This is going to seem like a strange thing to want, but is it possible to slow down a Z80 pod? I have a 9010 and the Z80 pod for it. As far as I know, both are working fine. I don't have anything with a Z80 in it to test to prove that. What I picked these up for is to use with a Signetics 2650 adapter (http://www.arcades.plus.com/s2650_fluke_adapter.htm) on Zaccaria boards. It seems like this should work, but in practice it's very unreliable. Test reading data from ROM on a known working board, sometimes I get correct data, most of the time I get random garbage. If I read a byte from the same address 10 times, maybe one time will be correct. Test RAM, it'll go a few bytes in to the test, then "fail". If I loop on that, it'll eventually pass, then fail a byte or two later. This seems to be essentially the same symptom as displayed by the ROM reading test. The adapter board is based on the old Fluke "Troubleshooter" article from 1983. As far as I can see, the board is doing what it's supposed to do, moving signals from the Z80 to where the 2650 expects them to be. So what I'm thinking is that the Z80, running at a faster clock speed than the 2650, is just over-running what a board designed for a slower processor can deliver. Is it possible to slow down the clock speed on the Z80 pod? _______________________________________________ Techtoolslist mailing list Techtoolslist@flippers.com https://pairlist7.pair.net/mailman/listinfo/techtoolslist FTP site is: ftp://ttl.arcadetech.org/TTL/Test_Equipment Archive site: http://seven.pairlist.net/pipermail/techtoolslist/
The pod normally uses the uut clock (and rate). Also the link you posted for the adapter does not work for me. On Sep 17, 2017 9:47 AM, "David Gersic" <info@zaccaria-pinball.com> wrote: So what I'm thinking is that the Z80, running at a faster clock speed than the 2650, is just over-running what a board designed for a slower processor can deliver. Is it possible to slow down the clock speed on the Z80 pod? _______________________________________________ Techtoolslist mailing list Techtoolslist@flippers.com https://pairlist7.pair.net/mailman/listinfo/techtoolslist FTP site is: ftp://ttl.arcadetech.org/TTL/Test_Equipment Archive site: http://seven.pairlist.net/pipermail/techtoolslist/ _______________________________________________ Techtoolslist mailing list Techtoolslist@flippers.com https://pairlist7.pair.net/mailman/listinfo/techtoolslist FTP site is: ftp://ttl.arcadetech.org/TTL/Test_Equipment Archive site: http://seven.pairlist.net/pipermail/techtoolslist/
Here’s the correct link: http://www.macros-arcade.com/s2650_fluke_adapter.htm Tony Jones writes:
The pod normally uses the uut clock (and rate). Also the link you posted for the adapter does not work for me.
On Sep 17, 2017 9:47 AM, "David Gersic" <info@zaccaria-pinball.com> wrote:
So what I'm thinking is that the Z80, running at a faster clock speed than the 2650, is just over-running what a board designed for a slower processor can deliver. Is it possible to slow down the clock speed on the Z80 pod?
_______________________________________________ Techtoolslist mailing list Techtoolslist@flippers.com https://pairlist7.pair.net/mailman/listinfo/techtoolslist FTP site is: ftp://ttl.arcadetech.org/TTL/Test_Equipment Archive site: http://seven.pairlist.net/pipermail/techtoolslist/ _______________________________________________ Techtoolslist mailing list Techtoolslist@flippers.com https://pairlist7.pair.net/mailman/listinfo/techtoolslist FTP site is: ftp://ttl.arcadetech.org/TTL/Test_Equipment Archive site: http://seven.pairlist.net/pipermail/techtoolslist/
_______________________________________________ Techtoolslist mailing list Techtoolslist@flippers.com https://pairlist7.pair.net/mailman/listinfo/techtoolslist FTP site is: ftp://ttl.arcadetech.org/TTL/Test_Equipment Archive site: http://seven.pairlist.net/pipermail/techtoolslist/
On Sunday, September 17, 2017 12:16:19 PM Tony Jones wrote:
The pod normally uses the uut clock (and rate). Also the link you posted for the adapter does not work for me.
