw the logic 1 on the bus signals lower than
1.6V.
Without getting into the whole can of worms about bus driver
choices, the prototype works, and that's a big deal.? Maybe it
doesn't work if you move your DMA device to the other end
of a 20 foot ribbon cable -- I doubt that's been tested.? On
the other hand, you shouldn't need to put it there, as this
thing potentially frees up a lot of slots.? I should say that
only the memory pins, not the whole Omnibus, are playing
fast and loose here, and the only devices connected to those
should be CPU, memory, and maybe a DMA device.? (Does
anyone actually put more than one data break device on
their Omnibus?)
From: Alan Hightower: Tuesday, March 04, 2014 5:05 PM
FYI, a normal ENIG finish is very thin.? Most
board houses will charge
extra for hard gold fingers which are designed for insertion durability.
These boards would have proper gold edge connectors, silk screen,
and solder mask, plated through holes, etc.? I've had some experience
with the board house, and their stuff has always been great.
4 layers is not necessary as most signals on the
bus transition or are
registered sub 1 MHz.
Agreed.? These would be 2 layer boards, with 16 mil signal traces, 32 mil
power and ground traces (well, battery power is 16 mil), and a fair amount
of polygon area for power and ground as well.
? ? Vince?
6 Mar
6 Mar
7:07 a.m.
New subject: Interest in a PDP-8/e/a/f/m 32K RAM kit?
I will as well.
Todd
* Dave <dfnr2 at yahoo.com> [140304 23:07]:
I will commit for for three.
Dave
On Tuesday, March 4, 2014 9:28 PM, Vincent Slyngstad <vrs at msn.com> wrote:
Various replies to various comments...
>
>From: Pontus Pihlgren: Tuesday, March 04, 2014 11:09 AM
>> It is tempting at that price point.
>>
>> With risk of increasing the price, wouldn't be nice to include Philipps
>> "OmniUSB" in the design? There is plenty of room for it.
>
>One of the constraints of a hardware Kickstarter, which I can
>agree with in principle, is that there must exist a working
>prototype.? So, my PCB will be based on a drawing of Steve's
>prototype, without substantive change.? That also effectively
>rules out FRAM, etc.? My drawing does include lots of
>prototype area, though, just like Steve's prototype.
>
>On Mar 4, 2014, at 2:20 PM, Alan Hightower wrote:
>> Most people I know who restore old machines, including me, have
>> an irrational disdain for batteries.
>
>As Tony has pointed out, one needn't, and probably shouldn't,
>put the batteries near anything important.? I've laid the board
>out like Steve's prototype, with the circuitry near the edge
>connectors, and the battery near the handles.? If that's not
>far enough away, then there is Tony's suggestion of a cable
>and a box :-).
>
>From: David Riley: Tuesday, March 04, 2014 1:03 PM
>> As far as the bus transceivers, it looks like he's using 26S10
>> ICs, which are at least still actively sourced by TI.
>
>I want to be clear that this is Steve's design, not one of mine.
>While mine typically do use the 26S10, Steve's gone a different
>direction, which he nicely describes on his page.? His design
>has the distinct advantage of a working prototype.
>
>From: Eric Smith: Tuesday, March 04, 2014 1:22 PM
>> There's a reason DEC specified the Vih of 1.4V, which is that on a heavily
>> loaded bus it's entirely possible for a valid logic 1 to be that low.? In
>> 1983 I tried to debug a large 8/e system which had been donated to a
>> community college after years of service controlling the scoreboard at
>> Kauffman Stadium. I often saw the logic 1 on the bus signals lower than
>> 1.6V.
>
>Without getting into the whole can of worms about bus driver
>choices, the prototype works, and that's a big deal.? Maybe it
>doesn't work if you move your DMA device to the other end
>of a 20 foot ribbon cable -- I doubt that's been tested.? On
>the other hand, you shouldn't need to put it there, as this
>thing potentially frees up a lot of slots.? I should say that
>only the memory pins, not the whole Omnibus, are playing
>fast and loose here, and the only devices connected to those
>should be CPU, memory, and maybe a DMA device.? (Does
>anyone actually put more than one data break device on
>their Omnibus?)
>
>From: Alan Hightower: Tuesday, March 04, 2014 5:05 PM
>> FYI, a normal ENIG finish is very thin.? Most board houses will charge
>> extra for hard gold fingers which are designed for insertion durability.
>
>These boards would have proper gold edge connectors, silk screen,
>and solder mask, plated through holes, etc.? I've had some experience
>with the board house, and their stuff has always been great.
>
>> 4 layers is not necessary as most signals on the bus transition or are
>> registered sub 1 MHz.
>
>Agreed.? These would be 2 layer boards, with 16 mil signal traces, 32 mil
>power and ground traces (well, battery power is 16 mil), and a fair amount
>of polygon area for power and ground as well.
>
>
>? ? Vince?
>
>
>
4 Mar
4 Mar
8:59 p.m.
New subject: Interest in a PDP-8/e/a/f/m 32K RAM kit?
On Mar 4, 2014 8:25 PM, "Vincent Slyngstad" <vrs at msn.com> wrote:
prototype
works, and that's a big deal. Maybe it doesn't work if you move
your DMA device to the other end of a 20 foot ribbon cable -- I doubt
that's been tested. On the other hand, you shouldn't need to put it there,
as this thing potentially frees up a lot of slots. I should say that only
the memory pins, not the whole Omnibus, are playing fast and loose here,
and the only devices connected to those should be CPU, memory, and maybe a
DMA device. (Does anyone actually put more than one data break device on
their Omnibus?)
My point isn't that there's anything wrong with making a board that doesn't
completely meet DEC specs, as long as that's understood. The problem is
trying to represent something as meeting DEC specs when it probably
doesn't, and using an incorrect analysis to try to support the claim that
it meets DEC specs.
Yes, it will most likely work in small or even medium-scale Omnibus systems.
No, it might not work reliably in a large system, such as I previously
described, which did have multiple data break devices. The length of the
bus is also a factor, even if you have few or no data break devices. The
system in question had one 40- and two 20-slot boxes, with fairly long
cables. To some extent the long cables may actually be better than short
ones.
If DEC didn't spec the maximum slew rate, that was probably because at the
time there weren't any cost-effective interface chips that had fast enough
slew rate to be a problem. There certainly are now, and the lack of a spec
doesn't mean that it's OK to use arbitrarily high slew rate drivers.
All that said, I'm happy to see the design published, and boards and/or
kits made available. I know what a large amount of work that can be. I'd
be happy to buy two of them, knowing that they most likely don't actually
meet DEC specs.
Eric
From: Eric Smith: Tuesday, March 04, 2014 1:22 PM
> There's a reason DEC specified the Vih of 1.4V, which is that on a
heavily
loaded bus
it's entirely possible for a valid logic 1 to be that low. In
1983 I tried to debug a large 8/e system which had been donated to a
community college after years of service controlling the scoreboard at
Kauffman Stadium. I often saw the logic 1 on the bus signals lower than
1.6V.
Without getting into the whole can of worms about bus driver choices, the 
19 Mar
19 Mar
4:18 a.m.
New subject: Interest in a PDP-8/e/a/f/m 32K RAM kit?
Am 05.03.2014 05:59, schrieb Eric Smith:
Yes, it will most likely work in small or even
medium-scale Omnibus systems.
Blah!
Who has a really large scale Omnibus system?!?
These discussions are sooooo annoying! There are not that many people
who make new Omnibus stuff today. What is the reason for discouraging them?
Everybody knows now that it is nearly impossible to exactly meet DEC
specs. And everybody knows that even in the old days companies sold
"incompatible" boards which exposed better overall performance than the
original DEC stuff.
Someone should count the frequency of bus driver discussions on this
list. Or even declare bus drivers as off topic.
Sorry for the yelling....
9:54 a.m.
New subject: Interest in a PDP-8/e/a/f/m 32K RAM kit?
On Mar 19, 2014 5:33 AM, "Philipp Hachtmann" <hachti at hachti.de> wrote:
These discussions are sooooo annoying! There are not
that many people who
make new Omnibus stuff today. What is the reason for
discouraging them?
It's not a matter of trying to discourage anyone from making, selling, or
buying such things. It's a matter of full disclosure.
Suppose you bought a "SOHO" Ethernet switch and it didn't work for you,
because the manufacturer knowingly made it in such a way that it didn't
meet the Ethernet specifications and couldn't work with cables longer than
25m. They didn't bother to document that limitation bacause it is obvious
that almost no one needs Ethernet cables longer than 25m at home, so they
conidered it totally unimportant that it didn't meet the specifications.
Would you say that it was OK for them to do that? Would you criticize a
third party who tried to publicize the problem, because that person was
trying to "discourage" the vendor?
