3rd Mar 2021, 7:29 pm | #781 | |
Dekatron
Join Date: Aug 2011
Location: Newcastle, Tyne and Wear, UK.
Posts: 11,482
|
Re: Non-working Commodore PET 3016
Quote:
In the absence of any obvious chip output driving the INIT line maybe it is just supposed to hold all the connected ICs off until the +5V supply is up? The main thing is that it should normally be high, ~ 4-5V, held in that state by R12. As per Mark: Check for +5V on UH3 pin 9 Check for 1MHz on UH3 pin 1 - if not there, look for it on UG5 pin 7. Check for 16MHz on UH3 pin 8 UH3 pins 3,4,5,6, 11,12,13 should all have 1MHz waveforms on as well. |
|
3rd Mar 2021, 9:07 pm | #782 | ||
Octode
Join Date: May 2012
Location: Perth, Scotland
Posts: 1,762
|
Re: Non-working Commodore PET 3016
UH3/9 - 5.3V
UH3/1 - 1Mhz confirmed - see attached UG5/7 - ditto UH3/8 - 16Mhz confirmed by the scope, but as a sine wave not a square wave https://drive.google.com/file/d/1efE...ew?usp=sharing However, I can get no readings on UH3 pins 3,4,5,6, 11,12,13. Colin. Quote:
|
||
3rd Mar 2021, 9:53 pm | #783 |
Dekatron
Join Date: Aug 2011
Location: Newcastle, Tyne and Wear, UK.
Posts: 11,482
|
Re: Non-working Commodore PET 3016
Odd. Could you check that IC UH3 is powered (meter red on pin 14, meter black on pin 7, measure voltage) and also measure the voltage on pin 2 of UH3? (Should be 5V).
I think the 16MHz 'sine wave' is probably just because that frequency is getting close to the upper frequency limit that your scope can display. We know it must be a good 16MHz signal because that signal also drives UG5 which divides it down to produce 1MHz out on pin 7. |
3rd Mar 2021, 10:09 pm | #784 |
Octode
Join Date: May 2012
Location: Perth, Scotland
Posts: 1,762
|
Re: Non-working Commodore PET 3016
UH3 - 5.04V.
UH3/2 - 5.43V Colin. |
3rd Mar 2021, 10:20 pm | #785 |
Dekatron
Join Date: Aug 2011
Location: Newcastle, Tyne and Wear, UK.
Posts: 11,482
|
Re: Non-working Commodore PET 3016
This seems to be suggesting a problem with UH3 itself. The output pins 3,4,5,6, 11,12,13 - if they don't have signals (waveforms) on them what do they have on them? 0V or 5V?
We've assumed that UH3 pin 9 is held steady-high and your voltage measurement seems to bear that out but could you look at it with a scope as well to see if there is a waveform on it, or whether it is just steady +5V? Try looking at the 16MHz signal on UI1 pin 4. Does that look like a sine wave, or a square wave? |
3rd Mar 2021, 11:38 pm | #786 |
Octode
Join Date: May 2012
Location: Perth, Scotland
Posts: 1,762
|
Re: Non-working Commodore PET 3016
The output pins 3,4,5,6, 11,12,13 - they all have a very low voltage on them - c. 0.1V but it's there.
The scope gives me no waveform on UH3/9 - rock steady 5.02V. UI1/4 is a steady 16Mhz frequency with a sine wave. Colin. |
3rd Mar 2021, 11:44 pm | #787 |
Dekatron
Join Date: Aug 2011
Location: Newcastle, Tyne and Wear, UK.
Posts: 11,482
|
Re: Non-working Commodore PET 3016
OK, it was possible that the 16MHz output from UG1 (6) was genuinely weak and, although able to drive UG5, not able to drive UH3.
If it looks the same back on UI1 (4) then it is probably just your scope not quite able to show the square shape of the 16MHz signal. Unfortunately this throws it back in the direction of UH3 being faulty, unless anyone else has any ideas? If UH3 was in a socket / removable you could write a little sketch to test it using your Arduino, but I imagine it is soldered. |
3rd Mar 2021, 11:59 pm | #788 | |
Octode
Join Date: May 2012
Location: Perth, Scotland
Posts: 1,762
|
Re: Non-working Commodore PET 3016
Yep - soldered, but I'm getting happier each time to change them out if that looks like the right thing to do.
Colin. Quote:
|
|
4th Mar 2021, 12:01 am | #789 |
Dekatron
Join Date: Aug 2011
Location: Newcastle, Tyne and Wear, UK.
Posts: 11,482
|
Re: Non-working Commodore PET 3016
It may come to that, let's see what the others think first.
|
4th Mar 2021, 12:11 am | #790 |
Octode
Join Date: May 2012
Location: Perth, Scotland
Posts: 1,762
|
Re: Non-working Commodore PET 3016
OK - which one?