Looks like he moved it to http://www.macros-arcade.com/s2650_fluke_adapter.htm
On Sep 17, 2017 9:47 AM, "David Gersic" <info@zaccaria-pinball.com> wrote:
So what I'm thinking is that the Z80, running at a faster clock speed than the 2650, is just over-running what a board designed for a slower processor can deliver. Is it possible to slow down the clock speed on the Z80 pod?
_______________________________________________ Techtoolslist mailing list Techtoolslist@flippers.com https://pairlist7.pair.net/mailman/listinfo/techtoolslist FTP site is: ftp://ttl.arcadetech.org/TTL/Test_Equipment Archive site: http://seven.pairlist.net/pipermail/techtoolslist/ _______________________________________________ Techtoolslist mailing list Techtoolslist@flippers.com https://pairlist7.pair.net/mailman/listinfo/techtoolslist FTP site is: ftp://ttl.arcadetech.org/TTL/Test_Equipment Archive site: http://seven.pairlist.net/pipermail/techtoolslist/
Techtoolslist mailing list Techtoolslist@flippers.com https://pairlist7.pair.net/mailman/listinfo/techtoolslist FTP site is: ftp://ttl.arcadetech.org/TTL/Test_Equipment Archive site: http://seven.pairlist.net/pipermail/techtoolslist/
That’s not really how it works. The CPU in the pod is driven by the clock on the PCB. If the pod CPU was running at a different clock speed, nothing would work at all. Based on the article, it sounds like perhaps the adapter just doesn’t match the timings of the original CPU to support anything other than individual reads, or perhaps there’s an issue with your adapter. I’d see if anyone else has used a similar setup and see what results they got. If you wanted to slow things down in the hopes of improving marginal timings, you’d need to alter the clock on the PCB, either by substituting a slowre crystal or by replacing the clock that controls the CPU and bus accesses with a slower signal. Either option almost certainly guarantees that the video hardware won’t work, and may mask other problems going on with the board. Are you aware of Paul Swan’s 2650 tester? That might be another option worth looking into. http://www.paulswan.me/arcade/ArduinoMegaICT.htm David Gersic writes:
This is going to seem like a strange thing to want, but is it possible to slow down a Z80 pod?
I have a 9010 and the Z80 pod for it. As far as I know, both are working fine. I don't have anything with a Z80 in it to test to prove that.
What I picked these up for is to use with a Signetics 2650 adapter (http://www.arcades.plus.com/s2650_fluke_adapter.htm) on Zaccaria boards. It seems like this should work, but in practice it's very unreliable.
Test reading data from ROM on a known working board, sometimes I get correct data, most of the time I get random garbage. If I read a byte from the same address 10 times, maybe one time will be correct.
Test RAM, it'll go a few bytes in to the test, then "fail". If I loop on that, it'll eventually pass, then fail a byte or two later. This seems to be essentially the same symptom as displayed by the ROM reading test.
The adapter board is based on the old Fluke "Troubleshooter" article from 1983. As far as I can see, the board is doing what it's supposed to do, moving signals from the Z80 to where the 2650 expects them to be.
So what I'm thinking is that the Z80, running at a faster clock speed than the 2650, is just over-running what a board designed for a slower processor can deliver. Is it possible to slow down the clock speed on the Z80 pod?
_______________________________________________ Techtoolslist mailing list Techtoolslist@flippers.com https://pairlist7.pair.net/mailman/listinfo/techtoolslist FTP site is: ftp://ttl.arcadetech.org/TTL/Test_Equipment Archive site: http://seven.pairlist.net/pipermail/techtoolslist/
_______________________________________________ Techtoolslist mailing list Techtoolslist@flippers.com https://pairlist7.pair.net/mailman/listinfo/techtoolslist FTP site is: ftp://ttl.arcadetech.org/TTL/Test_Equipment Archive site: http://seven.pairlist.net/pipermail/techtoolslist/
On Sunday, September 17, 2017 12:24:07 PM Ian Eure wrote:
That’s not really how it works. The CPU in the pod is driven by the clock on the PCB. If the pod CPU was running at a different clock speed, nothing would work at all.
Hm. So the adapter ports CLOCK straight across, from Z80 pin 6 so 2650 pin 38. The Z80, being capable of running at faster clock speeds, should be perfectly happy to run slower.
Based on the article, it sounds like perhaps the adapter just doesn’t match the timings of the original CPU to support anything other than individual reads, or perhaps there’s an issue with your adapter. I’d see if anyone else has used a similar setup and see what results they got.