Now consider the same question if the vendor explicitly claimed that it did
meet the specification, and put on their web site a well-intentioned but
flawed analysis of that claim. Would you complain about someone that
pointed out the flaw in the analysis?
As I've said, I'm delighted to see people making new products for use
Omnibus, and I'll buy some myself, but I think it is entirely appropriate
to point out that that they almost certainly don't meet the bus specs, even
though that probably will affect only a small fraction of users.
20 Mar
20 Mar
4:55 a.m.
New subject: No more bus driver discussion, PLEASE!
Am 19.03.2014 17:54, schrieb Eric Smith:
but I think it is entirely appropriate
to point out that that they almost certainly don't meet the bus specs, even
though that probably will affect only a small fraction of users.
Who is that fraction? For me it currently looks like the empty set. So
for me the bus driver discussion is as much of secondary interest as it
could be.
And I insist: Throwing up this bus driver thing again and again
discourages people.
The bus driver yelling fraction could consider to invest its capacities
into more productive aspects of criticism...
:-)
And BTW: Be informed that OmniUSB doesn't even trie to meet the DEC
specs. But it seems to work well so far.
10:05 a.m.
New subject: No more bus driver discussion, PLEASE!
On Thu, Mar 20, 2014 at 12:55:14PM +0100, Philipp Hachtmann wrote:
So for me the bus driver discussion is as much of
secondary interest
as it could be.
So ... why not skip those posts? You seem weirdly hostile about it.
John Wilson
10:24 a.m.
New subject: No more bus driver discussion, PLEASE!
On Thu, Mar 20, 2014 at 12:05 PM, John Wilson <wilson at dbit.com> wrote:
On Thu, Mar 20, 2014 at 12:55:14PM +0100, Philipp
Hachtmann wrote:
I personally have used DEC and non-DEC boards in my PDP-8/E and PDP-8/M
that don't have the "correct" bus drivers with great results, and my
sentiments are shared by many, including Philipp.
Has anyone actually been affected by this issue with large machines? If so,
what bus drivers were being used?
Kyle
So for me the bus driver discussion is as much of
secondary interest
as it could be.
So ... why not skip those posts?
20 Mar
20 Mar
10:19 a.m.
New subject: No more bus driver discussion, PLEASE!
On Mar 20, 2014 6:00 AM, "Philipp Hachtmann" <hachti at hachti.de> wrote:
In this thread several people with relatively large Omnibus machine
configurations have self-identified. These are people who are potentially
affected by this issue.
Am 19.03.2014 17:54, schrieb Eric Smith:
> but I think it is entirely appropriate
> to point out that that they almost certainly don't meet the bus specs,
even
though that
probably will affect only a small fraction of users.
Who is that fraction? For me it currently looks like the empty set. And BTW: Be informed that OmniUSB doesn't even
trie to meet the DEC
specs. But it seems to work well so far.
And since you've been up front about that from the start, no one has, as
you put it, "yelled" about it.
22 Mar
22 Mar
3:55 p.m.
New subject: No more bus driver discussion, PLEASE!
On 20.03.2014 18:19, Eric Smith wrote:
In this thread several people with relatively large
Omnibus machine
configurations have self-identified.
What means relatively large Omnibus
configurations? My test machine here has two
backplanes, nearly full with different stuff. No expansion.
These are people who are potentially
affected by this issue.
But are they? It would be great to know if real issues
exist. Not theoretically.
Nobody denies that the perfect drivers would probably work better in
configurations near the maximum. But that should be something like three to four
boxes with two backplanes each, filled with cards... And not to forget a many
feet long cable between them...
And since you've been up front about that from the
start, no one has, as
you put it, "yelled" about it.
Oh, probably yelling is the wrong word
(English is not my primary language!).
And not "at me" - but those bus driver discussions... just don't like them.
Everything about bus drivers seems to be said. And has been said long before I
even had the idea to build something like the USB board.
All that could be interesting in that area would be practical experience with
different drivers, hardware configurations....
:-)
6:02 p.m.
New subject: No more bus driver discussion, PLEASE!
From: Philipp Hachtmann: Saturday, March 22, 2014 3:55 PM
If I am understanding the configuration guide at
http://www.bitsavers.org/pdf/dec/pdp8/omnibus/EK-0PDP8-SP-001_PDP-8_Family_…
correctly, the maximal configuration consists of an 8/E with it's two
internal Omnibus backplanes, cabled to a BA8-C cabinet containing
another Omnibus backplane. That seems to give 60 slots, with 51
slots left over once you count the CPU, front panel, and backplane
interconnects.
Interestingly, the configuration with two 8/E boxes does not seem
to be allowed. Perhaps 80 slots was just too much?
In this thread
several people with relatively large Omnibus machine
configurations have self-identified.
What means relatively large Omnibus
configurations? My test machine here has
two backplanes, nearly full with different stuff. No expansion. Nobody denies that the perfect drivers would probably
work better in
configurations near the maximum. But that should be something like three to
four boxes with two backplanes each, filled with cards... And not to forget a
many feet long cable between them...
Apparently Omnibus systems with more than 2 boxes were not
guaranteed to work. I'm not sure what lengths the BC80C cable
assembly came in.
Vince
11:21 p.m.
New subject: No more bus driver discussion, PLEASE!
From: Vincent Slyngstad: Saturday, March 22, 2014 6:02 PM
Apparently Omnibus systems with more than 2 boxes were
not
guaranteed to work. I'm not sure what lengths the BC80C cable
assembly came in.
Found it. A BC80C cable is apparently about 6 feet from edge
connector to edge connector, with about 4 feet of flexprint
between the paddle cards.
http://bitsavers.trailing-edge.com/pdf/dec/pdp8/pdp8a/MP00140_BA8-C_Unit_As…
(page 56 of 60)
So the total length of an Omnibus wasn't supposed to be terribly
long.
Vince
23 Mar
23 Mar
1:03 p.m.
New subject: No more bus driver discussion, PLEASE!
----- Original Message -----
From: "Vincent Slyngstad" <vrs at msn.com>
To: "General Discussion: On-Topic and Off-Topic Posts"
<cctalk at classiccmp.org>
Sent: Sunday, March 23, 2014 6:21 AM
Subject: Re: No more bus driver discussion, PLEASE!
From: Vincent Slyngstad: Saturday, March 22, 2014 6:02
PM
Apparently the requirement for expansion of the omnibus via cable was a
rather late extra, well after the basic spec had been set and the design
done, it mainly involved a rework of the bus loads.
see tale 14 of this very interesting artical by one of the main PDP8E
design engineeres.
http://www.ieeeghn.org/wiki/index.php/First-Hand:PDP-8/E_OMNIBUS_Ride
Dave H
Apparently Omnibus systems with more than 2 boxes
were not
guaranteed to work. I'm not sure what lengths the BC80C cable
assembly came in.
Found it. A BC80C cable is apparently about 6 feet from edge
connector to edge connector, with about 4 feet of flexprint
between the paddle cards.
http://bitsavers.trailing-edge.com/pdf/dec/pdp8/pdp8a/MP00140_BA8-C_Unit_As…
(page 56 of 60)
So the total length of an Omnibus wasn't supposed to be terribly
long.
Vince
-----
No virus found in this message.
Checked by AVG - www.avg.com
Version: 2012.0.2247 / Virus Database: 3722/6731 - Release Date: 03/22/14
5:11 a.m.
New subject: No more bus driver discussion, PLEASE!
On Sat, Mar 22, 2014 at 7:02 PM, Vincent Slyngstad <vrs at msn.com> wrote:
Interestingly, the configuration with two 8/E boxes
does not seem
to be allowed. Perhaps 80 slots was just too much?
The BA8-A System Expander Box came with a single 20-slot OMNIBUS backplane,
but you could add (as a fully supported configuration) a BE8-A OMNIBUS
Expander, which added another 20-slot backplane and M935 bus jumpers, for a
total of 80 slots (six taken up by bus jumpers and cables).
Three systems I worked with were the 80 slot configuration. Two were
originally sold by DEC in that configuration, and one was field upgraded by
DEC.
It was not uncommon to see the BA8-A expander used with the PDP-8/A, though
any hex-width options had to live in the 8/A chassis, since the BA8-A only
accommodates quad width.
Eric
9:37 a.m.
New subject: No more bus driver discussion, PLEASE!
From: Eric Smith: Sunday, March 23, 2014 5:11 AM
On Sat, Mar 22, 2014 at 7:02 PM, Vincent Slyngstad
<vrs at msn.com> wrote:
I can't seem to find where this was officially sanctioned, but it does seem
that it was done often enough.
Since the BC08H cable came in 3 and 4 foot lengths, it would seem then
that the maximal Omnibus is two 40 slot 8/E chassis interconnected with
a 4 foot flexprint cable. (Each 8/E chassis with two 20 slot BE8-A backplanes
connected by M935 interconnecting cards.) Presumably with 70 slots
worth of option cards.
