17th Mar 2021, 8:22 pm | #1101 |
Nonode
Join Date: May 2004
Location: Halifax, West Yorkshire, UK.
Posts: 2,587
|
Re: Non-working Commodore PET 3016
Nothing wrong with the approach being described and in the end it will of course be down to Colin to decide how he wants to proceed. I just think there are some advantages to an alternative approach as follows:
1. Acquire eight 4116s from the eBay seller or Refugee. 2. Remove all sixteen 4108s. 3. Fit sockets to the eight RAM positions in row I. 4. Reconfigure the jumpers. 7. Insert the eight 4116s into the sockets. 6. Switch on and say a prayer (if applicable). This would eliminate faulty system RAMs from the equation. If the PET still refuses to say READY and flash its cursor the PIAs would be my next port of call. These are easy to test assuming they're socketed. Just remove them both and if one's faulty the PET will say READY without flashing a cursor. The faulty chip can then be identified by refitting them one at a time. Alan |
17th Mar 2021, 8:31 pm | #1102 | |
Octode
Join Date: May 2012
Location: Perth, Scotland
Posts: 1,803
|
Re: Non-working Commodore PET 3016
Meter set at 20K.
Measure from +5V (red lead on +5V) to each of UI10 pins 8-9 - 2.25 UI10 pins 6-7 - 1.64 UI10 pins 4-5 - 2.33 UI10 pins 2-3 - 1.65 UI11 pins 8-9 - 2.37 UI11 pins 6-7 - 2.20 UI11 pins 4-5 - 2.33 UI11 pins 2-3 - 1.65 Did you mean black on 0V (eg Pin 8 pad on UG4) for the second set of readings? If so, then I can't get any reading at all for those pins. Scope readings coming. Colin. Quote:
|
|
17th Mar 2021, 8:40 pm | #1103 | |
Dekatron
Join Date: Aug 2011
Location: Newcastle, Tyne and Wear, UK.
Posts: 11,553
|
Re: Non-working Commodore PET 3016
Quote:
That might be a moderately trivial operation for you or I, but for Colin realistically the safest way to get the 4108s out is to destroy them, possibly without any good reason. We can't be sure at the moment whether any of the system RAM is faulty. I would like to be slightly more sure before bulldozing all of the original system RAM. However, if preservation of the original parts is of less concern and the replacement 4116s are affordable, then of course that is another possible way forward. It must be said that there is a risk that the 16 4116s, wherever they come from, may also have one or two faulty devices lurking among them. Several of us here have bought old RAMs recently only to find one or two of them faulty. It might be wise to have some independent means of testing them if this approach is taken. |
|
17th Mar 2021, 8:45 pm | #1104 |
Octode
Join Date: May 2012
Location: Perth, Scotland
Posts: 1,803
|
Re: Non-working Commodore PET 3016
|
17th Mar 2021, 8:49 pm | #1105 | |
Dekatron
Join Date: Aug 2011
Location: Newcastle, Tyne and Wear, UK.
Posts: 11,553
|
Re: Non-working Commodore PET 3016
Quote:
Obviously there are noticeable differences between three of the pin pairs and the others, but not enough to be conclusive by themselves. Let's see what you find on the scope when looking at those lines. Edit: Just caught your last, looking now. |
|
17th Mar 2021, 8:49 pm | #1106 | ||
Octode
Join Date: May 2012
Location: Perth, Scotland
Posts: 1,803
|
Re: Non-working Commodore PET 3016
My principle here, such as it is, is that I'd like to leave as many original parts in as possible. Counting right now, I've changed 11 ICs so far, but I believe they all were faulty and therefore needed changing anyway.
So I'd rather use 4108s if possible, but I recognise that may not be possible. I contacted Cricklewood and they have said they don't have any 4108s and recommended 4116s, so that may be the direction I have to take. I guess let's see what happens with UI11 and UI10 first. Colin. Quote:
|
||
17th Mar 2021, 8:53 pm | #1107 | ||
Octode
Join Date: May 2012
Location: Perth, Scotland
Posts: 1,803
|
Re: Non-working Commodore PET 3016
I tried every setting on the meter and the readings stayed stubbornly at 1
Colin. Quote:
|
||
17th Mar 2021, 9:03 pm | #1108 |
Dekatron
Join Date: Aug 2011
Location: Newcastle, Tyne and Wear, UK.
Posts: 11,553
|
Re: Non-working Commodore PET 3016
I have to be honest, nothing is really jumping out at me there.
Can you try scoping the BD0... BD7 lines, you can find them on the first of each of these pairs of connected pins. UI10 pins 11-12 UI10 pins 13-14 UI10 pins 15-16 UI10 pins 17-18 UI11 pins 11-12 UI11 pins 13-14 UI11 pins 15-16 UI11 pins 17-18 Also scope Pin 3 of each of the 4108s and pins 4 and 15 of each 4108 as well. No need to post those results, just confirm that all of them have activity on them with proper high / low logic levels. |
17th Mar 2021, 9:05 pm | #1109 |
Dekatron
Join Date: Aug 2011
Location: Newcastle, Tyne and Wear, UK.
