UK Vintage Radio Repair and Restoration Powered By Google Custom Search Vintage Radio and TV Service Data

Go Back   UK Vintage Radio Repair and Restoration Discussion Forum > Specific Vintage Equipment > Vintage Computers

Notices

Vintage Computers Any vintage computer systems, calculators, video games etc., but with an emphasis on 1980s and earlier equipment.

Closed Thread
 
Thread Tools
Old 10th Mar 2021, 9:47 pm   #881
Timbucus
Octode
 
Join Date: Mar 2019
Location: Barry, Vale of Glamorgan, Wales, UK.
Posts: 1,362
Default Re: Non-working Commodore PET 3016

Quote:
Originally Posted by ScottishColin View Post
11-14 on UF7 and UF8 are steady high.

Colin.

Quote:
Originally Posted by SiriusHardware View Post
Quote:
check the shape of SD0-SD3 UF8 11-14 and the same for UF7 pins to see if they are nice clean 0-4V square waves
Those pins (four from UF7, four from UF8) go to the eight inputs on UF9 and Colin says he has steady high on all of those.

Doesn't mean that it is not also worth looking on UF7 pins 11-14 and UF8 pins 11-14 but it's hard to imagine where all 8 of those signals would go missing on the way to UF9.
Well it was worth looking - that is information in itself - if they are all 1 then the output of UF9 must be 3B constantly to generate that character - or something related. Can we confirm that the bit pattern is as we expect so I would think:

0 - Pin 16
1 - Pin 19
1 - Pin 12
1 - Pin 9
0 - Pin 6
1 - Pin 5
1 - Pin 2
Timbucus is offline  
Old 10th Mar 2021, 9:49 pm   #882
ScottishColin
Octode
 
Join Date: May 2012
Location: Perth, Scotland
Posts: 1,762
Default Re: Non-working Commodore PET 3016

I cut each pin above the motherboard and remove the carcass of the chip - it's a one-way process...

Then I de-solder each pin from underneath individually and remove each pin one by one - takes 10-15 minutes or so if I'm careful.

Then I pop in a turned pin socket and solder that in from underneath. if I have lost a pad (which I don't do much nowadays), I use your trick of putting some wires through the hole before I put the socket in place.

Then I go through the schematics and perform continuity tests for each pin to make sure I have an electrical connection before I put any chips into sockets.

Colin.


Quote:
Originally Posted by SiriusHardware View Post
Quote:
Yep - all RAM chips are soldered on this motherboard.
Which lands us with a bit of a headache, because the simple tests which Tim and AJ suggested are not easily possible.

So far when we have suspected an IC might be dud they have always been something quite run of the mill and with the exception of the CPU, relatively cheap and easy to replace. With something as exotic as the 6114 RAMs I
really prefer not to take the approach of hacking them to pieces on the off chance that they may be faulty.

PET specialists among us - is it possible that the video RAM is being intentionally filled with FF during the early stages of initialisation?

Colin: You're obviously a lot more comfortable with removing and replacing ICs now, but how have you been doing that, by cutting the pins, removing the chip body and then removing the pins individually?
ScottishColin is online now  
Old 10th Mar 2021, 9:50 pm   #883
Mark1960
Octode
 
Join Date: Mar 2020
Location: Kitchener, Ontario, Canada
Posts: 1,264
Default Re: Non-working Commodore PET 3016

Shortly after power on there should be activity on UF7-10 and UF8-10, before it becomes steady high level. This is assuming the 6502 is running the display clear routine.

If you are able to see activity on UF7-10 and UF8-10, then use the second probe to look at the data on BD0 to BD7, and then also on SD0 to SD7.

I think you should see hex 20 the space character on both sets of data lines while UF7-10 and UF8-10 are at low level. This would be all data lines low except for SD5 and BD5 which should both be high.

This is probably going to mean turning the power off and back on for each test. So do this with About 30 seconds between turn off and turn back on, don’t turn the power on and off quickly.

Last edited by Mark1960; 10th Mar 2021 at 9:55 pm. Reason: Add note on power cycling
Mark1960 is online now  
Old 10th Mar 2021, 9:52 pm   #884
ScottishColin
Octode
 
Join Date: May 2012
Location: Perth, Scotland
Posts: 1,762
Default Re: Non-working Commodore PET 3016

Here's a page on PETSCII. the semi-colon is 3B which is 0011101 in binary

https://en.wikipedia.org/wiki/PETSCII

Colin.