Vince
Interestingly, the configuration with two 8/E
boxes does not seem
to be allowed. Perhaps 80 slots was just too much?
The BA8-A System Expander Box came with a single 20-slot OMNIBUS backplane,
but you could add (as a fully supported configuration) a BE8-A OMNIBUS
Expander, which added another 20-slot backplane and M935 bus jumpers, for a
total of 80 slots (six taken up by bus jumpers and cables).
3:30 p.m.
New subject: No more bus driver discussion, PLEASE!
On Sun, Mar 23, 2014 at 10:37 AM, Vincent Slyngstad <vrs at msn.com> wrote:
I can't seem to find where this was officially
sanctioned, but it does
seem that it was done often enough.
One of the handbooks specifically says that you can do that. I'm pretty
sure the BA8-A manual said that as well, but I no longer have that.
4:51 a.m.
New subject: No more bus driver discussion, PLEASE!
At 06:55 PM 3/22/2014, Philipp Hachtmann wrote:
Everything about bus drivers seems to be said. And has
been said long before I
even had the idea to build something like the USB board.
All that could be interesting in that area would be practical experience with
different drivers, hardware configurations....
--- Hi Philipp, I agree with you that info on practical experience with various
configurations and modern bus interfaces would be interesting. However, I for one, have
benefited from discussion here of bus drivers/receivers. As a result of the questions
raised, I did the additional research which revealed that DEC had made prominent usage of
ordinary TTL gates as Omnibus receivers. I believe that fact wasn't widely known.
Since the modern bus transceivers meet or exceed the output current, output voltage levels
and input loading of the best DEC parts, the only remaining questionable specs were the
receiver input levels. Showing that DEC endorsed TTL input levels for Omnibus receiving
established modern transceivers as fully qualified for the RAM Board project.
The positive result which has emerged is that there is no reason to suspect that Omnibus
interfacing with modern parts (at least in applications similar to the RAM Board) is at a
disadvantage relative to DEC practice.
Steve Lafferty
20 Mar
20 Mar
11:16 a.m.
New subject: No more bus driver discussion, PLEASE!
From: Philipp Hachtmann: Thursday, March 20, 2014 4:55 AM
I think your perspective may be Omnibus-centric. I basically agree
with you for the Omnibus, Posibus, and Posibus Data Break busses.
Those busses are relatively poorly specified, and generally fairly
short. Indeed, the Omnibus configuration guide only discusses
systems with 1 or 2 boxes.
However, discussions of bus drivers for Unibus and Qbus still seem
relevant.
Vince
but I think it
is entirely appropriate
to point out that that they almost certainly don't meet the bus specs, even
though that probably will affect only a small fraction of users.
Who is that fraction? For me it currently looks like the empty set. So
for me the bus driver discussion is as much of secondary interest as it
could be.
And I insist: Throwing up this bus driver thing again and again
discourages people. 
11:55 a.m.
New subject: No more bus driver discussion, PLEASE!
On Mar 20, 2014, at 11:16 AM, Vincent Slyngstad <vrs at msn.com> wrote:
From: Philipp Hachtmann: Thursday, March 20, 2014 4:55
AM
I think your perspective may be Omnibus-centric. I basically agree with you for the
Omnibus, Posibus, and Posibus Data Break busses.
Those busses are relatively poorly specified, and generally fairly short. Indeed, the
Omnibus configuration guide only discusses systems with 1 or 2 boxes.
I haven't done a lot with Omnibus stuff (yet), so I don't have a lot of experience
with it. But I tend to be conservative with designs and will probably do whatever DEC did
on their boards as far as drivers go. ;-)
but I
think it is entirely appropriate
to point out that that they almost certainly don't meet the bus specs, even
though that probably will affect only a small fraction of users.
Who is that
fraction? For me it currently looks like the empty set. So for me the bus driver
discussion is as much of secondary interest as it could be.
And I insist: Throwing up this bus driver thing again and again discourages people.
However, discussions of bus drivers for Unibus and Qbus still seem relevant.
Given that I've had *loads* of issues related to BC11A cables (still planning on
making some of my own), I'd be really hesitant to use anything other than the DEC
specified parts.
TTFN - Guy
2:35 p.m.
New subject: No more bus driver discussion, PLEASE!
On Thu, Mar 20, 2014 at 2:55 PM, Guy Sotomayor <ggs at shiresoft.com> wrote:
Given that I've had *loads* of issues related to
BC11A cables (still planning on making some of my own), I'd be really hesitant to use
anything other than the DEC specified parts.
Some vendors did use non-DEC parts (ISTR seeing 74LS240s as Qbus
address/data transceivers). That doesn't mean the boards worked in
multiple-cabinet arrangements. At Software Results, we paid the big
bucks and used real DEC DC013s and DC005s and Natl Semi 8641s, etc.,
in all our products. We were already a 3rd-party product and didn't
need any fingers pointed at us when things didn't work right (the only
issue I ever remember was related to bus timeout issues in the last
Qbus VAXen that had a DLT drive on a particular DEC controller - our
board wasn't tolerant enough of the SCSI controller holding the bus
longer than any other DEC product). We certainly didn't want to play
the game of "it doesn't work - must be your random bus drivers causing
it" and having to prove it right or wrong.
Speaking for myself, I read each and every message on this list as it
relates to classic DEC busses and how to make modern peripherals for
them. I don't always have a comment to make, but I read each and
every one to see if we've found a way to make "in spec" peripherals.
Carry on
-ethan
21 Mar
21 Mar
5:35 a.m.
New subject: No more bus driver discussion, PLEASE!
On 03/20/2014 02:16 PM, Vincent Slyngstad wrote:
From: Philipp Hachtmann: Thursday, March 20, 2014 4:55
AM
I think your perspective may be Omnibus-centric. I basically agree
with you for the Omnibus, Posibus, and Posibus Data Break busses.
Those busses are relatively poorly specified, and generally fairly
short. Indeed, the Omnibus configuration guide only discusses systems
with 1 or 2 boxes.
However, discussions of bus drivers for Unibus and Qbus still seem
relevant.
Vince
What is being forgotten here is that even DEC had problems around this.
Using the
DEC parts did NOT assure nominal operation. It leads to both QA and a
whole process
creation called FA/T (factory assemble and test) because machines were
being configured
sent to the field and while it all plugged in it far to often didn't
work. Sometimes it was bad
boards and cables but more often it was what DEC would call in later years
"unsupported configurations". People that build and use multi box
systems understand
this.
Omnibus wasn't well specified and back then things like slew rates were
just becoming
a question and far from controlled. Same for the Utililogic and TTL
parts. Multi cabinet
systems were because boards were big and technology hadn't made them
smaller but
things like propagation delay, crosstalk, and cable lengths were
issues. Its up side
Omnibuss made it easy for users to make their own boards and it was
often done. DEC
made kits to aid that process but people did often go off the campus and
do it their way
sometimes good and occasionally not.
The fact that we have seen a few useful boards and an attempt to put in
parts that should
work fine is critical. A simple thing like memory has stopped many
pdp-8s from seeing
use or even bother to power them up.
Allison/An old mill rat
but I
think it is entirely appropriate
to point out that that they almost certainly don't meet the bus
specs, even
though that probably will affect only a small fraction of users.
Who is that fraction? For me it currently looks like the empty set.
So for me the bus driver discussion is as much of secondary interest
as it could be.
And I insist: Throwing up this bus driver thing again and again
discourages people.
4 Mar
4 Mar
11:20 p.m.
New subject: Interest in a PDP-8/e/a/f/m 32K RAM kit?
On Tue, Mar 04, 2014 at 07:21:24PM -0800, Vincent Slyngstad wrote:
My drawing does include lots of prototype area,
though, just like
Steve's prototype.
I fold :) I'm interested to see what people will do with that area.
/P
5 Mar
5 Mar
12:21 p.m.
New subject: Interest in a PDP-8/e/a/f/m 32K RAM kit?
As Tony has pointed out, one needn't, and probably
shouldn't,
put the batteries near anything important. I've laid the board
out like Steve's prototype, with the circuitry near the edge
connectors, and the battery near the handles. If that's not
far enough away, then there is Tony's suggestion of a cable
and a box :-).
It it were going in my system [1]. that I wouldn't want to do even that.
In the 8/e. the boards go in vertically, with the handles at the top. So
a leaking battery could drip down the board and into the backplane. I
would vertianly fit the battery remotely.
[1] It won't be, for the simple reason I have the full 32K words of core
in mine.
-tony
19 Mar
19 Mar
4:13 a.m.
New subject: Interest in a PDP-8/e/a/f/m 32K RAM kit?
Am 05.03.2014 04:21, schrieb Vincent Slyngstad:
(Does anyone actually put more than one
data break device on their Omnibus?)
Yes, me!
