View Full Version : SOl-20 help

November 5th, 2016, 10:48 PM
I think I'm suffering from a partial short of the 5v line to GND.
I was wondering if someone with a SOL-20 would be able to take a resistance reading between the GND and the 5v lines (maybe with the J10 connector to the PSU removed).
I've just replaced a bad 74LS04, and presently I'm getting about 95 ohms between 5V and GND which still seems very low. I don't want to plug it into the SOL PSU just yet until I'm sure the problem is solved. I've plugged the PSU in previously when trying to find the fault, but I kept blowing the 2N2222 on the PSU board. I've replaced the 2N2222 3 times now, and don't want to have to do it again until I'm sure the SOL motherboard is good.

Any help would be greatly appreciated.

Much thanks.


November 6th, 2016, 01:47 AM
Hi Phil,

Just browsing through the manual, I do not have a SOL-20 computer, just going on general experience here:

95 ohms is rather quite high assuming resistive load on 5 Volt, the power supply is designed to deliver 3 Amps max. so the resistance can be as low as 1.67 ohms... Alas digital circuits are not resistors so measuring resistance is not the best way to check. It can help in finding short circuits though, but that would be anything below 5 ohms.

Make sure Q1 is ok and has the unregulated voltage on its collector otherwise Q2 is trying to deliver all the output power through the base of Q1 and Q2 can not handle that.

You can check the powersupply with some powerful resistors in the range 1.5 to 10 ohms to check if the supply is really capable of delivering 3 amps. The circuit has a current limiter (through R1) so normally a short on the output should not result in damage of the components of the supply. Overvoltage on the output is shorted to ground by means of the 'crowbar' circuit round SCR1.

On the rest of the system: tantalum capacitors are known to fail, they usually short out or even burn...


November 6th, 2016, 07:57 AM
Well before I start looking at the board, I'd check the supply. There is a crowbar so you will need a high wattage power resistor plugged into the supply to put load against it. If I recall when I was troubleshooting one of my Sol-20 supplies, I had one of the large capacitors short out. So check them.

Now to my rule of thumb with all processor technology stuff. Change every single tantalum capacitor. In the power supply, on the board and on the personality module. They will fail at some point. I also change every single one on any PT S100 cards. Everyone knows I'm all about authenticity and making everything look factory fresh. Here is the exception to that rule.


November 6th, 2016, 12:36 PM
Wow...thanks guys for the quick replies. Really appreciate it.

Maybe I should take a step back and explain my symptoms first, and what I have done, and what I discovered just last night.

- About 2 weeks ago, I powered up the SOL, and it was dead.
- Checked the voltages, and I was getting about ~1v on the 5v line
- unplugging the PSU and tried the voltage at the connector. I was getting 5v!
- plugged it back in, and was still getting ~1v, but it then suddenly dropped to ~100mv !!
- Replacing the 2N2222 brought the voltage from the PSU (still unplugged from the motherboard) back to 5v.
- thinking that was the issue I plugged the PSU back into the board, but I was still only getting ~1v on the 5v line
- ...blew the 2N2222 again, and the voltage dropped to -100mv again.
- Replace the 2N2222.

So until last night, I was thinking the problem was the motherboard, but last night I hooked up some chunky resistors (as Gert recommended) to the PSU to apply a decent load. Again the voltage was 5v with no load and ~1v with the load.
...then the 2N2222 blew again and the voltage dropped further. Arrrrgh!

So reading all your posts above, I should do the following:
- Replace all the tantalum capacitors on the PSU
- check Q1, an possibly replace. What should I expect the voltages to be on the Base, Emitter and Collector?
...oh, and replace the 2N2222 again!

On a side note, one of the pads on the 2N2222 is just starting to lift, not surprising considering all the soldering/desoldering it's had recently. Does anyone have a recommendation for some epoxy that I can 'glue' the trace back down with?

If after reading my notes above, you have any further suggestions on what the PSU issue could be, these will be greatly received.

Thanks again for your time and support.


November 6th, 2016, 02:21 PM
Don't forget to check the large can caps for shorts in the supply. They do go bad.

November 6th, 2016, 02:25 PM
Thanks Corey, is that just a case of seeing if they have continuity across the terminals?


November 6th, 2016, 02:44 PM
Yes but disconnect them when you do

November 6th, 2016, 08:18 PM
Glueing back the pad/trace will not work, as soon as you put the soldering iron on it, the glue will burn and the pad will lift again. There are some special repair kits for this but for a single pad/trace I would not bother. Bend the pin coming through the pad slightly so it will hold the pad against the PCB, then solder it. If the pad breaks of the trace you can make a new 'pad' from one of the wire ends. Bend them into an 'question' mark shape with the straight end covering the end of the trace and the round part acting as a new pad. The solder it to the trace and component pin. If the trace is covered you need to clear the coating first (scalpel or such) revealing blank copper.

As for the 2N2222 that keeps blowing:

Since with no load the voltage is correct and assuming it is Q2 which keeps burning out, my guess is that Q1 is not working anymore. Its job is to amplify the current from Q2 and if it doesn't Q2 will fail. With respect to ground the voltage on the collector of Q1 has to be higher than 5 Volts and the same as the voltage on the large external capacitor (the voltage difference times the output current is dissipated in Q1 as heat).
If there is more than 5 Volts on the collector then it probably just failed internally and needs to be replaced.


November 7th, 2016, 10:03 PM
Thanks again guys,
I've replaced the tantalum caps, and the TIP41 (Q1). I've carefully removed the busted 2N2222 without further lifting the trace, and I've actually ordered a T0-18 3 pin socket that I'm going to install before putting in a new 2N2222.
Tonight (after I put my son to bed) I'm going to test the electrolytic caps for continuity.
So, slowly making progress. Stay tuned.


November 16th, 2016, 06:39 PM
Last of the parts arrived yesterday.
Hoping to power up the PSU this evening.
Hoping it doesn't explode!

UPDATE: No explosions...thats always a good sign.
I'm using some chunky resistors to applied a decent load on the 5v and 12v lines.
I'm getting about 4.8v. To get the voltage closer to 5v I was thinking of doing the resistor fix as mentioned here:


October 17th, 2017, 03:18 AM
I'm experiencing the identical behavior with my power supply and was happy to stumble on this thread. Bringing the PSU up slowly, I was happy to see the (unloaded) voltages within specs, but it dropped out to 1V when connected to the motherboard. The 12 and -12V lines are fine.

At first I thought the motherboard was shorted, too. But I measured 90 ohms across the 5V-GND, which is close to the manual spec of 100 ohms (Sol systems manual, p. III-11).

Looking forward to seeing what is going on with Q1 and Q2 on the regulator board.