Quote:
Originally Posted by Timbucus View Post
Quote:
Originally Posted by ScottishColin View Post
11-14 on UF7 and UF8 are steady high.

Colin.

Quote:
Originally Posted by SiriusHardware View Post

Those pins (four from UF7, four from UF8) go to the eight inputs on UF9 and Colin says he has steady high on all of those.

Doesn't mean that it is not also worth looking on UF7 pins 11-14 and UF8 pins 11-14 but it's hard to imagine where all 8 of those signals would go missing on the way to UF9.
Well it was worth looking - that is information in itself - if they are all 1 then the output of UF9 must be 3B constantly to generate that character - or something related. Can we confirm that the bit pattern is as we expect so I would think:

0 - Pin 16
1 - Pin 19
1 - Pin 12
1 - Pin 9
0 - Pin 6
1 - Pin 5
1 - Pin 2
ScottishColin is online now  
Old 10th Mar 2021, 9:54 pm   #885
ScottishColin
Octode
 
Join Date: May 2012
Location: Perth, Scotland
Posts: 1,762
Default Re: Non-working Commodore PET 3016

UF7/10 and UF8/10 go straight high at power on to 4.08V and stay there.

Colin.


Quote:
Originally Posted by Mark1960 View Post
Shortly after power on there should be activity on UF7-10 and UF8-10, before it becomes steady high level. This is assuming the 6502 is running the display clear routine.

If you are able to see activity on UF7-10 and UF8-10, then use the second probe to look at the data on BD0 to BD7, and then also on SD0 to SD7.

I think you should see hex 20 the space character on both sets of data lines while UF7-10 and UF8-10 are at low level. This would be all data lines low except for SD5 and BD5 which should both be high.
ScottishColin is online now  
Old 10th Mar 2021, 10:01 pm   #886
Timbucus
Octode
 
Join Date: Mar 2019
Location: Barry, Vale of Glamorgan, Wales, UK.
Posts: 1,362
Default Re: Non-working Commodore PET 3016

Actually yes you are correct I messed up my Pin numbers on UF9 for the signals...

0 - Pin 16
0 - Pin 19
1 - Pin 12
1 - Pin 9
1 - Pin 6
0 - Pin 5
1 - Pin 2

the can also be checked on the 6316 as well obviously

But - you have probably worked that out by now...

Quote:
Originally Posted by ScottishColin View Post
Here's a page on PETSCII. the semi-colon is 3B which is 0011101 in binary

https://en.wikipedia.org/wiki/PETSCII

Colin.


Quote:
Originally Posted by Timbucus View Post
Quote:
Originally Posted by ScottishColin View Post
11-14 on UF7 and UF8 are steady high.

Colin.
Well it was worth looking - that is information in itself - if they are all 1 then the output of UF9 must be 3B constantly to generate that character - or something related. Can we confirm that the bit pattern is as we expect so I would think:

0 - Pin 16
1 - Pin 19
1 - Pin 12
1 - Pin 9
0 - Pin 6
1 - Pin 5
1 - Pin 2
Timbucus is offline  
Old 10th Mar 2021, 10:15 pm   #887
SiriusHardware
Dekatron
 
Join Date: Aug 2011
Location: Newcastle, Tyne and Wear, UK.
Posts: 11,482
Default Re: Non-working Commodore PET 3016

Quote:
I cut each pin above the motherboard and remove the carcass of the chip
Your whole approach is 100% correct, given the equipment that you have. I just wondered if you had possibly progressed to trying to get them out in one piece (for low-cost TTL that just wouldn't be worth the risk of damaging the PCB).

As Mark says, if the processor is succeeding in running ROM code one of the first things it will do is to clear the display RAM, so shortly after power-on or reset there should be a series of writes to the display RAM. From what you say there is no write activity to the RAMs during the start up phase, but in that case I would expect the content of the RAM to be random, and for the screen display to be a random jumble of characters instead of one repeated character.