RK8E controller on the Omnibus and FPP12 via positive bus interface and
data break interface boards :-) :-) :-)
6:31 a.m.
New subject: Interest in a PDP-8/e/a/f/m 32K RAM kit?
From: Philipp Hachtmann: Wednesday, March 19, 2014 4:13 AM
Am 05.03.2014 04:21, schrieb Vincent Slyngstad:
Good point! I even have an FPP12, though my hope is to someday get
it working with my 8/I, and not my Omnibus gear.
Vince
(Does anyone actually put more than one
data break device on their Omnibus?)
Yes, me!
RK8E controller on the Omnibus and FPP12 via positive bus interface and
data break interface boards :-) :-) :-)
10:42 a.m.
New subject: Interest in a PDP-8/e/a/f/m 32K RAM kit?
On Tue, Mar 04, 2014 at 07:21:24PM -0800, Vincent Slyngstad wrote:
(Does anyone actually
put more than one data break device on their Omnibus?)
Sure -- RF08 and TC08, each with its own KD8E.
(But not powered on in a while.)
John
4 Mar
4 Mar
11:11 a.m.
New subject: Interest in a PDP-8/e/a/f/m 32K RAM kit?
On Mar 4, 2014 10:56 AM, "Steve Lafferty" <steve at tronola.com> wrote:
I've heard from a fellow, Vince S., who has created a PCB layout for the
PDP-8/e 32K RAM design presented in my article on Tronola.com. He's
considering a Kickstarter launch to offer a $95 kit of parts and PCB but is
wondering if there is sufficient interest. If so, please let it be known.
I'll be interested in one or two.
Kyle
11:15 a.m.
New subject: Interest in a PDP-8/e/a/f/m 32K RAM kit?
I'd buy a couple but I think the price ($95) is too low given the size of
the board and the # of boards that I think would be ordered (anywhere from
10 - 50).
Marc
On Tue, Mar 4, 2014 at 5:50 AM, Steve Lafferty <steve at tronola.com> wrote:
I've heard from a fellow, Vince S., who has created a PCB layout for the
PDP-8/e 32K RAM design presented in my article on Tronola.com. He's
considering a Kickstarter launch to offer a $95 kit of parts and PCB but is
wondering if there is sufficient interest. If so, please let it be known.
Features include:
- Battery backup with estimated 10-year data retention.
- Inhibit on power fail to prevent memory corruption.
- Most of the full size Omnibus board offers a prototyping layout.
- Cooperation with any mix of core memory boards. Switches disable 4K
fields.
- Compatibility with Omnibus specs.
- All through-hole parts, for easy assembly.
I'm not associated with the project, other than having published the
article and encouraging his efforts. Vince has published lots of
PDP-8-related and other layouts on his website and will make this one
available as well:
http://so-much-stuff.com/pdp8/cad/cad.php
The article is at:
http://www.tronola.com/html/ram_for_pdp-8e.html
Thanks,
Steve Lafferty
http://www.tronola.com/
1:05 p.m.
New subject: Interest in a PDP-8/e/a/f/m 32K RAM kit?
I just don't want to see someone doing a good deed get burned. Another
suggestion is that as long as he's doing this it should probably be a 4
layer board. Doesn't increase the price that much and does wonders for
signal integrity.
Marc
On Tue, Mar 4, 2014 at 12:10 PM, Al Kossow <aek at bitsavers.org> wrote:
On 3/4/14 11:15 AM, Marc Howard wrote:
I think the price ($95) is too low
WOW
That's a first on THIS list!
1:49 p.m.
New subject: Interest in a PDP-8/e/a/f/m 32K RAM kit?
On 3/4/14, 3:05 PM, Marc Howard wrote:
I just don't want to see someone doing a good deed
get burned. Another
suggestion is that as long as he's doing this it should probably be a 4
layer board. Doesn't increase the price that much and does wonders for
signal integrity.
4 layer boards at that size are very expensive.
I'd like to complement him on the writeup on his site. Really a great
read and excellent web content, diagrams, etc.
-Ryan
2:37 p.m.
New subject: Interest in a PDP-8/e/a/f/m 32K RAM kit?
|On 3/4/14, 3:05 PM, Marc Howard wrote:
I just don't want to see someone doing a good deed
get burned. Another
suggestion is that as long as he's doing this it should probably be a 4
layer board. Doesn't increase the price that much and does wonders for
signal integrity.
|4 layer boards at that size are very expensive.
Depends. About 2x which is fair given the you've double the # of layers.
A $20 board would now cost $40. The signal integrity of boards with
integral ground planes is much better and so is routing.
I never look at that end of the segment where I work. Most of our boards
are about 500mm x 600mm, have 22-28 layers, buried vias, buried caps, etc.
The 1-3 piece price will buy you a nice car.
Marc
On Tue, Mar 4, 2014 at 1:49 PM, Ryan Brooks <ryan at hack.net> wrote:
On 3/4/14, 3:05 PM, Marc Howard wrote:
I just don't want to see someone doing a good
deed get burned. Another
suggestion is that as long as he's doing this it should probably be a 4
layer board. Doesn't increase the price that much and does wonders for
signal integrity.
4 layer boards at that size are very expensive.
I'd like to complement him on the writeup on his site. Really a great
read and excellent web content, diagrams, etc.
-Ryan
4:48 p.m.
New subject: Interest in a PDP-8/e/a/f/m 32K RAM kit?
opinion...
4 layer????
This is absurd. PDP-8 omnibus was and is very tolerant and also by
todays speeds SLOW.
This machine is from the era of 50 mil traces (back then .025" was very
fine lines!).
The logic used was mostly straight TTL which back then was slow and not
terribly noise
immune. I've wire wrapped boards for my 8f (a ram card using 1uS 2102s
all 48 of them).
The board will be fine at two layers if done reasonably. Also it doe
not need to be full height
which saves both copper and cost.
However properly plated fingers is a must. That's Nickel over Copper
then Gold. Why?
My first Altair had a poisoned bus some of the early MITS cards were tin
plate, some
Gold over Nickel and the nasty ones were Gold over Copper. Ever see
gold turn green?
That's what happens you get copper migration and the connectors
especially gold plated
ones get contaminated by that... then its a remove and reinsert before
using forever
or crash. The fix for the Altair was replace the backplane. Not
likely to hppen on a
PDP-8.
Allison
On 03/04/2014 05:37 PM, Marc Howard wrote:
|On 3/4/14, 3:05 PM, Marc Howard wrote:
I just don't want to see someone doing a good
deed get burned. Another
suggestion is that as long as he's doing this it should probably be a 4
layer board. Doesn't increase the price that much and does wonders for
signal integrity.
|4 layer boards at that size are very expensive.
Depends. About 2x which is fair given the you've double the # of layers.
A $20 board would now cost $40. The signal integrity of boards with
integral ground planes is much better and so is routing.
I never look at that end of the segment where I work. Most of our boards
are about 500mm x 600mm, have 22-28 layers, buried vias, buried caps, etc.
The 1-3 piece price will buy you a nice car.
Marc
On Tue, Mar 4, 2014 at 1:49 PM, Ryan Brooks <ryan at hack.net> wrote:
> On 3/4/14, 3:05 PM, Marc Howard wrote:
>
>> I just don't want to see someone doing a good deed get burned. Another
>> suggestion is that as long as he's doing this it should probably be a 4
>> layer board. Doesn't increase the price that much and does wonders for
>> signal integrity.
>>
> 4 layer boards at that size are very expensive.
>
> I'd like to complement him on the writeup on his site. Really a great
> read and excellent web content, diagrams, etc.
>
> -Ryan
>
> 
5:05 p.m.
New subject: Interest in a PDP-8/e/a/f/m 32K RAM kit?
FYI, a normal ENIG finish is very thin. Most board houses will charge
extra for hard gold fingers which are designed for insertion durability.
Most of my ENIG ISA boards wear through with just a few insertions.
HASL is more durable.
4 layers is not necessary as most signals on the bus transition or are
registered sub 1 MHz.
-Alan
On 2014-03-04 19:48, allison wrote:
However properly plated fingers is a must. That's
Nickel over Copper
then Gold. Why?
My first Altair had a poisoned bus some of the early MITS cards were
tin
plate, some
Gold over Nickel and the nasty ones were Gold over Copper. Ever see
gold turn green?
That's what happens you get copper migration and the connectors
especially gold plated
ones get contaminated by that... then its a remove and reinsert before
using forever
or crash. The fix for the Altair was replace the backplane. Not
likely to hppen on a
PDP-8.
Allison
2:26 p.m.
New subject: Interest in a PDP-8/e/a/f/m 32K RAM kit?
On Tue, Mar 4, 2014 at 3:05 PM, Marc Howard <cramcram at gmail.com> wrote:
I just don't want to see someone doing a good deed
get burned. Another
suggestion is that as long as he's doing this it should probably be a 4
layer board. Doesn't increase the price that much and does wonders for
signal integrity.