Posts: 11,553
|
Re: Non-working Commodore PET 3016
Also check pins 1 and 19 of UI10 / UI11 for activity as well.
|
17th Mar 2021, 9:16 pm | #1110 | |
Octode
Join Date: May 2012
Location: Perth, Scotland
Posts: 1,803
|
Re: Non-working Commodore PET 3016
UI10/11 high pins attached. UI11 has some interesting readings for you - UI11 pins 11/12 kept jumping up the screen; I managed to capture one of the moments when it did so. Pins 13/14 don't look so good either.
I will go look at the 4108s now. https://drive.google.com/file/d/10m_...ew?usp=sharing Colin. Quote:
|
|
17th Mar 2021, 9:26 pm | #1111 |
Octode
Join Date: May 2012
Location: Perth, Scotland
Posts: 1,803
|
Re: Non-working Commodore PET 3016
|
17th Mar 2021, 9:27 pm | #1112 |
Dekatron
Join Date: Aug 2011
Location: Newcastle, Tyne and Wear, UK.
Posts: 11,553
|
Re: Non-working Commodore PET 3016
I see what you mean. Unfortunately those lines are attached to both the UI10 / UI11 buffers and the UE10 / UE11 buffers and a few other things besides. I'll let you do those other checks as well.
|
17th Mar 2021, 9:35 pm | #1113 |
Dekatron
Join Date: Aug 2011
Location: Newcastle, Tyne and Wear, UK.
Posts: 11,553
|
Re: Non-working Commodore PET 3016
From #1111, system RAM looks like it is being written to (pin 19 active) but not read from at all (pin 1 inactive) - or the signal which should be driving pin 1 low is missing in action, possibly. Of course we don't know exactly where the CPU is stuck or what it is trying to do.
Try monitoring UI10 / pin 1 for activity in the period immediately after reset and seeing if there is any activity on it then. If we never see any attempt to write to system RAM there may be a problem with the circuitry that generates the write signal for the system RAM. |
17th Mar 2021, 9:38 pm | #1114 | |
Octode
Join Date: May 2012
Location: Perth, Scotland
Posts: 1,803
|
Re: Non-working Commodore PET 3016
RAM chips pins 3,4 and 15 are consistent across all chips, but that doesn't look like activity to me on pins 4 and 15 unless I have my settings wrong?
https://drive.google.com/file/d/1Vxz...ew?usp=sharing Colin. Quote:
|
|
17th Mar 2021, 9:42 pm | #1115 |
Dekatron
Join Date: Aug 2011
Location: Newcastle, Tyne and Wear, UK.
Posts: 11,553
|
Re: Non-working Commodore PET 3016
Try looking at the pin 4 and 15 signals with the horizontal trace speed slowed down quite a bit. If you still get constant low and high levels on those lines we'll have to think about why that is.
|
17th Mar 2021, 9:42 pm | #1116 | |
Octode
Join Date: May 2012
Location: Perth, Scotland
Posts: 1,803
|
Re: Non-working Commodore PET 3016
This is what I get on UI10 and UI11 pin1. It doesn't stop after power on. There's no 'normal' square waves that I see.
Colin. Quote:
|
|
17th Mar 2021, 9:49 pm | #1117 |
Dekatron
Join Date: Aug 2011
Location: Newcastle, Tyne and Wear, UK.
Posts: 11,553
|
Re: Non-working Commodore PET 3016
Can you check with a meter or scope to confirm that UG7 pin 10 is held high (4-5V).
If it is can you scope UG7 pins 8 and 11 both at the same time, one above the other on your scope. Just trying to follow the system RAM read signal backwards for a while. Edit, sorry, typo: UG7 Pins 9 and 11 both at the same time. Hope I caught you in time. Last edited by SiriusHardware; 17th Mar 2021 at 9:55 pm. |
17th Mar 2021, 10:05 pm | #1118 |
Octode
Join Date: Mar 2020
Location: Kitchener, Ontario, Canada
Posts: 1,294
|
Re: Non-working Commodore PET 3016
/CAS provides the chip select function on dynamic ram, I think Colin verified he was getting RAS, CAS0 and CAS1 square waves in an earlier post.
|
17th Mar 2021, 10:07 pm | #1119 | |
Octode
Join Date: May 2012
Location: Perth, Scotland
Posts: 1,803
|
Re: Non-working Commodore PET 3016
UG7/10 settles at 5.3V after a few seconds.
Re UG7 9 and 11 - see attached. I get a signal on UG7/9 (green) but nothing on UG9/11 (yellow). Colin. Quote:
|
|
17th Mar 2021, 10:11 pm | #1120 | |
Dekatron
Join Date: Aug 2011
Location: Newcastle, Tyne and Wear, UK.
Posts: 11,553
|
Re: Non-working Commodore PET 3016
Quote:
|
|