The post about the processor slot makes me wonder if the OP is experiencing something I did a time ago when I tried to use Superglue as a thermal compound (I was inspired by this idiotic practice done by gluing a fan right to a overdrive chip) - got some of that superglue in the socket, used Acetone and a .024W guitar string to clean the pins off. The CPU in the computer was trashed though, could never get all the pins clean, still have it though.

Might pay to get a very thin wound guitar string and CAREFULLY use it to clean the pins on the processor socket just in case of corrosion/dirt/something leaked in there and is preventing good contact with the CPU.

GPA wrote:

good man, don't trash it! measure resistance to ground from those pins, if they're not shorted you can solder wires directly fr […]
Show full quote

good man, don't trash it!
measure resistance to ground from those pins, if they're not shorted you can solder wires
directly from voltage regulator to them. If regulator works.
Although i presume, there are more problems with this board hehe

Yeah, gonna be an awesome for decoration. Fact is when I fail to recover motherboards,
I just want to get rid of them. They make me remember about all the mistakes.

Then I will do that in each VCC (If I know what the hell I am doing - listen if the multimeter
emits beeps, right ?). More problems ? The readings tell you all that ? Where? 😢
Thank you for your honest.

FesterBlatz wrote:

I just repaired a motherboard with a similar fault on Monday. The large output transistor for the CPU series pass regulator had shorted the 5-volt rail to CPU VCC. Look for a large transistor...

I've spoted the transistor. Never took the time to measure it.
It is not screwed down to the motherboard. The holes don't match.
Uhh...Maybe someone did some poor job there! I will compare with photos on net.

FesterBlatz wrote:

Look up the part number and see if you can find a datasheet to show which pin is which. When the board jumpered for 3.3-volt CPU and powered up, carefully measure the voltages on the transistor's pins.
...
What's puzzling is not all the VCC pins showing voltage. Are you sure your probe was making a good clean contact with those pins?

I'll be honest with you. I don't even know how to measure VCC pins with a multimeter. I stuffed a paper clip into the holes - A needle works, too. I don't see why trust them is an issue. And I always repeat my measures to be sure.
Be patient and tell me the right way to do that. 😎

You certainly are giving me a hard time.
I will do my best to follow your command. I do valorize the little things I have. 😀

Frasco wrote:

FesterBlatz wrote:

I'll be honest with you. I don't even know how to measure VCC pins with a multimeter. I stuffed a paper clip into the holes - A […]
Show full quote

I'll be honest with you. I don't even know how to measure VCC pins with a multimeter. I stuffed a paper clip into the holes - A needle works, too. I don't see why trust them is an issue. And I always repeat my measures to be sure.
Be patient and tell me the right way to do that. 😎

You certainly are giving me a hard time.
I will do my best to follow your command. I do valorize the little things I have. 😀

Be careful, a paperclip is pretty large and may be over-extending or damaging contacts in the CPU socket. A safer bet might be to measure the pins on the back of the motherboard instead.

What is the part number of the large transistor?

creepingnet wrote:

The post about the processor slot makes me wonder if the OP is experiencing something I did a time ago when I tried to use Superglue as a thermal compound (I was inspired by this idiotic practice done by gluing a fan right to a overdrive chip) - got some of that superglue in the socket, used Acetone and a .024W guitar string to clean the pins off. The CPU in the computer was trashed though, could never get all the pins clean, still have it though.

Might pay to get a very thin wound guitar string and CAREFULLY use it to clean the pins on the processor socket just in case of corrosion/dirt/something leaked in there and is preventing good contact with the CPU.

I used superglue to stick a fan on a Voodoo3 heatsink. Works like a charm, but I guess it doesn't make any difference in terms of longevity or whatever. Besides 3DFX workers know their stuff.

You know what ? I could open the socket 3 and see what's going on inside. I've been told this is
a risk operation. (Tetrium ?) and need some practicing with another dud motherboard.

An stiletto will mow the superglue in your processor.

To Fasterblatz:
The paperclip is so small, but I got the point - just not confortable measuring in the
back of motherboard, I guess. The part number is:
542
D882-Y

Frasco wrote:

The part number is:
542
D882-Y

Ah, an NPN transistor--in that case, the load (emmiter) and source (collector) are reversed from what I mentioned earlier.

Pin 3 should be 3.3v to 3.45v
Pin 2 doesn't matter
Pin 1 should be 5v.

If Pin 1 and 3 both measure 5V, then the transistor has shorted.

Just to be safe I would recommend trying to redump the bios. Big 'ol solder blob to fix the leg and multiple dumps to average out and check for errors.

FesterBlatz wrote:

Ah, an NPN transistor--in that case, the load (emmiter) and source (collector) are reversed from what I mentioned earlier. […]
Show full quote

Frasco wrote:

The part number is:
542
D882-Y

Ah, an NPN transistor--in that case, the load (emmiter) and source (collector) are reversed from what I mentioned earlier.

Capture.PNG

Pin 3 should be 3.3v to 3.45v
Pin 2 doesn't matter
Pin 1 should be 5v.

If Pin 1 and 3 both measure 5V, then the transistor has shorted.

My readings: Pin 3 = 4.2v and Pin 1 = 5v

And I compared my transistor via photos on net and I was right: Someone replaced it and didn't install a screw. 😠

xplus93 wrote:

Just to be safe I would recommend trying to redump the bios. Big 'ol solder blob to fix the leg and multiple dumps to average out and check for errors.