Why would four layers be a concern, though? All of the boards in my PDP-8/E
and /M boards are two-layer, with manually routed traces and no ground
plane.
It would make sense to add an auto-loader, in my opinion. I have a board
that works well, made by Tennecomp. It can be configured to work with an
external push-button or the SW switch. The original design calls for a
1702A, which I think would be just a bit too difficult for most people to
reprogram these days. My second board has been white-wired to accept a
2716. I'm thinking for ease of use, a socketed EEPROM might be good, but
another neat idea might be an in-circuit serial reprogrammable device.
If folks are interested, I'll see about reverse engineering the board. I
count 34 ICs on the board, along with three transistors, five diodes, and
various resistors and capacitors. Probably could be really reduced with a
PAL Or just put it all on a microcontroller.
Kyle
3:14 p.m.
New subject: Interest in a PDP-8/e/a/f/m 32K RAM kit?
On Tue, Mar 4, 2014 at 2:26 PM, Kyle Owen <kylevowen at gmail.com> wrote:
On Tue, Mar 4, 2014 at 3:05 PM, Marc Howard
<cramcram at gmail.com> wrote:
The low speed and use of standard TTL does cover a multitude of sins. The
problem I worry about is when you mix contemporary devices with old school
logic. What looks like harmless ringing in 70's TTL land can look like a
clock pulse to modern CMOS device. But you're right, the PDP8 stuff works
(for the most part) with two layer boards, rather simple bypassing, and not
much impedance control.
Marc
I just don't want to see someone doing a good
deed get burned. Another
suggestion is that as long as he's doing this it should probably be a 4
layer board. Doesn't increase the price that much and does wonders for
signal integrity.
Why would four layers be a concern, though? All of the boards in my PDP-8/E
and /M boards are two-layer, with manually routed traces and no ground
plane.
1:10 p.m.
New subject: Interest in a PDP-8/e/a/f/m 32K RAM kit?
On Tue, Mar 4, 2014 at 1:15 PM, Marc Howard <cramcram at gmail.com> wrote:
I'd buy a couple but I think the price ($95) is
too low given the size of
the board and the # of boards that I think would be ordered (anywhere from
10 - 50).
Marc
In quantity 50, pcbcart.com is estimating around $20 per board, averaging
in the tooling cost as well. Estimating a BOM of $30 to $50, that's still
below $95 as of now, so I think $95 is quite reasonable.
Kyle
11:20 a.m.
New subject: Interest in a PDP-8/e/a/f/m 32K RAM kit?
I don't think you are going to find enough PDP8 owners for a
kickstarter.
Sounds like an interesting project. However if you include 1 SMT chip,
you could do it with FRAM or MRAM - which are in a way a modern day
equivalent of core memory. A nice link to the past.
Most people I know who restore old machines, including me, have an
irrational disdain for batteries.
On 2014-03-04 08:50, Steve Lafferty wrote:
I've heard from a fellow, Vince S., who has
created a PCB layout for
the PDP-8/e 32K RAM design presented in my article on Tronola.com.
He's considering a Kickstarter launch to offer a $95 kit of parts and
PCB but is wondering if there is sufficient interest. If so, please
let it be known.
Features include:
- Battery backup with estimated 10-year data retention.
- Inhibit on power fail to prevent memory corruption.
- Most of the full size Omnibus board offers a prototyping layout.
- Cooperation with any mix of core memory boards. Switches disable 4K
fields.
- Compatibility with Omnibus specs.
- All through-hole parts, for easy assembly.
I'm not associated with the project, other than having published the
article and encouraging his efforts. Vince has published lots of
PDP-8-related and other layouts on his website and will make this one
available as well:
http://so-much-stuff.com/pdp8/cad/cad.php
The article is at:
http://www.tronola.com/html/ram_for_pdp-8e.html
Thanks,
Steve Lafferty
http://www.tronola.com/
12:38 p.m.
New subject: Interest in a PDP-8/e/a/f/m 32K RAM kit?
On Mar 4, 2014, at 2:20 PM, Alan Hightower <alan at alanlee.org> wrote:
I don't think you are going to find enough PDP8 owners for a kickstarter.
Sounds like an interesting project. However if you include 1 SMT chip, you could do it
with FRAM or MRAM - which are in a way a modern day equivalent of core memory. A nice
link to the past.
Most people I know who restore old machines, including me, have an irrational disdain for
batteries.
Irrational? Given the damage I've seen batteries do when they leak, I'd
hardly call it irrational.
- Dave
2:25 p.m.
New subject: Interest in a PDP-8/e/a/f/m 32K RAM kit?
I don't
think you are going to find enough PDP8 owners for a kickstarter.
Sounds like an interesting project. However if you include 1 SMT
chip, you could do it with FRAM or MRAM - which are in a way a modern
day equivalent of core memory. A nice link to the past.
Most people I know who restore old machines, including me, have an
irrational disdain for batteries.
Irrational? Given the damage I've seen batteries do when they leak, I'd
hardly call it irrational. 
4:43 p.m.
New subject: Interest in a PDP-8/e/a/f/m 32K RAM kit?
Why not use an FRAM (e.g. Cypress FM28V020) and completely dispense with
the batteries? About $15 ea at Mouser. Looks like an SRAM and is 32K x 8.
-Gary
On 03/04/2014 12:38 PM, David Riley wrote:
On Mar 4, 2014, at 2:20 PM, Alan Hightower <alan at
alanlee.org> wrote:
--
-Gary
I don't think you are going to find enough
PDP8 owners for a kickstarter.
Sounds like an interesting project. However if you include 1 SMT chip, you could do it
with FRAM or MRAM - which are in a way a modern day equivalent of core memory. A nice
link to the past.
Most people I know who restore old machines, including me, have an irrational disdain for
batteries.
Irrational? Given the damage I've seen batteries do when they leak,
I'd
hardly call it irrational.
- Dave
4:44 p.m.
New subject: Interest in a PDP-8/e/a/f/m 32K RAM kit?
(Apologies if duplicate - sent original with wrong source email address.)
Why not use an FRAM (e.g. Cypress FM28V020) and completely dispense with
the batteries? About $15 ea at Mouser. Looks like an SRAM and is 32K x 8.
-Gary
On 03/04/2014 12:38 PM, David Riley wrote:
On Mar 4, 2014, at 2:20 PM, Alan Hightower <alan at
alanlee.org> wrote:
--
-Gary
I don't think you are going to find enough
PDP8 owners for a kickstarter.
Sounds like an interesting project. However if you include 1 SMT chip, you could do it
with FRAM or MRAM - which are in a way a modern day equivalent of core memory. A nice
link to the past.
Most people I know who restore old machines, including me, have an irrational disdain for
batteries.
Irrational? Given the damage I've seen batteries do when they leak,
I'd
hardly call it irrational.
- Dave
5:07 p.m.
New subject: Interest in a PDP-8/e/a/f/m 32K RAM kit?
My irrational was a poke at some.
The FRAM would have the be level translated, however 74LVC245s come in
DIPs.
-Alan
On 2014-03-04 19:44, Gary Oliver wrote:
(Apologies if duplicate - sent original with wrong
source email
address.)
Why not use an FRAM (e.g. Cypress FM28V020) and completely dispense
with
the batteries? About $15 ea at Mouser. Looks like an SRAM and is 32K
x 8.
-Gary
On 03/04/2014 12:38 PM, David Riley wrote:
> On Mar 4, 2014, at 2:20 PM, Alan Hightower <alan at alanlee.org> wrote:
>
>> I don't think you are going to find enough PDP8 owners for a
>> kickstarter.
>>
>> Sounds like an interesting project. However if you include 1 SMT
>> chip, you could do it with FRAM or MRAM - which are in a way a modern
>> day equivalent of core memory. A nice link to the past.
>>
>> Most people I know who restore old machines, including me, have an
>> irrational disdain for batteries.
> Irrational? Given the damage I've seen batteries do when they leak,
> I'd
> hardly call it irrational.
>
>
> - Dave
>
>
5 Mar
5 Mar
6:28 a.m.
New subject: Interest in a PDP-8/e/a/f/m 32K RAM kit?
Many thanks to all for the enthusiastic responses to the feeler regarding Vince's RAM
kit!
Regarding comments on the design, Design Requirement #3 precludes SMT parts, so the FRAM
option is out. Actually, given the factors of its limited cycles and 10-year retention,
it's not clear whether that would be a better solution. Lithium coin cells are far
less prone to leaking than the cylindrical dry cells that are notorious for that. No
doubt, they could leak if abused but they are routinely designed for ten-year life span.
They have had enormous usage in PCs and I've never seen a problem with them. Lithium
coin cells don't worry me as much as some other components do.
