27 Sep
2021
27 Sep
'21
8:23 a.m.
While restoring and repairing a Data General Nova 2/10 I found a bad
bipolar PROM on the CPU board. The PROM has open-collector outputs and is
organized as 32 words by 8 bits. It appears that one of the open-collector
driver transistors is faulty (but it could also be that a fuse has
"healed").
The part is an Intersil IM5600CP, but these were also made by others, for
example Signetics and Philips made the 82S23 and TI and NTE made the faster
SN74S188N. Some vendors still sell these parts and there are even a few on
Ebay.
How do I program these PROMs? I found one somewhat obscure description of
the algorithm in the NTE datasheet, but I suspect that each manufacturer
had (somewhat) different algorithms.
Is there an affordable commercial programmer out there which can program
these PROMs?
Is there a simple design out there which I could breadboard for a one-off
programming job (maybe using an Arduino to control the programming
sequence)?
Thanks and best regards
Tom Hunter
27 Sep
27 Sep
8:28 a.m.
I can do them with the Data I/O 29B and UniPak 2B. Some of the other UniPaks can handle
them too. They are one of those devices that most of the newer commercial programmers
can't do. I wouldn't call the Data I/O really affordable, though.
If you want to send us a blank we can program it for you, or we probably have some blanks
for that part # on hand.
Watch out buying blanks on eBay, they're often pulls and are actually already
programmed. Also the yield was not 100% even back in the day, apparently, so buy several
to ensure you do get a good one.
Thanks,
Jonathan
??????? Original Message ???????
On Monday, September 27th, 2021 at 11:23, Tom Hunter via cctalk <cctalk at
classiccmp.org> wrote:
While restoring and repairing a Data General Nova 2/10
I found a bad
bipolar PROM on the CPU board. The PROM has open-collector outputs and is
organized as 32 words by 8 bits. It appears that one of the open-collector
driver transistors is faulty (but it could also be that a fuse has
"healed").
The part is an Intersil IM5600CP, but these were also made by others, for
example Signetics and Philips made the 82S23 and TI and NTE made the faster
SN74S188N. Some vendors still sell these parts and there are even a few on
Ebay.
How do I program these PROMs? I found one somewhat obscure description of
the algorithm in the NTE datasheet, but I suspect that each manufacturer
had (somewhat) different algorithms.
Is there an affordable commercial programmer out there which can program
these PROMs?
Is there a simple design out there which I could breadboard for a one-off
programming job (maybe using an Arduino to control the programming
sequence)?
Thanks and best regards
Tom Hunter
8:48 a.m.
Jon wrote...
I can do them with the Data I/O 29B and UniPak 2B. Some of the other UniPaks can handle
them too. They are one of those devices that most of the newer commercial programmers
can't do. I wouldn't call the Data I/O really affordable, though.
------
Same, I use a data i/o 29B for these. Jon is right, the 29B seems to have held significant
value $-wise still on ebay. But you can find a reasonable deal once in a while.
And....
Watch out buying blanks on eBay, they're often pulls and are actually already
programmed. Also the yield was not 100% even back in the day, apparently, so buy several
to ensure you do get a good one.
------
I gave up trying to find any quantities of the 3 or 4 different bipolar proms used for the
HP 2100/21MX line on ebay. I had significant luck contacting arcade game repair houses,
and got a reasonable quantity of most all the flavors I needed. I checked back a couple
years ago and all the usual arcade places no longer had them ?
I do believe I saw a writeup on the web somewhere where someone took a different chip and
programmed it to act like a (now) rare bipolar prom. I don't recall how successful it
was - I'd imagine the speed of the bipolars would be an issue, plus ISTR height was an
issue too (new chip was on a carrier of sorts, creating headroom issues on some systems).
J
9:16 a.m.
I recently picked up an EPROM+ programming unit (
https://www.arlabs.com/eprom_plus.html) from eBay used in order to program
a couple of PROMs. Unfortunately, I haven't actually tried it yet but the
82S23 and others are supported. The owner of the company seems to be
extremely responsive to any and all questions and has a lot of
documentation on the web site.
On Mon, Sep 27, 2021 at 11:48 AM jwest--- via cctalk <cctalk at classiccmp.org>
wrote:
Jon wrote...
I can do them with the Data I/O 29B and UniPak 2B. Some of the other
UniPaks can handle them too. They are one of those devices that most of the
newer commercial programmers can't do. I wouldn't call the Data I/O really
affordable, though.
------
Same, I use a data i/o 29B for these. Jon is right, the 29B seems to have
held significant value $-wise still on ebay. But you can find a reasonable
deal once in a while.
And....
