*sadface*

[Dave]

Hi all,

I was going to list a second Gold Card, because I have a loaner SGC. I know the SGC came from a system that had some problems, but, alas, the problems were worse than I imagined, so I am going to hold off selling the second Gold Card until either the SGC or the Thor XVI is working.

When I plug the SGC into an unexpanded QL, and power up, the QL just hisses and squeals, and doesn't produce any standard display. This is also true when the 68008 is removed.

I am aware that the surface mount IC that was under the battery had some poor rework so I cleaned and remounted it and checked there were no bridges, etc. The result is the same.

I now suspect this card was the root of the problems with the system it was in originally.

So, if anyone out there has any experience with fault-tracing SGCs and can point me in the right direction, I can hopefully get this card running and free up a Gold Card for someone else (and more funding for my hardware endeavors...)

In the meantime, I will be posting a couple of other bits and bobs on SellMyRetro.com.

[RWAP]

It could be either the memory or the Altera chip (unfortunately).

The reason why you may need to check the memory is that I had a SGC with a similar issue some time ago - luckily I could try the Altera chip in another SGC and that was working.

I then thought about it - the SGC appeared to briefly allow the QL to start up and then the QL just crashed with a loud squeal...

I checked each memory chip and on removing them one by one, I found that one memory chip left a leg in the socket - there was probably just enough power getting across the broken leg to let the QL start up but then it failed miserably...

[vanpeebles]

Hehe usually is the QL users that are legless, not the QLs

[Dave]

I will check on this later this evening. Luckily the second GC is here, so I can simply swap the known good RAM onto the SGC and see if it makes a difference.

I just know this is Doug LaVerne's spare SGC which he loaned me so it gets some use. That is what freed up the second GC for sale. It makes me sad, because Doug absolutely cherishes his hardware, and although he said he knew the system had problems, it pains me to see such a precious part not working.

[Dave]

Unfortunately, it's not the memory. I removed the memory and checked the pins, tested the functionality on as compatible board from another device in pairs, all the memory testing okay.

I believe this SGC was hand reworked previously. The CPU was hand soldered and the flux had been left in place. The job was messy. I cleaned it up the best I could.

What would really help would be if someone could take a very clear photo of the 68020 (the SMD under the battery) on their SGC and post it, so I can check the solder bridges are all in the correct place, or even intended. I reproduced them verbatim.

I would also consider testing it against another SGC's components.

Finally, I am curious about the voltage regulation on this card. I haven't removed the heat sink to see what's underneath, but I note it has a 5V input. Does the design handle 5V up to unregulated input, and how does it maintain 5V output when getting 5V input from a backplane?

Aside: I always felt it was one of the great mistakes of the Sandy SuperQBoard that the 7805 and heat sink were at the right end of the board, dropping their heat well inside the case next to the CPU. I would have much preferred to place this circuitry on the left side of the board, outside the case, but I was overruled because this would conflict with the plastic covers we were using.

[RWAP]

Finally, I am curious about the voltage regulation on this card. I haven't removed the heat sink to see what's underneath, but I note it has a 5V input. Does the design handle 5V up to unregulated input, and how does it maintain 5V output when getting 5V input from a backplane?

It doesn't - if you are powering the SGC via a backplane, you have to add a wire to circumvent the onboard regulator.

[Dave]

This SGC was used unreliably in a cased, QPlane'd system. It was powered off the backplane, so was probably receiving inadequate power the whole time. There was no 5V flying lead to the card.

For those following, the manual says:

EXTERNAL POWER INPUT

If you are using your system with a standard power supply ignore this
section. If you wish to power your system from a switch mode power
supply 5V may be fed into the Super Gold Card directly via a 2.1mm
power connector located under the heatsink. The heatsink should not be
removed and the regulator should be left in place.

Wire the plug with +5V on the outside and the centre grounded. A
suitable plug is available from Miracle Systems.

....which if you ask me, is a little ambiguous.

I do have a replacement 68EC020 on hand, and can obtain NOS date range correct EP1810LC-25 if anyone happens to have the JEDEC file for it.

Also,why did they run the 020 at 24MHz when it runs natively at 25MHz?

Still looking for a photo of the 68020 in a SGC to see if these bridges are intentional or not.

[Dave]

Well well well.

I removed the heat sink and investigated underneath. I found a floating single whisker of what looked like cotton but ended up being copper wire sitting on the pins of the 7805.... I removed that, reassembled, and now it boots.

It's possible that getting intermittent unregulated 8-9v for the time it was being tested has done harm, so I'll do a soak test to see. If it passes, that will free up another Gold Card for sale

[RWAP]

Excellent news Good job it wasn't a SGC - they tend to be irreparable if plugged into a QL with the 5V regulator bypassed and no backplane.

[Dave]

It's a good job this QL was running off a custom PSU and not the factory one.

I'll post the other Gold Card on Sunday as I am working at the Circuit of the Americas today and tomorrow for the Grand Am race. It has the original box.