Is there any evidence that slewrate is a problem for the Omnibus? I've been using the
RAM Board for about a year and a half now and have been all over the bus with wideband
scopes. No problems seen. The RAM has been absolutely solid and the computer has been
glitch-free. This topic is further addressed in the article.
Regarding technicalities of meeting Omnibus threshold specs, I didn't claim in the
article that it does so, only that I judged it to be close enough. Moreover, it's not
completely clear what the Omnibus specs are, as is covered in the article. The fact is,
the Omnibus swings good TTL levels and the chips used in this design have guaranteed TTL
input levels of 2.2V and 0.8V. That's about all I would expect from a nominal
threshold of 1.5V. So whether the target threshold is exactly that or not, I'm
satisfied that its a decent receiver for the Omnibus.
The idea that 1.6V on the bus is an acceptable logic-1, isn't right. If a heavily
loaded bus has (50) DEC5380 receivers, the expected current is about 4.5mA. That will have
negligible effect on DEC pullups #2, #3 and #4 [see article], since they source far more
current and are clamped at 3.5V. DEC pullup #1 will drop from 3.8V to 3.3V. So an Omnibus
computer with 1.6V high levels simply needs repair.
Thank you all for the stimulating discussion.
Steve Lafferty
2:50 p.m.
New subject: Interest in a PDP-8/e/a/f/m 32K RAM kit?
FWIW if I did the calculation correctly, the FRAM I mentioned should be
good for about 30 years at a continuous 1 microsecond access rate before
you exceed it's cycle limit. 10**14 is a pretty big number. However,
the datasheet for the FM18L08 (32kx8) lists retention at 45 years with
no cycle limit.
The FM18L08 devices are available in DIP 28 pin packages from a number
of eBay sellers (old parts, I presume since the current data sheets only
describe surface mount parts.) Costs are about $10 to $15 and there
appears to be a bunch available. I've purchased a few from these
dealers and have had no problems.
-Gary
On 03/05/2014 06:28 AM, Steve Lafferty wrote:
Many thanks to all for the enthusiastic responses to
the feeler regarding Vince's RAM kit!
Regarding comments on the design, Design Requirement #3 precludes SMT parts, so the FRAM
option is out. Actually, given the factors of its limited cycles and 10-year retention,
it's not clear whether that would be a better solution. Lithium coin cells are far
less prone to leaking than the cylindrical dry cells that are notorious for that. No
doubt, they could leak if abused but they are routinely designed for ten-year life span.
They have had enormous usage in PCs and I've never seen a problem with them. Lithium
coin cells don't worry me as much as some other components do.
Is there any evidence that slewrate is a problem for the Omnibus? I've been using the
RAM Board for about a year and a half now and have been all over the bus with wideband
scopes. No problems seen. The RAM has been absolutely solid and the computer has been
glitch-free. This topic is further addressed in the article.
Regarding technicalities of meeting Omnibus threshold specs, I didn't claim in the
article that it does so, only that I judged it to be close enough. Moreover, it's not
completely clear what the Omnibus specs are, as is covered in the article. The fact is,
the Omnibus swings good TTL levels and the chips used in this design have guaranteed TTL
input levels of 2.2V and 0.8V. That's about all I would expect from a nominal
threshold of 1.5V. So whether the target threshold is exactly that or not, I'm
satisfied that its a decent receiver for the Omnibus.
The idea that 1.6V on the bus is an acceptable logic-1, isn't right. If a heavily
loaded bus has (50) DEC5380 receivers, the expected current is about 4.5mA. That will have
negligible effect on DEC pullups #2, #3 and #4 [see article], since they source far more
current and are clamped at 3.5V. DEC pullup #1 will drop from 3.8V to 3.3V. So an Omnibus
computer with 1.6V high levels simply needs repair.
Thank you all for the stimulating discussion.
Steve Lafferty
6:59 p.m.
New subject: Interest in a PDP-8/e/a/f/m 32K RAM kit?
I think the FRAM issue is a moot point. The Ramtron FM18L08 has the same
pinout and package as the SRAM shown in the schematic.
I think you could just leave out the battery, jumper the power diode,
and install FRAM.
Rob.
On 3/5/2014 3:50 PM, Gary Oliver wrote:
FWIW if I did the calculation correctly, the FRAM I
mentioned should be
good for about 30 years at a continuous 1 microsecond access rate before
you exceed it's cycle limit. 10**14 is a pretty big number. However,
the datasheet for the FM18L08 (32kx8) lists retention at 45 years with
no cycle limit.
The FM18L08 devices are available in DIP 28 pin packages from a number
of eBay sellers (old parts, I presume since the current data sheets only
describe surface mount parts.) Costs are about $10 to $15 and there
appears to be a bunch available. I've purchased a few from these
dealers and have had no problems.
-Gary
On 03/05/2014 06:28 AM, Steve Lafferty wrote:
> Many thanks to all for the enthusiastic responses to the feeler
regarding
Vince's RAM kit!
>
> Regarding comments on the design, Design Requirement #3 precludes
> SMT
parts, so the FRAM option is out. Actually, given the factors of its
limited cycles and 10-year retention, it's not clear whether that would
be a better solution. Lithium coin cells are far less prone to leaking
than the cylindrical dry cells that are notorious for that. No doubt,
they could leak if abused but they are routinely designed for ten-year
life span. They have had enormous usage in PCs and I've never seen a
problem with them. Lithium coin cells don't worry me as much as some
other components do.
>
> Is there any evidence that slewrate is a problem for the Omnibus?
I've been
using the RAM Board for about a year and a half now and have
been all over the bus with wideband scopes. No problems seen. The RAM
has been absolutely solid and the computer has been glitch-free. This
topic is further addressed in the article.
>
> Regarding technicalities of meeting Omnibus threshold specs, I
> didn't
claim in the article that it does so, only that I judged it to be
close
enough. Moreover, it's not completely clear what the Omnibus specs are,
as is covered in the article. The fact is, the Omnibus swings good TTL
levels and the chips used in this design have guaranteed TTL input
levels of 2.2V and 0.8V. That's about all I would expect from a nominal
threshold of 1.5V. So whether the target threshold is exactly that or
not, I'm satisfied that its a decent receiver for the Omnibus.
>
> The idea that 1.6V on the bus is an acceptable logic-1, isn't
> right.
If a heavily loaded bus has (50) DEC5380 receivers, the expected
current
is about 4.5mA. That will have negligible effect on DEC pullups #2, #3
and #4 [see article], since they source far more current and are clamped
at 3.5V. DEC pullup #1 will drop from 3.8V to 3.3V. So an Omnibus
computer with 1.6V high levels simply needs repair.
Thank you all for the stimulating discussion.
Steve Lafferty
7:13 p.m.
New subject: Interest in a PDP-8/e/a/f/m 32K RAM kit?
On Wed, Mar 5, 2014 at 8:59 PM, Rob Doyle <radioengr at gmail.com> wrote:
I think the FRAM issue is a moot point. The Ramtron
FM18L08 has the same
pinout and package as the SRAM shown in the schematic.
I think you could just leave out the battery, jumper the power diode,
and install FRAM.
Same pinout, yes, but it's only rated to 3.6V according to the datasheet
here:
http://pdf.datasheetcatalog.com/datasheet/RamtronInternational/mXuwyyw.pdf
I don't think that will be a drop-in replacement in Steve's original
circuit.
Kyle
10:19 p.m.
New subject: Interest in a PDP-8/e/a/f/m 32K RAM kit?
On 3/5/2014 8:13 PM, Kyle Owen wrote:
On Wed, Mar 5, 2014 at 8:59 PM, Rob Doyle
<radioengr at gmail.com> wrote:
Sorry. It's the FM1808 that is 5V. Cypress bought Ramtron...
http://www.cypress.com/?docID=42544
Rob.
I think the FRAM issue is a moot point. The
Ramtron FM18L08 has the same
pinout and package as the SRAM shown in the schematic.
I think you could just leave out the battery, jumper the power diode,
and install FRAM.
Same pinout, yes, but it's only rated to 3.6V according to the datasheet
here:
http://pdf.datasheetcatalog.com/datasheet/RamtronInternational/mXuwyyw.pdf
I don't think that will be a drop-in replacement in Steve's original
circuit.
Kyle
6 Mar
6 Mar
10:04 a.m.
New subject: Interest in a PDP-8/e/a/f/m 32K RAM kit?
To make things more confusing and frustrating, the 5v parts have a
"cycle limit" of 10**12 while the 3.6v parts are "unlimited"
according
to the datasheets I have. The 5v parts appear to be about 3 years back
and are not in the current 'catalog'.
For my experience, the 5v parts have been truly "drop in replacement"
for the volatile parts. I have no experience with drop-in on the 3.6
volt part, since I crafted the circuit specifically for that device.
