12th Oct 2021, 9:42 pm | #61 |
Dekatron
Join Date: Aug 2011
Location: Newcastle, Tyne and Wear, UK.
Posts: 11,553
|
Re: Commodore PET 2001 voltage regulators
There is also an outside chance that the solder joints at the tapered ends of CR1 / CR2 are cracked, and are only making contact with the PCB pads when you press your meter lead against them, bending them sideways slightly.
AJ suggested earlier that you should reflow the solder joints on all the diodes but I don't think you got that far - better do that next, and while you have the board out thoroughly inspect not only the JP4 connector pin 4, the solder joint, the connection between the pad and the track running up to it - is there a crack across the track where they join? Look into the pin 4 socket in the JP8 plug connector, do the contacts look as though they are spread apart, do they look bright, or dark or burnt? Look at pin 4 on the JP8 PCB pin connector, does it look bright or dull, dark or burnt? If it looks dull, try scraping it clean. |
12th Oct 2021, 9:59 pm | #62 |
Nonode
Join Date: May 2004
Location: Halifax, West Yorkshire, UK.
Posts: 2,587
|
Re: Commodore PET 2001 voltage regulators
Excellent and well done to you both. A textbook illustration of how tricky it can be to track down intermittent faults. Have a close look at the pin 4 connector in the JP8 plug as just suggested by Sirius.
I'll probably be even less able to help with the next phase as I'm used to the earlier 2001 with static rather than dynamic RAM. However it would be useful to know the specifics of the RAM in your machine. Are they 4116s and are there two banks of eight? Alan |
12th Oct 2021, 10:40 pm | #63 |
Nonode
Join Date: May 2004
Location: Halifax, West Yorkshire, UK.
Posts: 2,587
|
Re: Commodore PET 2001 voltage regulators
Thinking out loud for a moment a very quick look at the schematic suggests that it's only the dynamic RAM chips that use the 12V and -5V supplies. Certainly all the main processing ICs use just the +5V supply. If I'm right the chips most likely to have suffered as a result of losing +5V are the dynamic RAMs. Hope that's logical.
Alan |
12th Oct 2021, 11:02 pm | #64 |
Dekatron
Join Date: Aug 2011
Location: Newcastle, Tyne and Wear, UK.
Posts: 11,553
|
Re: Commodore PET 2001 voltage regulators
Possibly - we have some RAM-independent test code which can be programmed into a 2716 EPROM and placed in the UD8 position to test the DRAMs. We'd better have a rock steady set of supply rails first, though.
|
13th Oct 2021, 11:19 am | #65 | |
Pentode
Join Date: Oct 2021
Location: Falkirk, Scotland, UK
Posts: 140
|
Re: Commodore PET 2001 voltage regulators
Quote:
|
|
13th Oct 2021, 11:22 am | #66 | |
Pentode
Join Date: Oct 2021
Location: Falkirk, Scotland, UK
Posts: 140
|
Re: Commodore PET 2001 voltage regulators
Quote:
|
|
13th Oct 2021, 11:37 am | #67 |
Octode
Join Date: Apr 2018
Location: Newbury, Berkshire, UK.
Posts: 1,287
|
Re: Commodore PET 2001 voltage regulators
Commodore in this era was a bit erratic with socketing chips, but socketed RAM was common. Sometimes it depends on when and where the board was made. Sometimes 16k would be soldered in with sockets for the extra 16k so they could manufacture boards for 2016 and 2032 (3016/3032) and decide what actual machines to send out depending on orders. It doesn't necessarily mean that someone has come along in the past and socketed them while trying to repair it.
|
13th Oct 2021, 11:48 am | #68 |
Octode
Join Date: Apr 2018
Location: Newbury, Berkshire, UK.
Posts: 1,287
|
Re: Commodore PET 2001 voltage regulators
The PD416 is a 4116 pin-compatible memory chip btw.
|
13th Oct 2021, 12:34 pm | #69 |
Nonode
Join Date: May 2004
Location: Halifax, West Yorkshire, UK.
