16 Dec
2016
16 Dec
'16
1:43 p.m.
Hi folks,
I have an 8085-based viewdata telephone system on the bench that's proving
to be a labour of love in trying to get it running with zero documentation -
there are only 6 known examples that I've come across and all but 2 of them
are in museums, none known to work. If any of you fancy searching it's an
STC Executel 3910 and at least two of the hits you'll get will be my machine
before I bought it.
http://www.binarydinosaurs.co.uk/stcexecutel01.jpg
I've been in contact with all museums who have one, no luck on docs though
the Museum of Computing which is local to me MAY have some (Jason the owner
has 2 3910s himself), I just need to get down there and search for it :)
Kind-of-fortunately the viewdata side of things seems standard - all 74LS
TTL with an MC3242AP running 16x 4116-2 DRAMs, 27128 EEPROMs etc. The
viewdata side of things is powered by a Plessey MR9735-002 teletext
processor supported by a pair of 2112 RAM chips and an SAA5070 "LUCY".
http://txlib.mb21.co.uk/main.php?g2_view=core.DownloadItem&g2_itemId=20…
ROM selection is done by a PAL but my EEPROM reader has verified that's OK
and I have a dump of it as well as all the ROMs.
Unfortunately it's suffered battery leakage and it was seemingly stored on
its back in a very damp environment so some of the capacitors at the rear of
the board have rotted and bits of the analogue board for the 5" TV had
rusted to nothing, though I've replaced those.
What I originally thought was an analogue board issue that I posted about
here has turned into a total lack of timing issue so I've been tracing out
all the circuits and building a schematic of the board. All was well until I
came to the two chips you can see in the centre of this pic that have no
markings (9B and 10B on the board):
http://www.binarydinosaurs.co.uk/stcexecuteltimingcircuits.jpg
They aren't 'standard' 14-pin DIPs in that they don't follow the
GND-on-pin-7-Vcc-on-pin-14 layout. Pins 6 and 7 on both are wired together
(not to GND) and form the RESET signal for the 8085 via the 7414 at 10A,
source for this signal is unknown currently. Pin 5 on both appears to be Vcc
and pin 10 is GND or at least are pulled high and low respectively.
The XTAL you can see with its supporting resistors is connected to the 7404
at 12B, pin 12 of which goes to pin 1 of 10B (one of the mystery chips) and
pin 1 (X1) of the 8085.
Any clues? I'm going to search for reference 8085 boards (and I guess 8086?)
layouts to see if there are any similarities in timing circuits but for now
I'm stumped.
Cheers!
--
Adrian/Witchy
Binary Dinosaurs creator/curator
Www.binarydinosaurs.co.uk - the UK's biggest private home computer
collection?
16 Dec
16 Dec
3:44 p.m.
On 12/16/2016 08:43 AM, Adrian Graham wrote:
Hi folks,
I have an 8085-based viewdata telephone system on the bench that's proving
to be a labour of love in trying to get it running with zero documentation -
there are only 6 known examples that I've come across and all but 2 of them
are in museums, none known to work. If any of you fancy searching it's an
STC Executel 3910 and at least two of the hits you'll get will be my machine
before I bought it.
http://www.binarydinosaurs.co.uk/stcexecutel01.jpg
I've been in contact with all museums who have one, no luck on docs though
the Museum of Computing which is local to me MAY have some (Jason the owner
has 2 3910s himself), I just need to get down there and search for it :)
Kind-of-fortunately the viewdata side of things seems standard - all 74LS
TTL with an MC3242AP running 16x 4116-2 DRAMs, 27128 EEPROMs etc. The
viewdata side of things is powered by a Plessey MR9735-002 teletext
processor supported by a pair of 2112 RAM chips and an SAA5070 "LUCY".
http://txlib.mb21.co.uk/main.php?g2_view=core.DownloadItem&g2_itemId=20…
ROM selection is done by a PAL but my EEPROM reader has verified that's OK
and I have a dump of it as well as all the ROMs.
Unfortunately it's suffered battery leakage and it was seemingly stored on
its back in a very damp environment so some of the capacitors at the rear of
the board have rotted and bits of the analogue board for the 5" TV had
rusted to nothing, though I've replaced those.
