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.

Reply
 
Thread Tools
Old 15th Jan 2021, 1:52 pm   #121
ScottishColin
Hexode
 
Join Date: May 2012
Location: Perth, Scotland
Posts: 294
Default Re: Non-working Commodore PET 3016

And here is the underside - I think I have lifted the track to pin 4, but I think I may be able to fix that when I solder the socket in; is that right?

And I also think I have completely removed the pad on the underside for pin 8 - if that connects to 0V on the top of the board, might I get away with that?

Thanks.
Attached Thumbnails
Click image for larger version

Name:	IMG_20210115_134659_1.jpg
Views:	50
Size:	104.0 KB
ID:	224666  
ScottishColin is offline   Reply With Quote
Old 15th Jan 2021, 2:39 pm   #122
SiriusHardware
Dekatron
 
Join Date: Aug 2011
Location: Newcastle, Tyne and Wear, UK.
Posts: 5,790
Default Re: Non-working Commodore PET 3016

If that's the first time you have ever tried to do that then you did pretty well. At work just now, more detail later.
SiriusHardware is online now   Reply With Quote
Old 15th Jan 2021, 6:02 pm   #123
Mark1960
Hexode
 
Join Date: Mar 2020
Location: Kitchener, Ontario, Canada
Posts: 285
Default Re: Non-working Commodore PET 3016

Check the through hole plating for continuity for any pins that have a trace connected on both sides of the board, if there are any.

Do you have any copper tape, that would be useful to fit a strip over the damaged ground trace on top side, with holes for the socket. If not then maybe reinforce the ground with a wire link.

Did you get a turned pin socket for the replacement? Then you can solder top and bottom side of the socket, but only where there is a top side trace.
Mark1960 is online now   Reply With Quote
Old 15th Jan 2021, 6:20 pm   #124
SiriusHardware
Dekatron
 
Join Date: Aug 2011
Location: Newcastle, Tyne and Wear, UK.
Posts: 5,790
Default Re: Non-working Commodore PET 3016

Just taking another look.

With regard to the pin 8 pad, although you've taken the bottom side pad off, the top side pad is still there and hopefully the conductive sleeve going through the PCB between the top side and bottom side is still there.

For pin 4, luckily that top side pad doesn't go to anything so you only need to worry about restoring the connection between the underside track / pad going to pin 4 and pin 4 itself.

Fit your IC socket (right way around) - I would suggest soldering pins 1 and 9 first, then looking to see if the socket is fitted perfectly flat against the board. When you are happy that it is, solder all the other undamaged pads to their pins.

Then, solder pin 4 of the socket to what's left of the lower side pin 4 pad.

For pin 8, although there is no pad to solder to, if you apply your iron and some solder to pin 8 of the IC socket it should run down the pin, down inside the sleeve lining the hole and it should connect the pin to the top side pad that way.

When you've done that, fit your new IC in the socket.

We'll ask you to make a couple of measurements afterwards to make sure these connections are intact.

If you decide to fit the IC directly, then for pin 8 of the IC you can solder that neatly to the top side pad although you may have to clean the top side pad first. I would suggest fitting a socket if you did get one because you don't ever want to have to desolder anything from this IC position again.

Mark: If pin 8 doesn't connect due to a damaged through-hole then I think a link from pin 8 - pin 9 on the underside is the simplest way to fix that, as pin 9 is clearly connected to 0V.

Last edited by SiriusHardware; 15th Jan 2021 at 6:41 pm.
SiriusHardware is online now   Reply With Quote
Old 15th Jan 2021, 9:44 pm   #125
SiriusHardware
Dekatron
 
Join Date: Aug 2011
Location: Newcastle, Tyne and Wear, UK.
Posts: 5,790
Default Re: Non-working Commodore PET 3016

Once you've fitted the IC socket and put the IC in, working on the top side and measuring on the actual IC pins, check you have continuity (zero ohms resistance) between the following pairs of pins:-

UG5 pin 8 and UG5 pin 9
UG5 pin 4 and UG5 pin 1

Let us know how that goes. Don't power the machine up until you are sure those connections are present.
SiriusHardware is online now   Reply With Quote
Old 16th Jan 2021, 12:44 pm   #126
ScottishColin
Hexode
 
Join Date: May 2012
Location: Perth, Scotland
Posts: 294
Default Re: Non-working Commodore PET 3016

Hi gents - I'll start today - I have bought a standard socket and a turned socket (mainly because I didn't know the difference) - is there one that is better than the other for this job?