Last I'll say on the topic. If I still had an 8E, I'd likely be more
involved ;-)
-Gary
On 03/05/2014 10:19 PM, Rob Doyle wrote:
On 3/5/2014 8:13 PM, Kyle Owen wrote:
On Wed, Mar 5, 2014 at 8:59 PM, Rob Doyle
<radioengr at gmail.com> wrote:
Sorry. It's the FM1808 that is 5V. Cypress bought Ramtron...
http://www.cypress.com/?docID=42544
Rob.
I think the FRAM issue is a moot point. The
Ramtron FM18L08 has the
same
pinout and package as the SRAM shown in the schematic.
I think you could just leave out the battery, jumper the power diode,
and install FRAM.
Same pinout, yes, but it's only rated to 3.6V according to the datasheet
here:
http://pdf.datasheetcatalog.com/datasheet/RamtronInternational/mXuwyyw.pdf
I don't think that will be a drop-in replacement in Steve's original
circuit.
Kyle
10:18 a.m.
New subject: Interest in a PDP-8/e/a/f/m 32K RAM kit?
On the question of whether F-RAM would be a suitable replacement for the CY62256
CMOS/battery solution: There was a longevity calculation posted that appears to have been
in error. The 1e14 cycle limit was said to be good for 30-years continuous access at a 1us
cycle rate. I calculate about 3-years. Perhaps 2us would be a more appropriate cycle rate
for the PDP-8 and that would make it 6-years.
That might be considered sufficient and I have added a section discussing the availability
of various F-RAMs to the article: http://tronola.com/html/ram_for_pdp-8e.html#F-RAM
Unfortunately, the suitable through-hole devices are now obsolete but are still available
from Hong Kong/China.
I would appreciate any comments you might have.
By the way, in response to comments here about how the Omnibus interface compatibility
issue was handled, I have rewritten parts of that section. I hope it meets critical
approval now. I would like to say thanks for holding my feet to fire. I think it has
improved the article.
Steve Lafferty
2:29 p.m.
New subject: Interest in a PDP-8/e/a/f/m 32K RAM kit?
On Thu, Mar 6, 2014 at 11:18 AM, Steve Lafferty <steve at tronola.com> wrote:
On the question of whether F-RAM would be a suitable
replacement for the
CY62256 CMOS/battery solution: There was a longevity calculation posted
that appears to have been in error. The 1e14 cycle limit was said to be
good for 30-years continuous access at a 1us cycle rate. I calculate about
3-years. Perhaps 2us would be a more appropriate cycle rate for the PDP-8
and that would make it 6-years.
The 5V FRAM devices have a lower rated endurance than the 3.3V devices,
10^12 cycles for the FM1808 (5V) vs. 10^16 cycles for the FM18L08 (3.3V) or
10^14 cycles for the FM18W08 (wide voltage range). For the 5V that's only
23 days at continuous 2us cycles.
Each read or write cycle counts against the endurance of a full "row" of
the FRAM, which is four bytes in the case of the FM1808 (5V), unspecified
for the FM18L08, and eight bytes in the case of the FM18W08.
Note also that FRAM parts latch the address on the falling edge of /CE,
unlike static RAMs.
2:43 p.m.
New subject: Interest in a PDP-8/e/a/f/m 32K RAM kit?
If folks are concerned about battery leakage, then why don't you just place the
battery inside a plastic box mounted remotely inside the cabinet (or even outside of it)
and wire it over to the circuit card? Then you can use SRAM with no write cycle limit,
without any risk of the battery leaking onto the circuit card or into the backplane.
--
Mark J. Blair, NF6X <nf6x at nf6x.net>
http://www.nf6x.net/
7 Mar
7 Mar
5:41 a.m.
New subject: Interest in a PDP-8/e/a/f/m 32K RAM kit?
On 03/06/2014 05:43 PM, Mark J. Blair wrote:
If folks are concerned about battery leakage, then why
don't you just place the battery inside a plastic box mounted remotely inside the
cabinet (or even outside of it) and wire it over to the circuit card? Then you can use
SRAM with no write cycle limit, without any risk of the battery leaking onto the circuit
card or into the backplane.
Its a lithium coin cell. Has anyone ever seen one leak, ever?
I will say here and now that I've replaced at least 30 of them in
various PCs, Mobile radios,
Amateur radios, and can say that while in most cases they go dead far
sooner than expected
(4-6years vs expected 7-10) I have never seen one leak.
If battery leakage were a worry there are two solutions, remove it when
not in use as
likely you have it sitting far too long anyway. Or attach two leads and
remote it.
Its not a problem. IF you have a PDP-8 keeping the power supply happy and
various other issues like having enough core to do anything is more of
an issue.
Me I did this twice, once using 2102s without backup as a car battery
would be needed for
more than a few hours. The second time I used 61256 cmos with a
battery. Removed all
my core to safe dry storage case and run on the ram. That saves 6 slot
spaces
(core is triple thick) and I can run easily with two slower quieter fans
rather than the
full load of noisy (its an 8f).
Allison
6 Mar
6 Mar
10:42 p.m.
New subject: Interest in a PDP-8/e/a/f/m 32K RAM kit?
On 3/6/2014 11:18 AM, Steve Lafferty wrote:
On the question of whether F-RAM would be a suitable
replacement for
the CY62256 CMOS/battery solution: There was a longevity calculation
posted that appears to have been in error. The 1e14 cycle limit was
said to be good for 30-years continuous access at a 1us cycle rate.
I calculate about 3-years. Perhaps 2us would be a more appropriate
cycle rate for the PDP-8 and that would make it 6-years.
That might be considered sufficient and I have added a section
discussing the availability of various F-RAMs to the article:
http://tronola.com/html/ram_for_pdp-8e.html#F-RAM
Unfortunately, the suitable through-hole devices are now obsolete
but are still available from Hong Kong/China.
I would appreciate any comments you might have.
By the way, in response to comments here about how the Omnibus
interface compatibility issue was handled, I have rewritten parts of
that section. I hope it meets critical approval now. I would like to
say thanks for holding my feet to fire. I think it has improved the
article.
Steve Lafferty
The other alternative is the Simtek STK16C88 quantum trap SRAM (now also
Cypress). This device has unlimited reads and writes but has 1e6
power-down cycle endurance. 100 year data retention.
This device is also pin compatible with the SRAMS and is 5v. It may be
a better choice than the FRAM.
http://www.cypress.com/?docID=34462
They are also available on EBAY for about $10.00.
Rob.
7 Mar
7 Mar
5:50 a.m.
New subject: Interest in a PDP-8/e/a/f/m 32K RAM kit?
The other alternative is the Simtek STK16C88 quantum
trap SRAM (now also
Cypress). This device has unlimited reads and writes but has 1e6
power-down cycle endurance. 100 year data retention.
This device is also pin compatible with the SRAMS and is 5v. It may be
a better choice than the FRAM.
http://www.cypress.com/?docID=34462
They are also available on EBAY for about $10.00.
Rob.
--- Thank you for the suggestion but that option was discussed in the article. Quoting:
I also looked at the 32Kx8 Cypress STK16C88 flash-type NVRAM, which costs $13 at Mouser.
It?s being phased out and the replacement is neither 5V nor through-hole. Nevertheless,
since the STK16C88 is still in production, I considered using it. After all, the prospect
of getting rid of the battery altogether is very attractive. Unfortunately, I could not
find any ESD specs in either the commercial or military versions of the datasheet. This,
along with the fact that it is being discontinued and has some spooky
voltage/timing/retention requirements, discouraged me from using the part. Not to mention
that the two required would cost $26, versus the $4 we would have paid for a pair of
CY7199s.
Steve Lafferty
19 Mar
19 Mar
4:36 a.m.
New subject: Interest in a PDP-8/e/a/f/m 32K RAM kit?
Am 06.03.2014 19:18, schrieb Steve Lafferty:
calculate about 3-years. Perhaps 2us would be a more
appropriate
cycle rate for the PDP-8 and that would make it 6-years.
I would say 600ns for
safety. But 1200ns 10 1400ns should be the usual
case. Not 2000ns
4:32 a.m.
New subject: Interest in a PDP-8/e/a/f/m 32K RAM kit?
Am 05.03.2014 15:28, schrieb Steve Lafferty:
Regarding comments on the design, Design Requirement
#3 precludes SMT
parts, so the FRAM option is out.
Why? SMT saves board space and is not that hard
to solder as some people
think!
Actually, given the factors of its
limited cycles and 10-year retention,
Fujitsu MB85R1001ANC-GE1, 1MBit FRAM:
A pdp8 has at best 1.2us cycle time. Each cycle has to be modeled as two
accesses. Makes an access cycle time of 600ns. Frequency is 1,6666...
MHz. Oops... Only 6000 seconds of operation!
10^10 sounded so much :-(
4 Mar
4 Mar
11:54 a.m.