What I originally thought was an analogue board issue that I posted about
here has turned into a total lack of timing issue so I've been tracing out
all the circuits and building a schematic of the board. All was well until I
came to the two chips you can see in the centre of this pic that have no
markings (9B and 10B on the board):
http://www.binarydinosaurs.co.uk/stcexecuteltimingcircuits.jpg
They aren't 'standard' 14-pin DIPs in that they don't follow the
GND-on-pin-7-Vcc-on-pin-14 layout. Pins 6 and 7 on both are wired together
(not to GND) and form the RESET signal for the 8085 via the 7414 at 10A,
source for this signal is unknown currently. Pin 5 on both appears to be Vcc
and pin 10 is GND or at least are pulled high and low respectively.
The XTAL you can see with its supporting resistors is connected to the 7404
at 12B, pin 12 of which goes to pin 1 of 10B (one of the mystery chips) and
pin 1 (X1) of the 8085.
IF I had to throw a wild guess either a baud rate generator chip, an
N-stage counter
like CD4040 or 4060 or one of many low number cmos (CD4xxx parts) wired
as oscillator.
most of those had non TTL style pin outs.
Allison
Any clues? I'm going to search for reference 8085
boards (and I guess 8086?)
layouts to see if there are any similarities in timing circuits but for now
I'm stumped.
Cheers!
5:57 p.m.
From: Adrian Graham: Friday, December 16, 2016 5:43 AM
They aren't 'standard' 14-pin DIPs in that
they don't follow the
GND-on-pin-7-Vcc-on-pin-14 layout. Pins 6 and 7 on both are wired together
(not to GND) and form the RESET signal for the 8085 via the 7414 at 10A,
source for this signal is unknown currently. Pin 5 on both appears to be Vcc
and pin 10 is GND or at least are pulled high and low respectively.
If you had Vcc on 10 and GND on 5, I'd suggest UA733. There aren't a
lot of 14 pin chips with the power pins on 5 and 10. (More with 4 and 11.)
Vince
6:19 p.m.
They aren't 'standard' 14-pin DIPs in that they don't follow the
GND-on-pin-7-Vcc-on-pin-14 layout. Pins 6 and 7 on both are wired together
(not to GND) and form the RESET signal for the 8085 via the 7414 at 10A,
source for this signal is unknown currently. Pin 5 on both appears to be Vcc
and pin 10 is GND or at least are pulled high and low respectively.
Could they be 74(LS)90? Those have Vcc on pin 5 and GND on pin 10.
Regards,
Peter Coghlan.
6:25 p.m.
On Fri, Dec 16, 2016 at 6:19 PM, Peter Coghlan <cctalk at beyondthepale.ie> wrote:
And IIRC pins 6 and 7 are reset _inputs_ on the 74x90 (I don't have the
pinouts to hand, what about the 74x92 or more likely 74x93?). But active
high, not active low. So it's possible the reset signal comes from elsewhere,
resets these counters (as some kind of clock divider chain) and is inverted
by that '14 to feed the 8085 reset input.
I doubt it's a uA733. I can think of no logical reason to have one of those
in this sort of circuit.
-tony
They aren't 'standard' 14-pin DIPs in that they don't follow the
GND-on-pin-7-Vcc-on-pin-14 layout. Pins 6 and 7 on both are wired together
(not to GND) and form the RESET signal for the 8085 via the 7414 at 10A,
source for this signal is unknown currently. Pin 5 on both appears to be Vcc
and pin 10 is GND or at least are pulled high and low respectively.
Could they be 74(LS)90? Those have Vcc on pin 5 and GND on pin 10. 
6:51 p.m.
If one of the 749X parts, one should be able to figure
it out by looking at the NC pins. The different chips look to
have different NC pins. If something is connected to one
for one of them, it would most likely be one of the others.