Thanks.
ScottishColin is offline   Reply With Quote
Old 16th Jan 2021, 1:30 pm   #127
SiriusHardware
Dekatron
 
Join Date: Aug 2011
Location: Newcastle, Tyne and Wear, UK.
Posts: 5,790
Default Re: Non-working Commodore PET 3016

Turned pin sockets are generally held to be better quality. They only have one drawback, and that is that they are very difficult to desolder from a double sided PCB once fitted. The reason is that the round shape of the upper part of the pin forms a perfect seal over the top end of the hole, making it impossible for a solder sucker to draw air+solder through the hole from the other side.

As Mark pointed out it is slightly easier to solder the pins on a turned-pin socket to its associated pads on the top side of the PCB should that be required - a 'normal' socket hides its pins under the plastic body of the socket.

I would say fit the turned-pin socket because we can effect any repairs which may prove necessary by using short wire links on the underside of the PCB. Remember to get the notched end of the socket at the right (pin 1) end - electrically it would make no difference, but if you fit the socket the wrong way around then sooner or later that socket will mislead you or someone else into fitting an IC the wrong way around.

The other thing about turned-pin sockets is that you have to get all the IC pins very precisely lined up before they will go into all 16 holes. Don't rush it or you will end up folding one or more of the pins over, and you can only straighten them maybe once or twice before they snap off, like the one on your PROM.
SiriusHardware is online now   Reply With Quote
Old 16th Jan 2021, 4:01 pm   #128
ScottishColin
Hexode
 
Join Date: May 2012
Location: Perth, Scotland
Posts: 294
Default Re: Non-working Commodore PET 3016

Done - confirmed:

UG5 pin 8 and UG5 pin 9 continuity with 0 resistance
UG5 pin 4 and UG5 pin 1 continuity with 0 resistance
ScottishColin is offline   Reply With Quote
Old 16th Jan 2021, 4:23 pm   #129
SiriusHardware
Dekatron
 
Join Date: Aug 2011
Location: Newcastle, Tyne and Wear, UK.
Posts: 5,790
Default Re: Non-working Commodore PET 3016

Right, let's find out if that has made a difference - hopefully it has. If it still doesn't work don't be too disheartened, it's not unusual for something like this to have a string of faults which have to be tracked down one after the other.

Power the machine up and see if it comes on - bear in mind that the monitor is a CRT type so it will probably take a few seconds to come on even if the machine is now working.

If it doesn't come up, then we'll get back on the original track - do you now have 1Mhz on UG5 pin 7?
SiriusHardware is online now   Reply With Quote
Old 16th Jan 2021, 6:25 pm   #130
ScottishColin
Hexode
 
Join Date: May 2012
Location: Perth, Scotland
Posts: 294
Default Re: Non-working Commodore PET 3016

So I guess I have proved I can solder and reconnect things, but no 1Mhz signal.

I do have 8 Mhz on pin 3, 4Mhz on pin 2 and 2Mhz on pin 6.

I'm in this for the long run so I'm not disappointed. Learning all the time.

Colin.
ScottishColin is offline   Reply With Quote
Old 16th Jan 2021, 8:25 pm   #131
SiriusHardware
Dekatron
 
Join Date: Aug 2011
Location: Newcastle, Tyne and Wear, UK.
Posts: 5,790
Default Re: Non-working Commodore PET 3016

Sorry to have to ask again, but can you (with power off) check between UG5 pins 7 and pin 8 for a short, and between UG5 pins 7 and 16 for a short?

If no short on either of those, remove UG5 from its socket, GENTLY bend pin 7 out to an angle of about 45-60 degrees so that it is slanting out sideways and then put the IC back in its socket with every pin fitted in its hole except for pin 7, which should be hanging out over the side of the socket.