Rather than destroy the RAMs at this stage, one possible approach is to replace / socket the latch UF9 and the buffers UE7 and UE8. With these made removable you will be able to scope what's coming out of the RAMs with nothing else connected to the SD0-SD7 bus. Any other ideas are welcome - basically, I'm looking for ways to prove that the RAMs are / are not OK without taking them out.
SiriusHardware is online now  
Old 10th Mar 2021, 10:16 pm   #888
ScottishColin
Octode
 
Join Date: May 2012
Location: Perth, Scotland
Posts: 1,762
Default Re: Non-working Commodore PET 3016

Here's what I have:

1 - pin 2
1 - pin 19
0 - pin 5
1 - pin 16
1 - pin 6
1 - pin 15
0 - pin 9
0 - pin 12

I can see the 111011 pattern in reverse. Is that how it's supposed to work with the address lines/PETSCII?

Colin.
ScottishColin is online now  
Old 10th Mar 2021, 10:25 pm   #889
Mark1960
Octode
 
Join Date: Mar 2020
Location: Kitchener, Ontario, Canada
Posts: 1,264
Default Re: Non-working Commodore PET 3016

Quote:
Originally Posted by ScottishColin View Post
Here's what I have:

1 - pin 2
1 - pin 19
0 - pin 5
1 - pin 16
1 - pin 6
1 - pin 15
0 - pin 9
0 - pin 12

I can see the 111011 pattern in reverse. Is that how it's supposed to work with the address lines/PETSCII?

Colin.
Yes that looks right, D0 is least significant bit and D7 is most significant bit.
Mark1960 is online now  
Old 10th Mar 2021, 10:34 pm   #890
Mark1960
Octode
 
Join Date: Mar 2020
Location: Kitchener, Ontario, Canada
Posts: 1,264
Default Re: Non-working Commodore PET 3016

I think Colin reported earlier that pins 11 to 14 on both ram chips were always high, but he has 3B on the output of the video latch. If this is confirmed then I think we can suspect the video latch is faulty and could be removed as the next step, to see if that also changes pins 11 to 14 on the ram chips.
Mark1960 is online now  
Old 10th Mar 2021, 10:37 pm   #891
ScottishColin
Octode
 
Join Date: May 2012
Location: Perth, Scotland
Posts: 1,762
Default Re: Non-working Commodore PET 3016

Thanks. I keep learning bit by bit.

No idea how to fix things, but I can do diagnostics.

Colin.

Quote:
Originally Posted by Mark1960 View Post
Quote:
Originally Posted by ScottishColin View Post
Here's what I have:

1 - pin 2
1 - pin 19
0 - pin 5
1 - pin 16
1 - pin 6
1 - pin 15
0 - pin 9
0 - pin 12

I can see the 111011 pattern in reverse. Is that how it's supposed to work with the address lines/PETSCII?

Colin.
Yes that looks right, D0 is least significant bit and D7 is most significant bit.
ScottishColin is online now  
Old 10th Mar 2021, 10:48 pm   #892
SiriusHardware
Dekatron
 
Join Date: Aug 2011
Location: Newcastle, Tyne and Wear, UK.
Posts: 11,482
Default Re: Non-working Commodore PET 3016

Quote:
I think we can suspect the video latch is faulty
I see where you are coming from in one respect, because what is being presented to the inputs of the latch is not being passed through to the output.

Removal and replacement of UF9 in a socket sounds like a cautious first step. That will also allow the possibility of manually applying different character codes to the outputs of UF9 to see if the expected characters are displayed.

However, this doesn't really explain why there are steady highs on the outputs of the RAMs / inputs to the latch, unless those highs are somehow back-driving out of the UF9 inputs.

If it's not the RAMs themselves which are supplying those steady highs, then the outputs of UE7 / UE8 are another possible source.
SiriusHardware is online now  
Old 10th Mar 2021, 11:16 pm   #893
ScottishColin
Octode
 
Join Date: May 2012
Location: Perth, Scotland
Posts: 1,762
Default Re: Non-working Commodore PET 3016

I have sockets and replacement ICs here for UE7/8 if you think it's worth replacing them.