Dwight
________________________________
From: cctalk <cctalk-bounces at classiccmp.org> on behalf of Tony Duell
<ard.p850ug1 at gmail.com>
Sent: Friday, December 16, 2016 10:25:21 AM
To: General Discussion: On-Topic and Off-Topic Posts
Subject: Re: Mystery 8085-related IC identification needed please
On Fri, Dec 16, 2016 at 6:19 PM, Peter Coghlan <cctalk at beyondthepale.ie> wrote:
And IIRC pins 6 and 7 are reset _inputs_ on the 74x90 (I don't have the
pinouts to hand, what about the 74x92 or more likely 74x93?). But active
high, not active low. So it's possible the reset signal comes from elsewhere,
resets these counters (as some kind of clock divider chain) and is inverted
by that '14 to feed the 8085 reset input.
I doubt it's a uA733. I can think of no logical reason to have one of those
in this sort of circuit.
-tony
They aren't 'standard' 14-pin DIPs in that they don't follow the
GND-on-pin-7-Vcc-on-pin-14 layout. Pins 6 and 7 on both are wired together
(not to GND) and form the RESET signal for the 8085 via the 7414 at 10A,
source for this signal is unknown currently. Pin 5 on both appears to be Vcc
and pin 10 is GND or at least are pulled high and low respectively.
Could they be 74(LS)90? Those have Vcc on pin 5 and GND on pin 10.
17 Dec
17 Dec
2:02 p.m.
On 16/12/2016 18:25, "Tony Duell" <ard.p850ug1 at gmail.com> wrote:
On Fri, Dec 16, 2016 at 6:19 PM, Peter Coghlan
<cctalk at beyondthepale.ie>
wrote:
And IIRC pins 6 and 7 are reset _inputs_ on the 74x90 (I don't have the
pinouts to hand, what about the 74x92 or more likely 74x93?). But active
high, not active low. So it's possible the reset signal comes from elsewhere,
resets these counters (as some kind of clock divider chain) and is inverted
by that '14 to feed the 8085 reset input.
I doubt it's a uA733. I can think of no logical reason to have one of those
in this sort of circuit.
Thanks all, the pinouts are matching the LS9x counters so I just need to
trace more lines to hopefully narrow it down. Pins 6 and 7 are definitely
inputs so you're right Tony, the reset must come from elsewhere. One of the
outputs is confusing though since it appears to come FROM 5V via a resistor,
I can't see anything in the datasheets that hints it may also be used as an
input or that pulling it high affects behaviour. Maybe I mistraced in
tiredness the other night.
Back to the DMM then!
Cheers folks :D
--
Adrian/Witchy
Binary Dinosaurs creator/curator
Www.binarydinosaurs.co.uk - the UK's biggest private home computer
collection?
They aren't 'standard' 14-pin DIPs in that they don't follow the
GND-on-pin-7-Vcc-on-pin-14 layout. Pins 6 and 7 on both are wired together
(not to GND) and form the RESET signal for the 8085 via the 7414 at 10A,
source for this signal is unknown currently. Pin 5 on both appears to be Vcc
and pin 10 is GND or at least are pulled high and low respectively.
Could they be 74(LS)90? Those have Vcc on pin 5 and GND on pin 10.
2:28 p.m.
On Sat, Dec 17, 2016 at 2:02 PM, Adrian Graham
<witchy at binarydinosaurs.co.uk> wrote:
> Thanks all, the pinouts are matching the
LS9x counters so I just need to
> trace more lines to hopefully narrow it down. Pins 6 and 7 are definitely
> inputs so you're right Tony, the reset must come from elsewhere. One of the
> outputs is confusing though since it appears to come FROM 5V via a resistor,
What value resistor?
It is possible that whatever it is driving needs a swing to +5V,
rather than just
a TTL high level. Adding a pull-up resistor is a way to kludge this.
The 7490 has 2 sets of reset inputs (4 pins total), one pair to reset
it to 0, the
other pair to reset it to 9. Since these inputs are active high, they
can't be left
floating, they must be conneccted to something (even if directly to ground). So
you could see if those pins go anywhwere, if not, then you can eliminate the
7490
-tony
> Back to the DMM then!
Given that these counters seem to be reset by the same signal that
resets the 8085, and that I assume the 8085 is not reset very often,
the counters will presumably count continuously. If you have a 'scope,
look at the input signals to the 2 sections of the counter. If you see
nice regular square waves there, then look at the outputs and see if
the frequencies fit the right division ratios for one of the 749x devices.
The ideal tool for this is a LogicDart which will display the waveform
and tell you the frequency (and will drop in your toolbox), but those
are not that easy to get now.
