Did you measure the voltage, are you sure that the CPU is getting 3.3 / 3.45V ? Maybe the MOSFET is dead.

Would the MOSFET be the voltage regulator looking thing that's screwed to a heatsink, next to the CPU? It has TIP 32C written on it. I thought it was a voltage regulator but apparently its a transistor, which is something I'm not super familiar with it.
That being said, I did test the voltage by touching the pins on the bottom of the motherboard. I included a photo with the pins numbered so my tests make sense.

Here's what I found:

With the CPU voltage set to 3.3V:

First I tested all three pins with the black lead of the multimeter connected to the PSU ground and the red lead touching each pin of the transistor. Pin 1 read 1.5V, pin 2 read 2V and pin 3 read 5V.

I then tested by touching both black and red leads to different pins on the transistor. Here are the results:

Black lead to Pin 1, red lead to pin 2= 0.43V
Black to 1, red to 3= 3.4V
Black to 2, red to 3= 3V.

With the CPU voltage set to 3.45V:

With the black lead connected to PSU ground, Pin 1 read 1.3 V, pin 2 was 2V and pin 3 was 5 V.

When touching both leads to different pins I got:

Black to pin 1, red to pin= 0.62V
Black to 1, red to 3= 3.69V
Black to 2, red to 3= 3V.

I don't know what this test shows, or if this is even the right test to do. I'm kinda learning as we go with this motherboard. Does this mean anything? Is my transistor bad? Should I be looking at something else?

Thank you

That looks suspect, like somebody messed with it.

fenderstrat wrote on 2021-05-30, 01:05:

First I tested all three pins with the black lead of the multimeter connected to the PSU ground and the red lead touching each pin of the transistor. Pin 1 read 1.5V, pin 2 read 2V and pin 3 read 5V.

As BitWrangler said it looks like someone messed with it (I see solder splatter on the resistors). That HS bolt does not look stock. Maybe it was just replaced.
You need to re-check from the Lug on that VReg (is the output) to ground with a 3.3v CPU installed and with jumpers set at 3.3v. If around 3.3v then it is working.
the Vreg chip is just a transistor... maybe the circuit to it is bad. Can you take a better picture of the cpu socket+ Vreg area ?
Edit: fixed a boo-boo

Horun wrote on 2021-05-30, 02:57:

As BitWrangler said it looks like someone messed with it (I see solder splatter on the resistors). That HS bolt does not look s […]
Show full quote

fenderstrat wrote on 2021-05-30, 01:05:

First I tested all three pins with the black lead of the multimeter connected to the PSU ground and the red lead touching each pin of the transistor. Pin 1 read 1.5V, pin 2 read 2V and pin 3 read 5V.

As BitWrangler said it looks like someone messed with it (I see solder splatter on the resistors). That HS bolt does not look stock. Maybe it was just replaced.
You need to re-check from the Lug on that VReg (is the output) to ground with a 3.3v CPU installed and with jumpers set at 3.3v. If around 3.3v then it is working.
the Vreg chip is just a transistor... maybe the circuit to it is bad. Can you take a better picture of the cpu socket+ Vreg area ?
Edit: fixed a boo-boo

Ok, I just need some clarification. Is the Lug one of the transistor's pins? If so, which one? Looking at a datasheet I see that pin 1 is base, pin 2 is collector and pin 3 is emitter. Which one do you mean? Which one would be output? Or is the Lug something else?

As for ground, any ground? I've been using the middle (black) pins on the power connector, is that good?

From what I can see in photos of other 4DPS's, that heatsink bolt seems to be common. In any case, here's a closeup of that area

Instead of checking the voltage on the MOSFET pins, even better would be to measure directly the voltage received by the CPU.
Remove the CPU, insert a paper clip in one of the VCC pins, push the socket retention lever, power on the motherboard.

I suspect that this motherboard was dead and someone else tried to fix it, was unable to, and... probably sold it to you as "untested"? 😀

fenderstrat wrote on 2021-05-30, 01:05:

With the CPU voltage set to 3.3V: […]
Show full quote

With the CPU voltage set to 3.3V:

Pin 1 read 1.5V, pin 2 read 2V and pin 3 read 5V.

I then tested by touching both black and red leads to different pins on the transistor. Here are the results:

Black lead to Pin 1, red lead to pin 2= 0.43V
Black to 1, red to 3= 3.4V
Black to 2, red to 3= 3V.