The old Fluke article seems like they were trying to match the timings, or at least get close enough. Maybe they were good enough for whatever it was they were testing when the article was written. I don't know anybody else that works on Zaccaria boards, so no, haven't got anybody else that can confirm or deny that this works at all.
If you wanted to slow things down in the hopes of improving marginal timings, you’d need to alter the clock on the PCB, either by substituting a slowre crystal or by replacing the clock that controls the CPU and bus accesses with a slower signal. Either option almost certainly guarantees that the video hardware won’t work, and may mask other problems going on with the board.
Right now, I'm testing on boards that I know to be good. I could try swapping in a slower crystal to see what happens, good or bad. But if the pod runs at UUT clock speed, I'm thinking that wouldn't be likely to make a difference.
Are you aware of Paul Swan’s 2650 tester? That might be another option worth looking into. http://www.paulswan.me/arcade/ArduinoMegaICT.htm
Yes, I've seen that too. _______________________________________________ Techtoolslist mailing list Techtoolslist@flippers.com https://pairlist7.pair.net/mailman/listinfo/techtoolslist FTP site is: ftp://ttl.arcadetech.org/TTL/Test_Equipment Archive site: http://seven.pairlist.net/pipermail/techtoolslist/
It works on a Zaccaria Invaders PCB, as I have used it on a few of those before. mind you these are nice and simple, no pages as everything fits into the first 2k !!
On 19 September 2017 at 06:12 David Gersic <info@zaccaria-pinball.com> wrote:
On Sunday, September 17, 2017 12:24:07 PM Ian Eure wrote:
That's not really how it works. The CPU in the pod is driven by the clock on the PCB. If the pod CPU was running at a different clock speed, nothing would work at all.
Hm. So the adapter ports CLOCK straight across, from Z80 pin 6 so 2650 pin 38. The Z80, being capable of running at faster clock speeds, should be perfectly happy to run slower.
Based on the article, it sounds like perhaps the adapter just doesn't match the timings of the original CPU to support anything other than individual reads, or perhaps there's an issue with your adapter. I'd see if anyone else has used a similar setup and see what results they got.
The old Fluke article seems like they were trying to match the timings, or at least get close enough. Maybe they were good enough for whatever it was they were testing when the article was written.
I don't know anybody else that works on Zaccaria boards, so no, haven't got anybody else that can confirm or deny that this works at all.
If you wanted to slow things down in the hopes of improving marginal timings, you'd need to alter the clock on the PCB, either by substituting a slowre crystal or by replacing the clock that controls the CPU and bus accesses with a slower signal. Either option almost certainly guarantees that the video hardware won't work, and may mask other problems going on with the board.
Right now, I'm testing on boards that I know to be good. I could try swapping in a slower crystal to see what happens, good or bad. But if the pod runs at UUT clock speed, I'm thinking that wouldn't be likely to make a difference.
Are you aware of Paul Swan's 2650 tester? That might be another option worth looking into. http://www.paulswan.me/arcade/ArduinoMegaICT.htm
Yes, I've seen that too.
_______________________________________________ Techtoolslist mailing list Techtoolslist@flippers.com https://pairlist7.pair.net/mailman/listinfo/techtoolslist FTP site is: ftp://ttl.arcadetech.org/TTL/Test_Equipment Archive site: http://seven.pairlist.net/pipermail/techtoolslist/
Techtoolslist mailing list Techtoolslist@flippers.com https://pairlist7.pair.net/mailman/listinfo/techtoolslist FTP site is: ftp://ttl.arcadetech.org/TTL/Test_Equipment Archive site: http://seven.pairlist.net/pipermail/techtoolslist/
David Gersic writes:
On Sunday, September 17, 2017 12:24:07 PM Ian Eure wrote:
That’s not really how it works. The CPU in the pod is driven by the clock on the PCB. If the pod CPU was running at a different clock speed, nothing would work at all.
Hm. So the adapter ports CLOCK straight across, from Z80 pin 6 so 2650 pin 38. The Z80, being capable of running at faster clock speeds, should be perfectly happy to run slower.
Agreed.