Posts: 2,587
|
Re: Commodore PET 2001 voltage regulators
Does the label above the keyboard say 3016 Series and the label on the back 2001-16N Tim? If so it might explain the aftermarket socketing as illustrated in your earlier photo of the board's underside. In other words someone may have upgraded the RAM from 16K to 32K and fitted sockets (except for one!) at the same time. Also the unsocketed chip looks to have been rather untidily soldered in position. Out of interest what did the machine quote as the available RAM when it was working, albeit briefly? As Sothie says the 416C and 4116 are pin compatible.
By the way my favourite contender for the intermittent power supply fault is the pin 4 connector in JP8. Alan |
13th Oct 2021, 5:14 pm | #70 |
Octode
Join Date: Mar 2020
Location: Kitchener, Ontario, Canada
Posts: 1,294
|
Re: Commodore PET 2001 voltage regulators
The unsocketed ram ic is soldered from the top side of the board. Probably because the pcb was damaged during a previous repair and it was fitted like this to allow traces on both sides of the board to be soldered to the ic legs.
|
13th Oct 2021, 5:35 pm | #71 |
Nonode
Join Date: May 2004
Location: Halifax, West Yorkshire, UK.
Posts: 2,587
|
Re: Commodore PET 2001 voltage regulators
Attached is a crop from one of Tim's earlier photos showing the underside of the board. The unsocketed RAM chip is in the bottom right corner. Someone had problems!
Alan |
13th Oct 2021, 5:44 pm | #72 |
Dekatron
Join Date: Aug 2011
Location: Newcastle, Tyne and Wear, UK.
Posts: 11,553
|
Re: Commodore PET 2001 voltage regulators
We can but hope that the DRAMs are actually OK and that the problems lie elsewhere. After all, they were OK on Colin's against all expectations, in spite of the rest of that machine turning out to be a chip graveyard.
If not, someone has at least done Tim the prior favour of socketing nearly all of the DRAMs and furthermore, we know that the hardware, even with all of that rough looking soldering on the DRAM sockets and the track repairs around the not-socketed IC, did work, if only for one run. But we're running ahead again. Repair the power rails first. |
13th Oct 2021, 6:54 pm | #73 | |
Pentode
Join Date: Oct 2021
Location: Falkirk, Scotland, UK
Posts: 140
|
Re: Commodore PET 2001 voltage regulators
Quote:
Tim. |
|
13th Oct 2021, 7:04 pm | #74 | |
Pentode
Join Date: Oct 2021
Location: Falkirk, Scotland, UK
Posts: 140
|
Re: Commodore PET 2001 voltage regulators
Quote:
The JP8 socket looked okay, but I cleaned with contact cleaner again. On putting everything back together, we then had life. Good news. I decided to power off/on again, but then no screen. On checking the voltages, I found that when pressing on the top metal part of the pin (pin 4 on JP8), the PET sprang into life. I've attached a picture of this pointed to by the probe. To test this again, I unplugged the connector, then turn the PET on - nothing. So then just touched the top of the pin and the screen came on. I then took JP8 off and pressed down on the metal connectors at the top with a screwdriver; since doing this the PET has turned on successfully on each attempt. So it looks like when the connector was being pushed on, it maybe moved slightly, thus not making a connection. As it stands, I've now a PET which displays a screen every time it is turned on. I'm not sure if there's much else I can do about the connector, or if anyone has any other advice, but at least it's some headway and seems to explain what's happening. Thanks again, Tim. |
|
13th Oct 2021, 7:29 pm | #75 |
Nonode
Join Date: May 2004
Location: Halifax, West Yorkshire, UK.
Posts: 2,587
|
Re: Commodore PET 2001 voltage regulators
Looking into the connector from the other side from the wires does the metal part of the pin 4 push-on clip look the same as the others? Can you close up the clip a little with something like a sewing needle?
Alan |
13th Oct 2021, 7:38 pm | #76 |
Dekatron
Join Date: Aug 2011
Location: Newcastle, Tyne and Wear, UK.