They look odd. I looked up the TIP32C and it's a PNP transistor. That means that if the Base (pin 1) is slightly below the Emitter (pin 3), so current flows from Emitter to Base, then current can also flow from the Emitter to the Collector (pin 2). So Vin is pin 3, Vout is pin 2 and switch control is pin 1. But the data sheet says the maximum Vbe should be 1.8V (and ought to be more like 1.2V), when you've measured 3.4V. At that voltage (3.4V) there should be a huge current flowing from the Emitter to the Base.

I don't know why the Base voltage would be like that, could be a faulty component. As well as the transistor there should be a biasing resistor connected to the Base that controls the amount of current that can flow (probably one of those two resistors right by the transistor), so you may want to check that the resistance across those matches the colour codes.

we have this on UH19:
http://www.win3x.org/uh19/motherboard/show/5805

bloodem wrote on 2021-05-30, 07:42:

Instead of checking the voltage on the MOSFET pins, even better would be to measure directly the voltage received by the CPU.
Remove the CPU, insert a paper clip in one of the VCC pins, push the socket retention lever, power on the motherboard.

I suspect that this motherboard was dead and someone else tried to fix it, was unable to, and... probably sold it to you as "untested"? 😀

Actually, it was sold to me as "tested and working", so I got my money back. Which is even more incentive to get it working!

Ok, so I tried this method (a paper clip was way too thick, I ended up clipping the leg off a capacitor and using that instead) and got absolutely nothing, tried different VCC holes, no voltage.

Next thing I did was taking the TIP 32C, to test it off the circuit. My wife used her desoldering gun (she's way better at soldering/desoldering than I) and gently pulled it out. As soon as she did, one of the pins broke off. So I'm pretty sure that regardless of anything else, this transistor had to be replaced. I tested it with the multimeter in diode mode (there was just enough of the broken pin left to do this) and I got absolutely nothing, no reading from any pins. Already ordered a replacement, hopefully it should be here soon.

I'm new to testing resistors, but i tested them all anyway (in circuit, not sure if that makes a difference). Most of them were perfect, but some of them did read out of range. One of them (circled in red on the photo) was reading about 12 M ohms when according to the stripe codes, it should be 25 M ohms (which apparently is a pretty uncommon value, so I'm not sure if I read that one right). The yellow arrow shows what I think is a resistor array. That one is also reading bad, two of the pins were showing a much lower value than the rest. So I guess I'll be getting some of those as well.

So I'll just have to wait until the new parts arrive for now. Thanks everyone for all your tips, I'll keep you posted!

A resistor can read lower when tested in circuit (depending on the actual circuit), but it should never measure higher. So that resistor is probably fine.
What about if you configure the motherboard for 5V, do you get any reading on the VCC pins?

fenderstrat wrote on 2021-05-30, 12:21:

i tested them all anyway (in circuit, not sure if that makes a difference).

It can do, if there are two resistors in parallel, and there'll be some non-0 resistance from power to ground through the rest of the board. But it can be handy for just checking if a resistor has gone completely open and if biasing resistors look about right.

One of them (circled in red on the photo) was reading about 12 M ohms when according to the stripe codes, it should be 25 M ohms (which apparently is a pretty uncommon value, so I'm not sure if I read that one right).

Looks to be something to do with what I think is the 32.768KHz oscillator input to the 85C497 on pin 34&35 (bias resistor across the crystal inputs to get the oscillator started?). I can't make out the colour bands in your photo, but the picture on UH19 looks like it's brown-green-blue, so 15M in that case. I suspect the absolute value isn't critical here, as long as it's >1M but not actually open circuit.

The yellow arrow shows what I think is a resistor array. That one is also reading bad, two of the pins were showing a much lower value than the rest. So I guess I'll be getting some of those as well.

Maybe not, could just be to do with what jumpers are fitted. Either way, you've found a definite faulty part with that transistor, so I'd probably just get that changed first. When you replace it, try without any CPU or RAM fitted, just to make sure that the voltages look right first.

Back when RadioShack still had a few components they only had about 10 transistor types in store, so if you wondered in saying "I need a transistor like this" you'd end up with one of those 10... and I seem to recall the TIP32 being one of them... so I am wondering if that is actually the right part number for that board. What it should be doing is acting as current amplifier to a reference voltage, that's set at 3.3 or is adjustable by jumpers switching in different resistor values for 3.6 etc. Possibly finding the actual reference regulator and looking up the application note might give us more clue what's happening and whether this transistor is right for the job.