Is your first suggestion for me to remove UF9 and resocket/replace it? if so I have already purchased this from Cricklewood in anticipation and I have the sockets here:

https://www.cricklewoodelectronics.com/74LS373.html

I can also purchase TMM314APL from eBay for £3.00 ea (if they're the right ones):

https://www.ebay.co.uk/itm/TMM314APL...IAAOSw7FBgMmcK

Colin.
ScottishColin is online now  
Old 10th Mar 2021, 11:37 pm   #894
Mark1960
Octode
 
Join Date: Mar 2020
Location: Kitchener, Ontario, Canada
Posts: 1,264
Default Re: Non-working Commodore PET 3016

TMM314APL-1 may be too slow, the 6114 is 90ns.
Mark1960 is online now  
Old 11th Mar 2021, 12:00 am   #895
SiriusHardware
Dekatron
 
Join Date: Aug 2011
Location: Newcastle, Tyne and Wear, UK.
Posts: 11,482
Default Re: Non-working Commodore PET 3016

In the first instance, since you have been so foresighted, try replacing UF9 only. Let us know if that makes any difference and we will take it from there.
SiriusHardware is online now  
Old 11th Mar 2021, 12:08 am   #896
julie_m
Dekatron
 
Join Date: May 2008
Location: Derby, UK.
Posts: 7,735
Default Re: Non-working Commodore PET 3016

That bit pattern looks like a semicolon, alright. 00111011 = &3B = decimal 59 = PETSCII (and ASCII) semicolon character. No surprise you'd get a screen full of semicolons.

With UF9 removed, I would expect the inputs it drives to float high and, if I read my PETSCII right, give a screen full of a checkerboard pattern.
__________________
If I have seen further than others, it is because I was standing on a pile of failed experiments.
julie_m is offline  
Old 11th Mar 2021, 12:22 am   #897
ajgriff
Nonode
 
ajgriff's Avatar
 
Join Date: May 2004
Location: Halifax, West Yorkshire, UK.
Posts: 2,583
Default Re: Non-working Commodore PET 3016

I don't know whether or not it helps at all but you get the checkerboard pattern (see attached) with the 2001-8 when both video RAM chips are removed. The video RAM/ROM circuit is very similar.

Alan
Attached Thumbnails
Click image for larger version

Name:	Checkerboard.jpg
Views:	53
Size:	182.2 KB
ID:	228870  
ajgriff is offline  
Old 11th Mar 2021, 12:24 am   #898
SiriusHardware
Dekatron
 
Join Date: Aug 2011
Location: Newcastle, Tyne and Wear, UK.
Posts: 11,482
Default Re: Non-working Commodore PET 3016

Quote:
if I read my PETSCII right, give a screen full of a checkerboard pattern.
As demonstrated by AJ, yes.

In addition to which, it should be possible to pull some of the lines low to fill the screen with known characters, to check that the char gen is working as expected.
SiriusHardware is online now  
Old 11th Mar 2021, 12:39 am   #899
ScottishColin
Octode
 
Join Date: May 2012
Location: Perth, Scotland
Posts: 1,762
Default Re: Non-working Commodore PET 3016

Ok. I'll get the old chip out tomorrow and the socket in ready. The replacement may not come until Monday but I'll put a post up when it does.

Thanks to you in the main and with the help of the others who have contributed, I've got characters on my screen today. I'm very happy tonight.

Thanks.

Colin.


Quote:
Originally Posted by SiriusHardware View Post
In the first instance, since you have been so foresighted, try replacing UF9 only. Let us know if that makes any difference and we will take it from there.
ScottishColin is online now  
Old 11th Mar 2021, 9:02 am   #900
SiriusHardware
Dekatron
 
Join Date: Aug 2011
Location: Newcastle, Tyne and Wear, UK.
Posts: 11,482
Default Re: Non-working Commodore PET 3016

The more PET-specific this gets the more I will have to fade into the background, that is when you will really need the help of those individuals who know the machine better. It has been good to be able to assist to get it to this point though, and it now seems inevitable that the machine will eventually work as long as you continue with the same level of dogged determination that you have so far. With the combined knowledge and effort of all the people here combined, I can't see it not working in the end.
SiriusHardware is online now  
Closed Thread

Thread Tools



All times are GMT +1. The time now is 4:30 pm.


All information and advice on this forum is subject to the WARNING AND DISCLAIMER located at https://www.vintage-radio.net/rules.html.
Failure to heed this warning may result in death or serious injury to yourself and/or others.


Powered by vBulletin®
Copyright ©2000 - 2024, vBulletin Solutions, Inc.
Copyright ©2002 - 2023, Paul Stenning.