Based on the article, it sounds like perhaps the adapter just doesn’t match the timings of the original CPU to support anything other than individual reads, or perhaps there’s an issue with your adapter. I’d see if anyone else has used a similar setup and see what results they got.
The old Fluke article seems like they were trying to match the timings, or at least get close enough. Maybe they were good enough for whatever it was they were testing when the article was written.
I don't know anybody else that works on Zaccaria boards, so no, haven't got anybody else that can confirm or deny that this works at all.
Paul Swan does, and has a huge amount of repair/restore logs at http://zzzaccaria.com/ -- I’ve emailed him before. He made the tester I linked, which he uses for all his repairs, so I don’t know if he’d be helpful in this case. But a good guy to know about if you’re working on Zaccaria boards.
If you wanted to slow things down in the hopes of improving marginal timings, you’d need to alter the clock on the PCB, either by substituting a slowre crystal or by replacing the clock that controls the CPU and bus accesses with a slower signal. Either option almost certainly guarantees that the video hardware won’t work, and may mask other problems going on with the board.
Right now, I'm testing on boards that I know to be good. I could try swapping in a slower crystal to see what happens, good or bad. But if the pod runs at UUT clock speed, I'm thinking that wouldn't be likely to make a difference.
It definitely doesn’t. It has to run from the PCB’s clocks. _______________________________________________ Techtoolslist mailing list Techtoolslist@flippers.com https://pairlist7.pair.net/mailman/listinfo/techtoolslist FTP site is: ftp://ttl.arcadetech.org/TTL/Test_Equipment Archive site: http://seven.pairlist.net/pipermail/techtoolslist/
On 2017/09/17 9:38 AM, David Gersic wrote:
This is going to seem like a strange thing to want, but is it possible to slow down a Z80 pod?
I have a 9010 and the Z80 pod for it. As far as I know, both are working fine. I don't have anything with a Z80 in it to test to prove that.
What I picked these up for is to use with a Signetics 2650 adapter (http://www.arcades.plus.com/s2650_fluke_adapter.htm) on Zaccaria boards. It seems like this should work, but in practice it's very unreliable.
Test reading data from ROM on a known working board, sometimes I get correct data, most of the time I get random garbage. If I read a byte from the same address 10 times, maybe one time will be correct.
Test RAM, it'll go a few bytes in to the test, then "fail". If I loop on that, it'll eventually pass, then fail a byte or two later. This seems to be essentially the same symptom as displayed by the ROM reading test.
The adapter board is based on the old Fluke "Troubleshooter" article from 1983. As far as I can see, the board is doing what it's supposed to do, moving signals from the Z80 to where the 2650 expects them to be.
So what I'm thinking is that the Z80, running at a faster clock speed than the 2650, is just over-running what a board designed for a slower processor can deliver. Is it possible to slow down the clock speed on the Z80 pod?
_______________________________________________ Techtoolslist mailing list Techtoolslist@flippers.com https://pairlist7.pair.net/mailman/listinfo/techtoolslist FTP site is: ftp://ttl.arcadetech.org/TTL/Test_Equipment Archive site: http://seven.pairlist.net/pipermail/techtoolslist/
Silly question, but have you tried this adapter on a known-to-be-good 2650 board? Your UUT may be defective such that you get those results (ducking). John :-#)# -- How to subscribe or unsubscribe from TTL http://seven.pairlist.net/mailman/listinfo/techtoolslist _______________________________________________ Techtoolslist mailing list Techtoolslist@flippers.com https://pairlist7.pair.net/mailman/listinfo/techtoolslist FTP site is: ftp://ttl.arcadetech.org/TTL/Test_Equipment Archive site: http://seven.pairlist.net/pipermail/techtoolslist/
On 2017/09/17 4:23 PM, John Robertson wrote:
On 2017/09/17 9:38 AM, David Gersic wrote:
This is going to seem like a strange thing to want, but is it possible to slow down a Z80 pod?
I have a 9010 and the Z80 pod for it. As far as I know, both are working fine. I don't have anything with a Z80 in it to test to prove that.
What I picked these up for is to use with a Signetics 2650 adapter (http://www.arcades.plus.com/s2650_fluke_adapter.htm) on Zaccaria boards. It seems like this should work, but in practice it's very unreliable.
Test reading data from ROM on a known working board, sometimes I get correct data, most of the time I get random garbage. If I read a byte from the same address 10 times, maybe one time will be correct.