Watch out buying blanks on eBay, they're often pulls and are actually
already programmed. Also the yield was not 100% even back in the day,
apparently, so buy several to ensure you do get a good one.
------
I gave up trying to find any quantities of the 3 or 4 different bipolar
proms used for the HP 2100/21MX line on ebay. I had significant luck
contacting arcade game repair houses, and got a reasonable quantity of most
all the flavors I needed. I checked back a couple years ago and all the
usual arcade places no longer had them ?
I do believe I saw a writeup on the web somewhere where someone took a
different chip and programmed it to act like a (now) rare bipolar prom. I
don't recall how successful it was - I'd imagine the speed of the bipolars
would be an issue, plus ISTR height was an issue too (new chip was on a
carrier of sorts, creating headroom issues on some systems).
J
10:05 a.m.
Most of these older fused PROMs are of comparable speed in newer EPROMs or E2PROMs. Open
collector is a hassle but not too much. One can create the needed circuits using surface
mount parts for size reduction.
A hassle but not out of the question. You usually have to go to a larger ROM size so it
can be used for additional storage with a switch or jumper select.
Dwight
________________________________
From: cctalk <cctalk-bounces at classiccmp.org> on behalf of Brian Marstella via
cctalk <cctalk at classiccmp.org>
Sent: Monday, September 27, 2021 9:16 AM
To: Jay West <jwest at classiccmp.org>; General Discussion: On-Topic and Off-Topic
Posts <cctalk at classiccmp.org>
Subject: Re: Programming Bipolar PROMs
I recently picked up an EPROM+ programming unit (
https://www.arlabs.com/eprom_plus.html) from eBay used in order to program
a couple of PROMs. Unfortunately, I haven't actually tried it yet but the
82S23 and others are supported. The owner of the company seems to be
extremely responsive to any and all questions and has a lot of
documentation on the web site.
On Mon, Sep 27, 2021 at 11:48 AM jwest--- via cctalk <cctalk at classiccmp.org>
wrote:
Jon wrote...
I can do them with the Data I/O 29B and UniPak 2B. Some of the other
UniPaks can handle them too. They are one of those devices that most of the
newer commercial programmers can't do. I wouldn't call the Data I/O really
affordable, though.
------
Same, I use a data i/o 29B for these. Jon is right, the 29B seems to have
held significant value $-wise still on ebay. But you can find a reasonable
deal once in a while.
And....
Watch out buying blanks on eBay, they're often pulls and are actually
already programmed. Also the yield was not 100% even back in the day,
apparently, so buy several to ensure you do get a good one.
------
I gave up trying to find any quantities of the 3 or 4 different bipolar
proms used for the HP 2100/21MX line on ebay. I had significant luck
contacting arcade game repair houses, and got a reasonable quantity of most
all the flavors I needed. I checked back a couple years ago and all the
usual arcade places no longer had them ?
I do believe I saw a writeup on the web somewhere where someone took a
different chip and programmed it to act like a (now) rare bipolar prom. I
don't recall how successful it was - I'd imagine the speed of the bipolars
would be an issue, plus ISTR height was an issue too (new chip was on a
carrier of sorts, creating headroom issues on some systems).
J
10:33 a.m.
On 9/27/21 10:05 AM, dwight via cctalk wrote:
Most of these older fused PROMs are of comparable
speed in newer
EPROMs or E2PROMs. Open collector is a hassle but not too much. One
can create the needed circuits using surface mount parts for size
reduction. A hassle but not out of the question. You usually have to
go to a larger ROM size so it can be used for additional storage with
a switch or jumper select.
In some cases, if you're handy with logic tables,
it's sometimes
possible to program a GAL to act as a bipolar PROM substitute.
--Chuck
10:07 a.m.
On Mon, 27 Sept 2021 at 17:23, Tom Hunter via cctalk
<cctalk at classiccmp.org> wrote:
While restoring and repairing a Data General Nova 2/10 I found a bad
bipolar PROM on the CPU board.
Maybe it was just feeling depressed this week. Try it again next week
and you might find it works... maybe even really fast?
(Sorry if my attempt at humour upsets anyone.)
--
Liam Proven ? Profile: https://about.me/liamproven
Email: lproven at cix.co.uk ? gMail/gTalk/gHangouts: lproven at gmail.com
Twitter/Facebook/LinkedIn/Flickr: lproven ? Skype: liamproven
UK: +44 7939-087884 ? ?R (+ WhatsApp/Telegram/Signal): +420 702 829 053
10:52 a.m.