-tony
3:06 p.m.
On 17/12/2016 14:28, Tony Duell wrote:
On Sat, Dec 17, 2016 at 2:02 PM, Adrian Graham
<witchy at binarydinosaurs.co.uk> wrote:
> Thanks all, the pinouts are matching the LS9x
counters so I just need to
> trace more lines to hopefully narrow it down. Pins 6 and 7 are definitely
> inputs so you're right Tony, the reset must come from elsewhere. One of the
> outputs is confusing though since it appears to come FROM 5V via a resistor,
What value resistor?
It is possible that whatever it is driving needs a swing to +5V,
rather than just
a TTL high level. Adding a pull-up resistor is a way to kludge this.
The CLOCK line on a Z80 is usually arranged like this, with a 330R (or
thereabouts) pullup.
--
Pete
Pete Turnbull
6:34 p.m.
On 17/12/2016 14:28, "Tony Duell" <ard.p850ug1 at gmail.com> wrote:
On Sat, Dec 17, 2016 at 2:02 PM, Adrian Graham
<witchy at binarydinosaurs.co.uk> wrote:
> Thanks all, the pinouts are matching
the LS9x counters so I just need to
> trace more lines to hopefully narrow it down. Pins 6 and 7 are definitely
> inputs so you're right Tony, the reset must come from elsewhere. One of the
> outputs is confusing though since it appears to come FROM 5V via a resistor,
What value resistor?
It's a 5-band red-red-black-black-violet so either 220R or 70k? Based on
what Pete said about the Z80 I'm going for 220R without pulling it out of
circuit.
It is possible that whatever it is driving needs a
swing to +5V,
rather than just
a TTL high level. Adding a pull-up resistor is a way to kludge this.
I thought that but it doesn't appear to go anywhere else. I'll keep looking,
its only a 2 layer board so there's nothing hidden.
The 7490 has 2 sets of reset inputs (4 pins total),
one pair to reset
it to 0, the
other pair to reset it to 9. Since these inputs are active high, they
can't be left
floating, they must be conneccted to something (even if directly to ground).
So
you could see if those pins go anywhwere, if not, then you can eliminate the
7490
One problem I have is that I've already found a few chips with dead outputs
so I've no idea if these will be any different. The pinouts I have match the
LS92 since pins 2/3/4/13 are NC. All testing so far has been done with a DMM
and cheap logic analyser. Since one of the possibly-LS92s is out of circuit
I'll build a little test circuit to see if it does actually count given a
clock source...
Cheers!
--
Adrian/Witchy
Binary Dinosaurs creator/curator
Www.binarydinosaurs.co.uk - the UK's biggest private home computer
collection?
6:37 p.m.
It is most likely an off brown, marking a 1% and not violet.
Sometimes their paints change with age.
Dwight
________________________________
From: cctalk <cctalk-bounces at classiccmp.org> on behalf of Adrian Graham
<witchy at binarydinosaurs.co.uk>
Sent: Saturday, December 17, 2016 10:34:01 AM
To: General Discussion: On-Topic and Off-Topic Posts
Subject: Re: Mystery 8085-related IC identification needed please
On 17/12/2016 14:28, "Tony Duell" <ard.p850ug1 at gmail.com> wrote:
On Sat, Dec 17, 2016 at 2:02 PM, Adrian Graham
<witchy at binarydinosaurs.co.uk> wrote:
> Thanks all, the pinouts are matching
the LS9x counters so I just need to
> trace more lines to hopefully narrow it down. Pins 6 and 7 are definitely
> inputs so you're right Tony, the reset must come from elsewhere. One of the
> outputs is confusing though since it appears to come FROM 5V via a resistor,
What value resistor?
It's a 5-band red-red-black-black-violet so either 220R or 70k? Based on
what Pete said about the Z80 I'm going for 220R without pulling it out of
circuit.
It is possible that whatever it is driving needs a
swing to +5V,
rather than just
a TTL high level. Adding a pull-up resistor is a way to kludge this.
I thought that but it doesn't appear to go anywhere else. I'll keep looking,
its only a 2 layer board so there's nothing hidden.