Test RAM, it'll go a few bytes in to the test, then "fail". If I loop on that, it'll eventually pass, then fail a byte or two later. This seems to be essentially the same symptom as displayed by the ROM reading test.
The adapter board is based on the old Fluke "Troubleshooter" article from 1983. As far as I can see, the board is doing what it's supposed to do, moving signals from the Z80 to where the 2650 expects them to be.
So what I'm thinking is that the Z80, running at a faster clock speed than the 2650, is just over-running what a board designed for a slower processor can deliver. Is it possible to slow down the clock speed on the Z80 pod?
_______________________________________________ Techtoolslist mailing list Techtoolslist@flippers.com https://pairlist7.pair.net/mailman/listinfo/techtoolslist FTP site is: ftp://ttl.arcadetech.org/TTL/Test_Equipment Archive site: http://seven.pairlist.net/pipermail/techtoolslist/
Silly question, but have you tried this adapter on a known-to-be-good 2650 board? Your UUT may be defective such that you get those results (ducking).
John :-#)#
Crap, you DID say KNOW TO BE WORKING 2650 board. !@#$. Just ignore me. John :-#(# -- How to subscribe or unsubscribe from TTL http://seven.pairlist.net/mailman/listinfo/techtoolslist _______________________________________________ Techtoolslist mailing list Techtoolslist@flippers.com https://pairlist7.pair.net/mailman/listinfo/techtoolslist FTP site is: ftp://ttl.arcadetech.org/TTL/Test_Equipment Archive site: http://seven.pairlist.net/pipermail/techtoolslist/
It may be easier to speed up the board....cool the memory you're trying to test. Might just give you enough timing-wise to see where the hard failures are. On Sep 17, 2017 6:24 PM, "John Robertson" <jrr@flippers.com> wrote:
On 2017/09/17 4:23 PM, John Robertson wrote:
On 2017/09/17 9:38 AM, David Gersic wrote:
This is going to seem like a strange thing to want, but is it possible to slow down a Z80 pod?
I have a 9010 and the Z80 pod for it. As far as I know, both are working fine. I don't have anything with a Z80 in it to test to prove that.
What I picked these up for is to use with a Signetics 2650 adapter (http://www.arcades.plus.com/s2650_fluke_adapter.htm) on Zaccaria boards. It seems like this should work, but in practice it's very unreliable.
Test reading data from ROM on a known working board, sometimes I get correct data, most of the time I get random garbage. If I read a byte from the same address 10 times, maybe one time will be correct.
Test RAM, it'll go a few bytes in to the test, then "fail". If I loop on that, it'll eventually pass, then fail a byte or two later. This seems to be essentially the same symptom as displayed by the ROM reading test.
The adapter board is based on the old Fluke "Troubleshooter" article from 1983. As far as I can see, the board is doing what it's supposed to do, moving signals from the Z80 to where the 2650 expects them to be.
So what I'm thinking is that the Z80, running at a faster clock speed than the 2650, is just over-running what a board designed for a slower processor can deliver. Is it possible to slow down the clock speed on the Z80 pod?
_______________________________________________ Techtoolslist mailing list Techtoolslist@flippers.com https://pairlist7.pair.net/mailman/listinfo/techtoolslist FTP site is: ftp://ttl.arcadetech.org/TTL/Test_Equipment Archive site: http://seven.pairlist.net/pipermail/techtoolslist/
Silly question, but have you tried this adapter on a known-to-be-good 2650 board? Your UUT may be defective such that you get those results (ducking).
John :-#)#
Crap, you DID say KNOW TO BE WORKING 2650 board. !@#$.
Just ignore me.
John :-#(#
-- How to subscribe or unsubscribe from TTL http://seven.pairlist.net/mail man/listinfo/techtoolslist
_______________________________________________ Techtoolslist mailing list Techtoolslist@flippers.com https://pairlist7.pair.net/mailman/listinfo/techtoolslist FTP site is: ftp://ttl.arcadetech.org/TTL/Test_Equipment Archive site: http://seven.pairlist.net/pipermail/techtoolslist/
participants (6)
-
David Gersic -
Ian Eure -
John Robertson -
Mike Coates -
Rodger Boots -
Tony Jones