BitWrangler wrote on 2021-05-30, 13:36:

Back when RadioShack still had a few components they only had about 10 transistor types in store, so if you wondered in saying "I need a transistor like this" you'd end up with one of those 10... and I seem to recall the TIP32 being one of them... so I am wondering if that is actually the right part number for that board. What it should be doing is acting as current amplifier to a reference voltage, that's set at 3.3 or is adjustable by jumpers switching in different resistor values for 3.6 etc. Possibly finding the actual reference regulator and looking up the application note might give us more clue what's happening and whether this transistor is right for the job.

Good point. There's an 8 pin chip by the edge of the board near the transistor. It's likely involved with switching the transistor, but I can't make out all the lettering on it. The TIP32C datasheet says it can be used for switching, so they may have used a PNP as part of a switched power supply. The UH19 photos say they were provided by Deksor, so maybe they can help confirm the ID of the transistor?

snufkin wrote on 2021-05-30, 13:51:

BitWrangler wrote on 2021-05-30, 13:36:

Back when RadioShack still had a few components they only had about 10 transistor types in store, so if you wondered in saying "I need a transistor like this" you'd end up with one of those 10... and I seem to recall the TIP32 being one of them... so I am wondering if that is actually the right part number for that board. What it should be doing is acting as current amplifier to a reference voltage, that's set at 3.3 or is adjustable by jumpers switching in different resistor values for 3.6 etc. Possibly finding the actual reference regulator and looking up the application note might give us more clue what's happening and whether this transistor is right for the job.

Good point. There's an 8 pin chip by the edge of the board near the transistor. It's likely involved with switching the transistor, but I can't make out all the lettering on it. The TIP32C datasheet says it can be used for switching, so they may have used a PNP as part of a switched power supply. The UH19 photos say they were provided by Deksor, so maybe they can help confirm the ID of the transistor?

That tiny 8 pin chip reads:
53KY
LP295
1ACMA

Does that help?

fenderstrat wrote on 2021-05-30, 14:09:

That tiny 8 pin chip reads: 53KY LP295 1ACMA […]
Show full quote

That tiny 8 pin chip reads:
53KY
LP295
1ACMA

There's an LP2951 voltage regulator that looks like it might be it. Check if pin 4 connects to Ground and if pin 1 connects to either +5V or +12V. The output from this chip might control the Base pin of the large transistor, with the Vout/Collector pin going through a resistor divider to provide the feedback back to this chip.

But if that is an LP2951 then I'm not sure at the moment if a PNP transistor is the right thing to use where the TIP32 was. If the output voltage from the TIP32 is too low, then that will increase the output voltage from the LP2951. Assuming that output is connected (through a resistor) to the Base of a TIP32, then increasing the Base voltage will reduce Vbe (Emitter was at +5V), and so the current flowing through the collector, which will reduce the output voltage from the transistor even further, when the voltage should be increasing. That said, it always takes me a while to get my head around analogue stuff, so I may have something back to front there.

It really would be good if anyone with the same board can confirm what that part should be.

Could you measure the resistance between pin 1 (output) of the LP2951 and the three pins of the TIP32 (particularly pin 1 (Base))? That'll check if the LP2951 is controlling the TIP32.
Also measure the resistance between pin 7 (feedback) of the LP2951 and the three pins of the TIP32 (particularly pin 2 (Collector)). That shows if the LP2951 should be changing its output based on the output from the TIP32. I'm hoping we can figure out if the TIP32 is actually the wrong part based on what it's connected to.

Bloodem's suggestion of testing a Vcc pin with the motherboard configured for 5V is a good one. I think that should bypass the regulator entirely.

TIP32C seems to be the correct transistor (Whether it's up to the job long term might be up for speculation)
Tomato 4DPS 2.0 + Cyrix DX2-66 - not work

Odd that never showed up until I was looking for TNT 64 info and quite serendipitous.

Ok, I set the CPU voltage to 5V with the jumpers, placed a pin into one of the VCC holes in the socket and I got 5 Volts! Finally, a step in the right direction. Tried another VCC phole and it also read 5V. So I guess the transistor does get bypassed when the voltage is set to 5. In any case, I have a replacement on the way, hopefully it'll get here soon. Thanks!

Great news! 😀

Ok, I replaced the TIP32 C and now it works like a charm! Thanks everyone for all your help!