On 2021-Sep-27, at 8:23 AM, Tom Hunter via cctalk wrote:
While restoring and repairing a Data General Nova 2/10
I found a bad
bipolar PROM on the CPU board. The PROM has open-collector outputs and is
organized as 32 words by 8 bits. It appears that one of the open-collector
driver transistors is faulty (but it could also be that a fuse has
"healed").
The part is an Intersil IM5600CP, but these were also made by others, for
example Signetics and Philips made the 82S23 and TI and NTE made the faster
SN74S188N. Some vendors still sell these parts and there are even a few on
Ebay.
How do I program these PROMs? I found one somewhat obscure description of
the algorithm in the NTE datasheet, but I suspect that each manufacturer
had (somewhat) different algorithms.
Is there an affordable commercial programmer out there which can program
these PROMs?
Is there a simple design out there which I could breadboard for a one-off
programming job (maybe using an Arduino to control the programming
sequence)?
The hardware & software for a one-off breadboarded programmer is, or should be, easy
enough with a ~ few-hours/1-day effort if you wanted to go that route.
I did this 8 years ago for a PROM in a HP9830 - a 256*4 Intel 3601 PROM, for which I
burned, and replaced with, a TI 74S387.
(In the end in turned out it was an unnecessary effort, one data-line from the PROM was
showing bad but the failure wasn't the PROM output, it was the input of the IC the
PROM was driving.)
Programmer hardware was 3 TTL ICs, 2 zeners, 6 transistors (more specifically
2+numDataLines) and a few resistors on a solderless breadboard, and a bench power supply
for the programming voltage. Software was a python program running on a RPi. All
program-pulse timing was done in user-land on the RPi.
I can send along the schematic and program if you wish, there will be some adaptation
required of course. Details of the 74S188 programming/burn algorithm is in the 1975 TI
Memory Databook for Design Engineers (available at bitsavers) (I expect the 188 is the
most likely type you'll find NOS today).
11:20 a.m.
On 2021-Sep-27, at 10:52 AM, Brent Hilpert via cctalk wrote:
On 2021-Sep-27, at 8:23 AM, Tom Hunter via cctalk
wrote:
> While restoring and repairing a Data General Nova 2/10 I found a bad
> bipolar PROM on the CPU board. The PROM has open-collector outputs and is
> organized as 32 words by 8 bits. It appears that one of the open-collector
> driver transistors is faulty (but it could also be that a fuse has
> "healed").
>
> The part is an Intersil IM5600CP, but these were also made by others, for
> example Signetics and Philips made the 82S23 and TI and NTE made the faster
> SN74S188N. Some vendors still sell these parts and there are even a few on
> Ebay.
>
> How do I program these PROMs? I found one somewhat obscure description of
> the algorithm in the NTE datasheet, but I suspect that each manufacturer
> had (somewhat) different algorithms.
>
> Is there an affordable commercial programmer out there which can program
> these PROMs?
>
> Is there a simple design out there which I could breadboard for a one-off
> programming job (maybe using an Arduino to control the programming
> sequence)?
...
Details of the 74S188 programming/burn algorithm is in the 1975 TI Memory Databook for
Design Engineers (available at bitsavers) (I expect the 188 is the most likely type
you'll find NOS today).
The 82S23 programming algorithm is in the 1975 Signetics Bipolar Memories databook (also
on bitsavers).
Looks like it may be a little more complex than that for the 188.
12:05 p.m.
On Mon, Sep 27, 2021 at 11:20 AM Brent Hilpert via cctalk
<cctalk at classiccmp.org> wrote:
The 82S23 programming algorithm is in the 1975 Signetics Bipolar Memories databook (also
on bitsavers).
Looks like it may be a little more complex than that for the 188.
http://www.bitsavers.org/components/signetics/_dataBooks/1977_Signetics_Bip…
Page 91, timing diagram on Page 92
82S23 PROGRAMMING PROCEDURE
1. Terminate all device outputs with a 10K Ohm resistor to Vcc.
2. Select the address to be programmed, and raise Vcc to Vccp = +10 ? 0.5V.
3. After 10us delay, apply Iout = 65 ? 3mA to the output to be
programmed. Program one output at a time.
4. After 10us delay, pulse the /CE input to logic low for 0.3 to 0.5us.
5. After 10us delay, remove Iout from the programmed output.
6. After 10us delay, return Vcc to 0V.
7. To verify programming, after 50us delay, raise Vcc to Vcch = +5.5 ?
0.2V, and apply a logic low level to the /CE input. The programmed
output should remain in the high state. Again, lower Vcc to Vccl =
+4.5 ? 0.2V, and verify that the programmed output remains in the high
state.