1) https://www.cricklewoodelectronics.com/74LS164.html 2) https://www.cricklewoodelectronics.com/74164.html Colin. |
4th Mar 2021, 12:23 am | #791 |
Dekatron
Join Date: Aug 2011
Location: Newcastle, Tyne and Wear, UK.
Posts: 11,482
|
Re: Non-working Commodore PET 3016
If you are just going to order one anyway, the 'plain' (non-LS) version is the one you want. Naturally, that's the older, more exotic and therefore slightly more expensive one.
I may have missed something though, so maybe wait for a second opinion? |
4th Mar 2021, 12:37 am | #792 |
Octode
Join Date: Mar 2019
Location: Barry, Vale of Glamorgan, Wales, UK.
Posts: 1,362
|
Re: Non-working Commodore PET 3016
I agree buy the 74164 (using an LS may cause issues if they rely on the switching speed of the plain TTL) - it is unlikely that all circuits downstream of the shift register have failed and if there is nothing on any of the outputs then there are no clock phases being generated to UH2, UG1 or UH4.
This is the source of RAS and CAS as well as part of the video so the 4116 Dynamic RAM memory will not work... This will seriously hinder the CPU... |
4th Mar 2021, 12:56 am | #793 |
Dekatron
Join Date: Aug 2011
Location: Newcastle, Tyne and Wear, UK.
Posts: 11,482
|
Re: Non-working Commodore PET 3016
What do you think the likely UH3 chip failure mode is though - failure of the clock input or the serial data input, perhaps? Certainly not failure of a single output.
This machine has definitely had a hard life. If only it could tell us its story. |
4th Mar 2021, 1:03 am | #794 | |
Octode
Join Date: Mar 2019
Location: Barry, Vale of Glamorgan, Wales, UK.
Posts: 1,362
|
Re: Non-working Commodore PET 3016
Quote:
I think it is likely internal either the AND gate (1/2) or the inverter on the clock input (8). As long as Pin 9 is held high then that has to be the cause. |
|
4th Mar 2021, 1:07 am | #795 |
Octode
Join Date: Mar 2019
Location: Barry, Vale of Glamorgan, Wales, UK.
Posts: 1,362
|
Re: Non-working Commodore PET 3016
My comment about downstream circuits was aimed at showing that it is unlikely with one of the downstream inputs failed causing a drive problem you would still expect the other phases to work although I suppose if pin (EDIT Pin 3) to UH4 is shorted then not sure if that would not halt the chain, we need Slothie's encyclopedic knowledge of TTL output circuits to see if that could cause a failure of the chain.
|
4th Mar 2021, 1:20 am | #796 |
Dekatron
Join Date: Aug 2011
Location: Newcastle, Tyne and Wear, UK.
Posts: 11,482
|
Re: Non-working Commodore PET 3016
OK, well, it seems we have at least two in favour of replacing UH3 (remember to get a 14-pin socket if you don't have one already).
|
4th Mar 2021, 2:22 am | #797 |
Octode
Join Date: Mar 2020
Location: Kitchener, Ontario, Canada
Posts: 1,264
|
Re: Non-working Commodore PET 3016
Was this a typo or is UH3-2 really 0.4 volts higher than the supply?
Max input voltage for ttl is 5.5 v, so 5.43 is getting a bit close to the limit. I seem to remember that design recommendations for ttl were to not connect inputs directly to 5v, so this seems to be a bad design from commodore and might explain why the chip failed. If we are voting then my vote is to replace UH3. |
4th Mar 2021, 11:44 am | #798 | |
Octode
Join Date: May 2012
Location: Perth, Scotland
Posts: 1,762
|
Re: Non-working Commodore PET 3016
It wasn't a typo - I'll measure again when replaced to see what I can see. Old chip coming out today and socket to be soldered in.
New chip on the way from Cricklewood so hopefully tomorrow or Saturday. For anyone else looking, Cricklewood's service has been awesome, even on the smallest of orders. i am impressed. Colin. Quote:
|
|
4th Mar 2021, 12:53 pm | #799 |
Dekatron
Join Date: Aug 2011
Location: Newcastle, Tyne and Wear, UK.
Posts: 11,482
|
Re: Non-working Commodore PET 3016
After you've put the socket in, power on briefly and measure the voltages on UH3 socket pins 14 and 2. It should be impossible for the voltage on pin 2 to be higher than the +5V supply voltage, that is why we are puzzled.
|
5th Mar 2021, 1:52 pm | #800 |
Octode
Join Date: May 2012
Location: Perth, Scotland
Posts: 1,762
|
Re: Non-working Commodore PET 3016
Checked the voltage again - both sitting at 5.04V. I'll leave it on for longer and try later on to see - I wrote that voltage down so I didn't make that one up (unlike some of my other responses).
In better news, the new chip/socket is in and I now get nice 1Mhz waves at pins 3,4,5,6, 11,12,13 on UH3. Still nothing on the actual screen though. Colin. |