Posts: 11,553
|
Re: Commodore PET 2001 voltage regulators
It sounds to me as though the contacts inside the lower part of the plug-in connector have spread open so they aren't a snug fit against the sides of the male pin on the motherboard. Pushing the contact assembly downwards from the top has moved it down just far enough so that the female contact assembly in the plug is touching some part of the male pin on the motherboard connector.
It also sounds to me as though, if you were to move that connector at all, the connection might go again. If we assume a worst case scenario - every time you lose the +5V, some of the chips may get damaged - we really want to make a robust and permanent repair to this +9V supply feed. Well done on proving that the JP8 / 4 connection does seem to be the problem, though. Anyone who has one of these - how feasible would it be to get that one contact in the plug-on connector out for inspection and to reform it back into shape? Are the female contacts held in place in the plug shell by spring latches which can be pressed in to release the contact? |
13th Oct 2021, 8:00 pm | #77 |
Octode
Join Date: Apr 2018
Location: Newbury, Berkshire, UK.
Posts: 1,287
|
Re: Commodore PET 2001 voltage regulators
Its probably worth mentioning that the pins might be oxidised too so a rub with a light abrasive might help. I'd also carefully inspect the plug as AJ and Sirius suggest, to see if the contacts are corroded or need bending a little. People traditionally suggest contact cleaner, but if you don't have any I'm not sure what else to suggest. Given the current in the +5v circuit I would imagine any intermittent connection would leave some trace though.
|
13th Oct 2021, 8:29 pm | #78 |
Dekatron
Join Date: Aug 2011
Location: Newcastle, Tyne and Wear, UK.
Posts: 11,553
|
Re: Commodore PET 2001 voltage regulators
To try to make things a bit clearer, one of my famous sketches - here we have a stylised drawing of the female contact (shown in dark grey) and the male pin (shown in brown).
Drawing 'A' shows the situation as it may have been originally, with the female contacts in the JP8 plug / connection #4 spread wider apart than they should be so there was no, or intermittent contact with the pin. 'B' shows the situation as it may be now - the contacts have been pushed down a bit so that they now come into contact with the pin, but maybe only just. Any movement of JP8 could make that connection disconnect again. 'C' and 'D' show the contacts more like they should be, close to the pin and touching it all the way down the sides. When like this, the relative height of the contact inside the housing does not matter so much, it will still make contact with the sides of the pin. Note that the actual contact in the actual plug may only touch the pin on one side, or on two sides (as shown) or it may even be a hollow 'box' which makes contact with the pin on all four sides. Whatever the actual situation, if the contacts in the female connector have been bent too far apart then they may need to be bent back closer inwards towards the male pin. AJ suggested looking to see if you can get a needle in there to bend them back in, I would also look to see if the contact can be unlatched / unlocked and lifted out of the plastic housing. |
13th Oct 2021, 8:42 pm | #79 |
Pentode
Join Date: Oct 2021
Location: Falkirk, Scotland, UK
Posts: 140
|
Re: Commodore PET 2001 voltage regulators
Hi all.
When I was cleaning the JP8 pins, I did clean (especially pin 4) them with a fibreglass pen, then IPA. For the JP8 plug, I do have some contact cleaner, so have put some of it in each of the holes for the pin (and also on the pins on the motherboard). I've had a look over the plug connector, and (to me) it looks okay. I've added a couple of pictures, so hope they are useful. I did try and flick the connector at the bottom with a sewing needle, but not sure if it made any difference. I've plugged/unplugged a number of times, and each time the PET has displayed the screen (I've been trying this since around 18:00 until now). However, I agree with SiriusHardware that if the connection was to go again, then a number of chips could potentially be lost. I'm happy it seems reliable at the moment, and it could be fixed; however, I'm really not sure. Just want to say thanks again for all your help (and patience!). Tim. |
13th Oct 2021, 9:01 pm | #80 |
Dekatron
Join Date: Aug 2011
Location: Newcastle, Tyne and Wear, UK.
Posts: 11,553
|
Re: Commodore PET 2001 voltage regulators
I'm afraid it's your call now Retromit: Only you, as the man on the spot, can really judge whether the connection looks clean, solid and reliable.
|