Power on again, now do you see 1MHz on UG5 pin 7?
SiriusHardware is online now   Reply With Quote
Old 16th Jan 2021, 10:33 pm   #132
ScottishColin
Hexode
 
Join Date: May 2012
Location: Perth, Scotland
Posts: 294
Default Re: Non-working Commodore PET 3016

OK - no shorts between UG5 (7) and UG5 (8), nor UG5 (7) and UG5 (16).

Awesome idea re pin 7; I hadn't thought of that. However, makes no change - no reading. pins 2,3 and 6 give me the same readings as in post #130.

Something's stopping pin 7 doing its stuff.....
ScottishColin is offline   Reply With Quote
Old 16th Jan 2021, 11:12 pm   #133
SiriusHardware
Dekatron
 
Join Date: Aug 2011
Location: Newcastle, Tyne and Wear, UK.
Posts: 5,790
Default Re: Non-working Commodore PET 3016

Did you get more than one IC?

There is a small chance that the new IC's QD output has been damaged by some as yet undiscovered problem on the CLK1 line. If you have a spare, try fitting it but with pin 7 sidelined right from the start so that the output can't be damaged, and see if that IC has 1MHz on its pin 7. If you don't have a spare we'll proceed for now on the assumption that your replacement IC is still OK and the cause is elsewhere.

Where, though? At the moment I don't know, I will have to browse the diagrams for a while. Any thoughts, anyone? When last checked the 'load' input (pin 11) on UG5 appeared to have a steady high (5V) on it, high being the 'inactive' state for the load input. If it was taken low momentarily, it would reset the counter to zero.

Colin measured for frequency on that pin as well, so if the counter is being periodically reset that should have showed up as a frequency with a high duty cycle,

One thing you can do, Colin, is to remove UG5 and (just working on the chip, and with the black lead on pin 8) measure the resistances from pins 2, 3, 6 and 7 to pin 8. Then with the red lead on pin 16, measure the resistances from pins 2, 3, 6 and 7 to pin 16.

As all of these pins of the device have the same function, I would expect the resistances from those pins to pin 8 to all be about the same value as each other. The resistance from the pins to pin 16 will be a different value but again, it should be about the same for all four pins. If one pin looks significantly different to the other three, that would suggest that pin was faulty. By the way, be careful not to measure yourself when taking these readings. If you have an experimenter's 'breadboard' (a white slab with holes to plug components into) plug the IC into that to stop it from skating around while you are trying to take measurements.

There is a way we can find out whether the chip is or is not counting all the way up to 15 (1111) but let's see what you get from the above first.
SiriusHardware is online now   Reply With Quote
Old 16th Jan 2021, 11:13 pm   #134
ScottishColin
Hexode
 
Join Date: May 2012
Location: Perth, Scotland
Posts: 294
Default Re: Non-working Commodore PET 3016

I bought two chips and tried them both - the second one has not been in while the PET was powered on.

I'll check resistances tomorrow.

thanks.
ScottishColin is offline   Reply With Quote
Old 16th Jan 2021, 11:45 pm   #135
SiriusHardware
Dekatron
 
Join Date: Aug 2011
Location: Newcastle, Tyne and Wear, UK.
Posts: 5,790
Default Re: Non-working Commodore PET 3016

After you've done those checks, a couple of other things to try with the IC reinserted normally and the machine powered up, please:

Try to measure frequency on:-
UG5 Pin 14 (should be 16MHz, but your meter may not manage it)
UG5 Pin 12 (Should be 16MHz with a very low duty cycle figure - the meter may not manage it)
UH3 pin 3 (should be 1MHz)
SiriusHardware is online now   Reply With Quote
Old 17th Jan 2021, 9:22 am   #136
SiriusHardware
Dekatron
 
Join Date: Aug 2011
Location: Newcastle, Tyne and Wear, UK.
Posts: 5,790
Default Re: Non-working Commodore PET 3016

And I have another question about your meter. We know it is autoranging but most autoranging meters allow you to manually select a range to override the meter's automatic choice, typically, there is a dedicated 'range' button but your meter does not appear to have one. Could you have a look at your meter's manual to see if there is a way to 'force' the meter to use a particular range such as Hz, KHz, MHz when in frequency measuring mode?
SiriusHardware is online now   Reply With Quote
Old 17th Jan 2021, 4:39 pm   #137
ScottishColin
Hexode
 