New subject: Interest in a PDP-8/e/a/f/m 32K RAM kit?
On Tue, Mar 4, 2014 at 8:50 AM, Steve Lafferty <steve at tronola.com> wrote:
I've heard from a fellow, Vince S., who has
created a PCB layout for the
PDP-8/e 32K RAM design presented in my article on Tronola.com. He's
considering a Kickstarter launch to offer a $95 kit of parts and PCB but is
wondering if there is sufficient interest. If so, please let it be known.
I'm totally interested in a $95 OMNIBUS memory board, especially a
project by Vince S., but I won't use Amazon payments. :-/
-ethan
18 Mar
18 Mar
8:10 a.m.
New subject: 32K RAM kits (was Re: Interest in a PDP-8/e/a/f/m 32K RAM kit?)
From: Ethan Dicks: Tuesday, March 04, 2014 12:54 PM
I'm totally interested in a $95 OMNIBUS memory
board, especially a
project by Vince S., but I won't use Amazon payments. :-/
I went to the trouble of setting up a Kickstarter, but when I went to
submit it for approval, it said I needed to set up Amazon payments,
and when I went to do that, it wanted my SSN and a whole bunch of
crap related to taxes, money laundering, etc. I won't be giving my
SSN to Amazon, so I won't be doing a Kickstarter.
Fortunately, there's a lot of interest in the boards, and I'm already
set up to take PayPal at this email. So, I've set up some PDF files at
http://svn.so-much-stuff.com/svn/trunk/Eagle/projects/32k-Omnibus/
Lafferty1sch.pdf is the schematic. Page 1 has the connections to the
Omnibus, page 2 has the RAM, page 3 has a timing diagram, and
pages 4 and 5 are the prototyping areas.
Lafferty1top.pdf and Lafferty1bot.pdf show views of the top and
bottom of the proposed boards, respectively. (The color schemes
in the PDF files are silly -- the actual boards should have gold edge
connectors, green solder mask, and white silkscreen.)
Anyway, my plan is to collect money through the end of the month,
then order the boards and the parts. I should get the boards
toward the end of next month, and be able to start sending stuff
out shortly after that. (I have sufficient funds that if PayPal gets its
knickers in a twist, I can pre-order the stuff and show the invoices
to get PayPal unstuck.)
Vince
8 p.m.
New subject: 32K RAM kits (was Re: Interest in a PDP-8/e/a/f/m 32K RAM kit?)
So...is your Paypal account the same as your email address?
Thanks,
Marc
On Tue, Mar 18, 2014 at 8:10 AM, Vincent Slyngstad <vrs at msn.com> wrote:
From: Ethan Dicks: Tuesday, March 04, 2014 12:54 PM
I'm totally interested in a $95 OMNIBUS
memory board, especially a
project by Vince S., but I won't use Amazon payments. :-/
I went to the trouble of setting up a Kickstarter, but when I went to
submit it for approval, it said I needed to set up Amazon payments, and
when I went to do that, it wanted my SSN and a whole bunch of crap related
to taxes, money laundering, etc. I won't be giving my SSN to Amazon, so I
won't be doing a Kickstarter.
Fortunately, there's a lot of interest in the boards, and I'm already set
up to take PayPal at this email. So, I've set up some PDF files at
http://svn.so-much-stuff.com/svn/trunk/Eagle/projects/32k-Omnibus/
Lafferty1sch.pdf is the schematic. Page 1 has the connections to the
Omnibus, page 2 has the RAM, page 3 has a timing diagram, and pages 4 and 5
are the prototyping areas.
Lafferty1top.pdf and Lafferty1bot.pdf show views of the top and bottom of
the proposed boards, respectively. (The color schemes
in the PDF files are silly -- the actual boards should have gold edge
connectors, green solder mask, and white silkscreen.)
Anyway, my plan is to collect money through the end of the month, then
order the boards and the parts. I should get the boards toward the end of
next month, and be able to start sending stuff out shortly after that. (I
have sufficient funds that if PayPal gets its knickers in a twist, I can
pre-order the stuff and show the invoices to get PayPal unstuck.)
Vince 
7 Mar
7 Mar
1:20 a.m.
New subject: Interest in a PDP-8/e/a/f/m 32K RAM kit?
Put me in the list of future purchasers, please.
2014-03-04 14:50 GMT+01:00 Steve Lafferty <steve at tronola.com>:
I've heard from a fellow, Vince S., who has created a PCB layout for the
PDP-8/e 32K RAM design presented in my article on Tronola.com. He's
considering a Kickstarter launch to offer a $95 kit of parts and PCB but is
wondering if there is sufficient interest. If so, please let it be known.
Features include:
- Battery backup with estimated 10-year data retention.
- Inhibit on power fail to prevent memory corruption.
- Most of the full size Omnibus board offers a prototyping layout.
- Cooperation with any mix of core memory boards. Switches disable 4K
fields.
- Compatibility with Omnibus specs.
- All through-hole parts, for easy assembly.
I'm not associated with the project, other than having published the
article and encouraging his efforts. Vince has published lots of
PDP-8-related and other layouts on his website and will make this one
available as well:
http://so-much-stuff.com/pdp8/cad/cad.php
The article is at:
http://www.tronola.com/html/ram_for_pdp-8e.html
Thanks,
Steve Lafferty
http://www.tronola.com/
--
Gracias | Regards
Saludos - Greetings - Freundliche Gr??e - Salutations
--
*Sergio Pedraja*
twitter: @sergio_pedraja
skype: Sergio Pedraja
http://plus.google.com/u/0/101292256663392735405
http://www.linkedin.com/in/sergiopedraja
http://www.quora.com/Sergio-Pedraja
http://spedraja.wordpress.com
https://www.xing.com/profile/Sergio_Pedraja <http://spedraja.wordpress.com/>
-----
No crea todo lo que ve, ni crea que est? vi?ndolo todo
4 Mar
4 Mar
7:52 a.m.
Hi, I did post my solution to the group but it does not seem to have got
there, I have noterd that sometimes my mails dont get through so here it is
again.
Dave H
Hi
The problem is with the gating of the IOT ( 6RK4L) before its applied to
the 74161 as CLK (2),
on drawing M7105-0-1 REV B , MAJOR REGISTERS (D06)
this is done by E47 , E41 and E28, the problen is that they are taking the
6RK4, the IOT, and trying to truncate it with a modified TP3, and if this
TP3 timeing is a little off then you get a spike at the trailing edge.
Different revisions of this board seem to have different ways of generating
this TP3, and mine has a wire patch in place .
Have a look at E18(2), the CLK while doing an DLCA scope loop and see what
you get.
My fix was as follows, "piggy back" a 7474 onto E41, just the gnd and
power , 7 and 14, bend the rest of the pins away.
On the 7474 connect both the D(2) and the CLR (1) to E41(10), then connect
the CK(3) to E41(9).
Cut the track from E41(8) to E28(12). Connect fron the 7474 op(5) to
E28(12).
The result of this is to latch the TP3 pulse so it cant go low before the
end of the IOT.
This shoul be it, I had to decifer my notes and look at the board so I hope
iv got it right!
Lets hope this is your problem, the RK8E sure is fun to work with , I had to
change several IC's to get my set working at all.
Cheers
Dave
----- Original Message -----
From: "Al Kossow" <aek at bitsavers.org>
To: <cctalk at classiccmp.org>
Sent: Monday, March 03, 2014 5:51 PM
Subject: Re: PDP 8/e, RK8E Weirdness - Scope Traces - Update
On 3/3/14 8:37 AM, Rick Bensene wrote:
If anyone wants the procedure for the
modification, I can Email it to
you. Just drop me an Email and I'll send it along.
Please just post it to the list so that is it archived.
-----
No virus found in this message.
Checked by AVG - www.avg.com
Version: 2012.0.2247 / Virus Database: 3705/6642 - Release Date: 03/02/14
3927
days inactive
3948
days old
75 comments
27 participants
participants (27)
-
aek＠bitsavers.org -
ajp166＠verizon.net -
alan＠alanlee.org -
ard＠p850ug1.demon.co.uk -
cramcram＠gmail.com -
david＠hheng.plus.com -
dfnr2＠yahoo.com -
ethan.dicks＠gmail.com -
fraveydank＠gmail.com -
ggs＠shiresoft.com -
go＠aerodesic.com -
go＠ao-cs.com -
hachti＠hachti.de -
jdr_use＠bluewin.ch -
kylevowen＠gmail.com -
nf6x＠nf6x.net -
pontus＠Update.UU.SE -
radioengr＠gmail.com -
rickb＠bensene.com -
ryan＠hack.net -
spacewar＠gmail.com -
spedraja＠gmail.com -
steve＠tronola.com -
tsg＠bonedaddy.net -
vrs＠msn.com -
wilson＠dbit.com -
wmaddox＠pacbell.net