The 7490 has 2 sets of reset inputs (4 pins total),
one pair to reset
it to 0, the
other pair to reset it to 9. Since these inputs are active high, they
can't be left
floating, they must be conneccted to something (even if directly to ground).
So
you could see if those pins go anywhwere, if not, then you can eliminate the
7490
One problem I have is that I've already found a few chips with dead outputs
so I've no idea if these will be any different. The pinouts I have match the
LS92 since pins 2/3/4/13 are NC. All testing so far has been done with a DMM
and cheap logic analyser. Since one of the possibly-LS92s is out of circuit
I'll build a little test circuit to see if it does actually count given a
clock source...
Cheers!
--
Adrian/Witchy
Binary Dinosaurs creator/curator
Www.binarydinosaurs.co.uk - the UK's biggest private home computer
collection?
6:49 p.m.
On Sat, Dec 17, 2016 at 6:34 PM, Adrian Graham
<witchy at binarydinosaurs.co.uk> wrote:
It's not a 7493 as the resets don't match up. I would guess at 74x92 at least
for the moment. See if the rest of the circuit makes any sense.
-tony
On 17/12/2016 14:28, "Tony Duell"
<ard.p850ug1 at gmail.com> wrote:
Yes. I would agree with that.
On Sat, Dec 17, 2016 at 2:02 PM, Adrian Graham
<witchy at binarydinosaurs.co.uk> wrote:
> Thanks all, the pinouts are matching
the LS9x counters so I just need to
> trace more lines to hopefully narrow it down. Pins 6 and 7 are definitely
> inputs so you're right Tony, the reset must come from elsewhere. One of the
> outputs is confusing though since it appears to come FROM 5V via a resistor,
What value resistor?
It's a 5-band red-red-black-black-violet so either 220R or 70k? Based on
what Pete said about the Z80 I'm going for 220R without pulling it out of
circuit.
It is possible
that whatever it is driving needs a swing to +5V,
rather than just
a TTL high level. Adding a pull-up resistor is a way to kludge this.
I thought that but it doesn't appear to go anywhere else. I'll keep looking,
its only a 2 layer board so there's nothing hidden.
The 7490 has 2 sets of reset inputs (4 pins
total), one pair to reset
it to 0, the
other pair to reset it to 9. Since these inputs are active high, they
can't be left
floating, they must be conneccted to something (even if directly to ground).
So
you could see if those pins go anywhwere, if not, then you can eliminate the
7490
One problem I have is that I've already found a few chips with dead outputs
so I've no idea if these will be any different. The pinouts I have match the
LS92 since pins 2/3/4/13 are NC. All testing so far has been done with a DMM
and cheap logic analyser. Since one of the possibly-LS92s is out of circuit
I'll build a little test circuit to see if it does actually count given a
clock source...
19 Dec
19 Dec
1:09 a.m.
On 17/12/2016 18:49, "Tony Duell" <ard.p850ug1 at gmail.com> wrote:
Yes. I would agree with that.
It connects to some telephone-based doohickey that can be seen as the
right-most light-green module lower centre left of this pic, the one with
the red capacitor in the middle of it:
http://www.binarydinosaurs.co.uk/stcexecutelboard.jpg
Initially I figured the board was divided half and half into phone and
digital viewdata with the viewdata part on the right. Now I'm tracing more
lines that blur between the pair of them right down the centre, including
the source of the 8085's RESET which would seem to be the circuitry in the
top-middle of that pic where you can see the battery (now removed).
The chip directly below the battery is the RTC, to its left is an MC14081
(quad input AND) and left again is an ICL7611 op-amp. Those 3 have their own
5V supply that's separate from the rest of the board, yay. Not found the
source of that yet.
I'm 99% sure they're 7492s now, outputs and inputs match other surrounding
chips.
Cheers!
--
Adrian/Witchy
Binary Dinosaurs creator/curator
Www.binarydinosaurs.co.uk - the UK's biggest private home computer
collection?
What value resistor?
It's a 5-band red-red-black-black-violet so either 220R or 70k? Based on
what Pete said about the Z80 I'm going for 220R without pulling it out of
circuit.
LS92 since
pins 2/3/4/13 are NC. All testing so far has been done with a DMM
and cheap logic analyser. Since one of the possibly-LS92s is out of circuit
I'll build a little test circuit to see if it does actually count given a
clock source...