8. Raise Vcc to Vccp = +10 ? 0.5V and repeat steps 3 through 7 to
program other bits at the same address.
9. After 10us delay, repeat steps 2 through 8 to program all other
address locations.
http://www.bitsavers.org/components/signetics/_dataBooks/1975_Signetics_Bip…
Page 24, timing diagram on Page 25
Same as above, except different timing in step 4.
4. After 10us delay, pulse the /CE input to logic "0" for 1 to 2 ms.
(of course in case of typos above, refer to the manuals)
10:57 a.m.
On Mon, Sep 27, 2021 at 8:23 AM Tom Hunter via cctalk
<cctalk at classiccmp.org> wrote:
The part is an Intersil IM5600CP, but these were also made by others, for
example Signetics and Philips made the 82S23 and TI and NTE made the faster
SN74S188N. Some vendors still sell these parts and there are even a few on
Ebay.
How do I program these PROMs? I found one somewhat obscure description of
the algorithm in the NTE datasheet, but I suspect that each manufacturer
had (somewhat) different algorithms.
Is there an affordable commercial programmer out there which can program
these PROMs?
How do you program a single bipolar PROM as a one time occurance? Find
someone willing to do it for you for the cost of the postage :)
Are there affordable commercial programmers which can reliably program
bipolar PROMs with device vendor approved programming algorithms? I
have an older Data I/O UniSite and a newer BP Microsystems BP-1610
which between the two of them can cover the majority of those older
devices, and the BP-1610 is currently supported with regular software
updates. I wouldn't call either of those two programmers "affordable",
and you have to be patient to pick one up on eBay.
The older parallel port interface BP-1200 and BP-1400 device
programmers probably have the same bipolar PROM support as the
BP-1610, and can more often be found at somewhat reasonable prices on
eBay, but software updates for new devices were discontinued for those
non-USB interface programmers several years ago.
8:38 p.m.
On 2021-09-27 9:23 a.m., Tom Hunter via cctalk wrote:
While restoring and repairing a Data General Nova 2/10
I found a bad
bipolar PROM on the CPU board. The PROM has open-collector outputs and is
organized as 32 words by 8 bits. It appears that one of the open-collector
driver transistors is faulty (but it could also be that a fuse has
"healed").
The part is an Intersil IM5600CP, but these were also made by others, for
example Signetics and Philips made the 82S23 and TI and NTE made the faster
SN74S188N. Some vendors still sell these parts and there are even a few on
Ebay.
How do I program these PROMs? I found one somewhat obscure description of
the algorithm in the NTE datasheet, but I suspect that each manufacturer
had (somewhat) different algorithms.
Is there an affordable commercial programmer out there which can program
these PROMs?
Is there a simple design out there which I could breadboard for a one-off
programming job (maybe using an Arduino to control the programming
sequence)?
Thanks and best regards
Tom Hunter
http://www.eprompro.com/Index.html
Does PROMS.
Ben.
28 Sep
28 Sep
8:14 a.m.
... and also has many rare ones in stock; reasonable prices, good guy!
On Mon, Sep 27, 2021 at 11:38 PM ben via cctalk <cctalk at classiccmp.org>
wrote:
On 2021-09-27 9:23 a.m., Tom Hunter via cctalk wrote:
While restoring and repairing a Data General Nova
2/10 I found a bad
bipolar PROM on the CPU board. The PROM has open-collector outputs and is
organized as 32 words by 8 bits. It appears that one of the
open-collector
driver transistors is faulty (but it could also
be that a fuse has
"healed").
The part is an Intersil IM5600CP, but these were also made by others, for
example Signetics and Philips made the 82S23 and TI and NTE made the
faster
SN74S188N. Some vendors still sell these parts
and there are even a few
on
Ebay.
How do I program these PROMs? I found one somewhat obscure description of
the algorithm in the NTE datasheet, but I suspect that each manufacturer
had (somewhat) different algorithms.
Is there an affordable commercial programmer out there which can program
these PROMs?
Is there a simple design out there which I could breadboard for a one-off
programming job (maybe using an Arduino to control the programming
sequence)?
Thanks and best regards
Tom Hunter
http://www.eprompro.com/Index.html
Does PROMS.
Ben.
943
days inactive
944
days old
12 comments
11 participants
participants (11)
-
bfranchuk＠jetnet.ab.ca -
bhilpert＠shaw.ca -
brian＠marstella.net -
cclist＠sydex.com -
ccth6600＠gmail.com -
dkelvey＠hotmail.com -
glen.slick＠gmail.com -
jwest＠classiccmp.org -
lists＠glitchwrks.com -
lproven＠gmail.com -
mhs.stein＠gmail.com