Join Date: May 2012
Location: Perth, Scotland
Posts: 294
Default Re: Non-working Commodore PET 3016

Hi. re post #133

Measured Pin 11 UG5 an have a steady 5.3V when powered on
Measured (black) pin 8 UG5-> pins 2,3,6 & 7 UG5; 0 resistance when powered off
Measured (red) pin 16 UG5 -> pins 2,3,6 & 7 and have 2.19 MΩ on 2 & 3 and 2.4 MΩon 6 & 7 when powered off

re post #135
UG5 pin 14 - I get 16Mhz at 49.8%
UG5 pin 12 - zero reading
UH3 pin 3 = 49.97 @ 49.8% which I assume is background mains wiring?

re post #136
There seems no way to force the range of reading on this meter - it seems to auto-adjust and there are no buttons I can find nor is there anything in the (pretty sparse) manual.
ScottishColin is offline   Reply With Quote
Old 17th Jan 2021, 5:32 pm   #138
SiriusHardware
Dekatron
 
Join Date: Aug 2011
Location: Newcastle, Tyne and Wear, UK.
Posts: 5,790
Default Re: Non-working Commodore PET 3016

Quote:
Measured (black) pin 8 UG5-> pins 2,3,6 & 7 UG5;
0 resistance when powered off
Measured (red) pin 16 UG5 -> pins 2,3,6 & 7 and have
2.19 MΩ on 2 & 3 and 2.4 MΩon 6 & 7 when powered off
These measurements were really meant to be made on the IC with the IC removed from circuit, the reason being that if you do them with the IC in-circuit you are not only measuring the IC, you are measuring whatever the IC is connected to as well. However the fact that they all read about the same anyway suggests that none of those four pins is significantly different from the others. (which is good).

For the first set whch I have quoted above, what is the meter really saying when you make those measurements because '0 resistance' is ambiguous - do you mean that the resistance is infinite (no resistance reading)? Usually if the resistance is too great for the meter to measure it will show a '1' in the left digit, the way it does when you have the meter in resistance mode and the probes not connected to anything.

Did you try your second (originally never powered) IC with pin 7 swung out to see if that can produce 1MHz on pin 7?

I'm actually beginning to wonder if your meter has some kind of 'blind spot' for 1MHz, maybe it can't decide whether to display it as 1000KHz or 1.000MHz (This is why I was hoping you could manually select the frequency range). To check this, put everything back to normal, power up, and with black lead on UG5 pin 8, measure the DC voltage on pin 7 of UG5. What do you get... 0V? or 5V? or 2.5V?
SiriusHardware is online now   Reply With Quote
Old 17th Jan 2021, 5:57 pm   #139
ScottishColin
Hexode
 
Join Date: May 2012
Location: Perth, Scotland
Posts: 294
Default Re: Non-working Commodore PET 3016

Apologies I didn't make it clear. These readings were all with the chip out and plugged into a breadboard I have:

"Measured (black) pin 8 UG5-> pins 2,3,6 & 7 UG5;
0 resistance when powered off
Measured (red) pin 16 UG5 -> pins 2,3,6 & 7 and have
2.19 MΩ on 2 & 3 and 2.4 MΩon 6 & 7 when powered off"

When I said 0 resistance, the meter showed no reading and stayed at '0'.

I've tried both chips with pin 7 out of the way and get the same result.

I will try to source another meter and see what readings I get and come back to you.

Thanks.
ScottishColin is offline   Reply With Quote
Old 17th Jan 2021, 6:18 pm   #140
ScottishColin
Hexode
 
Join Date: May 2012
Location: Perth, Scotland
Posts: 294
Default Re: Non-working Commodore PET 3016

Does this look like it might do the job?

https://www.amazon.co.uk/Almabner-1H...1&rps=1&sr=8-1
ScottishColin is offline   Reply With Quote
Reply

Thread Tools



All times are GMT. The time now is 9:15 am.


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 - 2021, vBulletin Solutions, Inc.
Copyright ©2002 - 2021, Paul Stenning.