It's not a 7493 as the resets don't match up. I would guess at 74x92 at least
for the moment. See if the rest of the circuit makes any sense. 
21 Dec
21 Dec
7:37 p.m.
On 2016-Dec-16, at 5:43 AM, Adrian Graham wrote:
viewdata side of things is powered by a Plessey
MR9735-002 teletext
processor supported by a pair of 2112 RAM chips and an SAA5070 "LUCY".
http://txlib.mb21.co.uk/main.php?g2_view=core.DownloadItem&g2_itemId=20…
On 2016-Dec-17, at 10:34 AM, Adrian Graham wrote:
One problem I have is that I've already found a
few chips with dead outputs
so I've no idea if these will be any different. The pinouts I have match the
LS92 since pins 2/3/4/13 are NC. All testing so far has been done with a DMM
and cheap logic analyser. Since one of the possibly-LS92s is out of circuit
I'll build a little test circuit to see if it does actually count given a
clock source...
The counters may be the beginning of a video timing divider chain or a clock divider for
the LUCY chip.
(e.g. a diagram in the document you mentioned shows a div-6 between a 6MHz crystal osc and
the LUCY chip.
The division factor may vary of course in the executel depending on the crystal.)
Also to keep in mind as you trace it out, 749x counters were often used with gates between
the outputs and reset inputs to change the count modulus / division factor.
On 2016-Dec-17, at 10:34 AM, Adrian Graham wrote:
It's a 5-band red-red-black-black-violet so either
220R or 70k? Based on
what Pete said about the Z80 I'm going for 220R without pulling it out of
circuit.
Looking at the photo in your earlier message
http://www.binarydinosaurs.co.uk/stcexecuteltimingcircuits.jpg
the wide red band on the resistors would be a 2% tolerance indicator.
As with Dwight's suggestion about colour confusion, if the resistors in question are
those to the left of the two unmarked chips (or similar), they look like
brown-black-black-red-wideRed, which would be standard 10K / 2%.
You can try measuring resistors in circuits like this with a DMM.
Switch the DMM leads to get both polarities through the resistor.
It's not guaranteed as it depends on the connected circuitry and the R value, but one
of the readings will often be your R value, or may confirm or give direction as to the
value.
7:44 p.m.
On Wed, Dec 21, 2016 at 7:37 PM, Brent Hilpert <hilpert at cs.ubc.ca> wrote:
The counters may be the beginning of a video timing
divider chain or a clock divider for the LUCY chip.
(e.g. a diagram in the document you mentioned shows a div-6 between a 6MHz crystal osc
and the LUCY chip.
The division factor may vary of course in the executel depending on the crystal.)
Also to keep in mind as you trace it out, 749x counters were often used with gates
between the
outputs and reset inputs to change the count modulus / division factor.
I seem to remember Adrian said that the reset pins on these counters (assuming
they are 7492s, and those pins are reset) come from an inverted version of
the 8085 reset. I think it's unlikely that the 8085 will be
continually reset, so
I would think the counters were not reset by a combnation of their outputs.
[Unknown resistor]
You can try measuring resistors in circuits like this
with a DMM.
Switch the DMM leads to get both polarities through the resistor.
It's not guaranteed as it depends on the connected circuitry and the R value, but
one of the readings will often be your R value, or may confirm or give direction as to
the value.
It's worth remembering that if you put the DMM leads across the resistor,
then the reading can only be affected by other paths in parallel, and these
will reduce the displayed resistance reading, never increase it.
So if you have a resistor that might be either 2k2 or 33k (think about it)
and you measure it in-circuit as 25.4k then, assuming it is not faulty,
it must be 33k.
-tony
8:45 p.m.
On 2016-Dec-21, at 11:44 AM, Tony Duell wrote:
On Wed, Dec 21, 2016 at 7:37 PM, Brent Hilpert
<hilpert at cs.ubc.ca> wrote:
He did mention they (pins 6,7 and 8085 reset) were in association with a 7414 but it
wasn't entirely clear or ensured they were part of the same logic path as opposed to
merely using separate inverters in the same chip. Timing divider chains typically
don't need to be reset in sync with the system microproc, so there's the
possibility, if not likelihood, they're logically unrelated; hence modulus alteration
is something to keep in mind.
The counters may be the beginning of a video
timing divider chain or a clock divider for the LUCY chip.
(e.g. a diagram in the document you mentioned shows a div-6 between a 6MHz crystal osc
and the LUCY chip.
The division factor may vary of course in the executel depending on the crystal.)
Also to keep in mind as you trace it out, 749x counters were often used with gates
between the
outputs and reset inputs to change the count modulus / division factor.
I seem to remember Adrian said that the reset pins on these counters (assuming
they are 7492s, and those pins are reset) come from an inverted version of
the 8085 reset. I think it's unlikely that the 8085 will be
continually reset, so
I would think the counters were not reset by a combnation of their outputs.
22 Dec
22 Dec
6:52 p.m.
On 21/12/2016 19:37, "Brent Hilpert" <hilpert at cs.ubc.ca> wrote:
On 2016-Dec-17, at 10:34 AM, Adrian Graham wrote:
That's a good point so I've revisited that section of board - one of the
things that started me down this whole path was a missing video sync before
I realised that the CPU wasn't running because of RESET and have spent many
days since tracing and measuring trying to find the source. I'm pretty sure
the LUCY chip is only there as a modem since the teletext display side of
things is handled by a Plessey MR9735 and a pair of 2114 RAM chips as page
store.
The board's also fairly logically laid out, split 50/50 between phone and
teletext with the teletext side split between CPU/glue, ROMs, RAM and
display logic. LUCY is well over to the left on the phone side.
One problem I have is that I've already found
a few chips with dead outputs
so I've no idea if these will be any different. The pinouts I have match the
LS92 since pins 2/3/4/13 are NC. All testing so far has been done with a DMM
and cheap logic analyser. Since one of the possibly-LS92s is out of circuit
I'll build a little test circuit to see if it does actually count given a
clock source...
The counters may be the beginning of a video timing divider chain or a clock
divider for the LUCY chip.
(e.g. a diagram in the document you mentioned shows a div-6 between a 6MHz
crystal osc and the LUCY chip.
The division factor may vary of course in the executel depending on the
crystal.) Also to keep in mind as you trace it out, 749x
counters were often used with
gates between the outputs and reset inputs to change the count modulus /
division factor.
Hopefully there's nothing complex here, the reset inputs are inverted via a
74LS14, input to that is from the CMOS MC14081B mentioned elsewhere through
two hops over pins 1-2 then 3-4 of a 74LS04N.
On 2016-Dec-17, at 10:34 AM, Adrian Graham wrote:
Those are pullups on an output from one of the '92s over to one of the 3
light green telephony bits that I've not looked up, marked 'NMC1515/6/7'
It's a 5-band red-red-black-black-violet so
either 220R or 70k? Based on
what Pete said about the Z80 I'm going for 220R without pulling it out of
circuit.
Looking at the photo in your earlier message
http://www.binarydinosaurs.co.uk/stcexecuteltimingcircuits.jpg
the wide red band on the resistors would be a 2% tolerance indicator.
As with Dwight's suggestion about colour confusion, if the resistors in
question are those to the left of the two unmarked chips (or similar), they
look like brown-black-black-red-wideRed, which would be standard 10K / 2%. You can try measuring resistors in circuits like this
with a DMM.
Switch the DMM leads to get both polarities through the resistor.
It's not guaranteed as it depends on the connected circuitry and the R value,
but one of the readings will often be your R value, or may confirm or give
direction as to the value.
I normally pull up one end and test out of circuit but I'll give that a go
too.
Cheers!
--
Adrian/Witchy
Binary Dinosaurs creator/curator
Www.binarydinosaurs.co.uk - the UK's biggest private home computer
collection?
2928
days inactive
2934
days old
16 comments
8 participants
participants (8)
-
ajp166＠verizon.net -
ard.p850ug1＠gmail.com -
cctalk＠beyondthepale.ie -
dkelvey＠hotmail.com -
hilpert＠cs.ubc.ca -
pete＠dunnington.plus.com -
v.slyngstad＠frontier.com -
witchy＠binarydinosaurs.co.uk