Think safety - my first dump went wrong. 😵 and even received an OK message !
After 4 equal dumps, I was ready to go. Now that you said errors, I noticed a big blank
space in the dump, but then again, I compared thoroughly those files.
Solder blob ? Are you talking about a trick ? My blob fall off all the time. I shake when dealing with it 🤣

Frasco wrote:

Would you mind explain how to measure the transistor ? Cause my readings are totally messed. Pin 3 = 0.40 and Pin 1 = 0.00 or 0. […]
Show full quote

FesterBlatz wrote:

Ah, an NPN transistor--in that case, the load (emmiter) and source (collector) are reversed from what I mentioned earlier. […]
Show full quote

Frasco wrote:

The part number is:
542
D882-Y

Ah, an NPN transistor--in that case, the load (emmiter) and source (collector) are reversed from what I mentioned earlier.

Capture.PNG

Pin 3 should be 3.3v to 3.45v
Pin 2 doesn't matter
Pin 1 should be 5v.

If Pin 1 and 3 both measure 5V, then the transistor has shorted.

Would you mind explain how to measure the transistor ?
Cause my readings are totally messed. Pin 3 = 0.40 and Pin 1 = 0.00 or 0.01

If this is "correct" readings, then I screwed big time !
And I compared my transistor via photos on net and I was right: Someone replaced it and didn't install a screw. 😠

xplus93 wrote:

Just to be safe I would recommend trying to redump the bios. Big 'ol solder blob to fix the leg and multiple dumps to average out and check for errors.

Think safety - my first dump went wrong. 😵 and even received an OK message !
After 4 equal dumps, I was ready to go. Now that you said errors, I noticed a big blank
space in the dump, but then again, I compared thoroughly those files.
Solder blob ? Are you talking about a trick ? My blob fall off all the time. I shake when dealing with it 🤣

With 4 matching dumps I would say it's good. But if you wanna give it a shot. Place chip in programmer, put in the broken off leg. Bend leg out just enough to be able to touch your iron between leg and stub. Place a very tiny bit of tin foil underneath where you'll be placing blob. That's both to protect your programmer and help hold solder so it doesn't flow away. Then heat up leg and stub and place your blob. Easy on paper, but i'll admit it's a bit tricky.

If the board is configured for 5V, it will bypass voltage regulator. You might try that, if you have 5V CPUs. I did that to one of my boards, the regulating transistor failed, i took it away and board works fine with 5V CPUs without it.
But that is if the transistor is guilty. Configure the board for 5V CPUs, then measure resistance between VCC pins and +5V rail pin in PSU connector on the motherboard to check. You should read close to 0 Ohms. If it's ok you may try running a 5V CPU in it and see if it posts. If it doesn't, leave the regulator alone, there are way more problems with the board.

Frasco wrote:

An stiletto will mow the superglue in your processor.

I will WHAT? 🤣

Stiletto wrote:

Frasco wrote:

An stiletto will mow the superglue in your processor.

I will WHAT? 🤣

You will remove the superglue on his processor - although this is well known and very simple.

Yeah, the moment I was writing "stiletto", I thought about what you did back there
- freaks everywhere (for google's sake)
😀
I saw you coming to my topic.
You know, I have a surprise for you in case I succeed.

GPA wrote:

If the board is configured for 5V, it will bypass voltage regulator. You might try that, if you have 5V CPUs. I did that to one of my boards, the regulating transistor failed, i took it away and board works fine with 5V CPUs without it.
But that is if the transistor is guilty. Configure the board for 5V CPUs, then measure resistance between VCC pins and +5V rail pin in PSU connector on the motherboard to check. You should read close to 0 Ohms. If it's ok you may try running a 5V CPU in it and see if it posts. If it doesn't, leave the regulator alone, there are way more problems with the board.

Configured for 5V, same results in regard of VCCs. And now the readings: Pin1=5V and Pin3=4.5V
I am very confused. Is this transistor what we call MOSFET nowadays ? I suck at electronic.

Let me tell you a DX4/100 (3.3v) or Cyrix 586 (3.3v) will work at 5V (for testing purpose). So I did it to no avail.
OK. There are tons of problems. Now tell me why?
I saw it with my own eyes POSTing, the Windows BIOS and everything...

Frasco wrote:

You know, I have a surprise for you in case I succeed.

... 😒 😢 😵

Stiletto wrote:

Frasco wrote:

You know, I have a surprise for you in case I succeed.

... 😒 😢 😵

It's good. Very funny. Too bad my chances are overwhelming...

OK. Let's get to the bottom. Seeing 3 things in the horizon:

GPA wrote:

measure resistance to ground from those pins, if they're not shorted you can
solder wires directly from voltage regulator to them. If regulator works.

There is no short in the pins. I need to know - Is my regulator in fact dead already?
Should I remove it as a last resort ?

2) Take heed to gerwin words:
Some experiences of mine that match a totally dead board:
- Voltage Step down chips overheated in the past, solder got loose, therefore the Voltage
Regulator Module gives no 'Power-Good' signal to the BIOS.

Module Regulator = Regulator ?
I know a simple trick about reading the "power good" on the Power ON connector and this is it.
How I check it out in my VL4200?

3) Scenario has changed:
VCC marked = 0.01v
The rest = 5v
Of course the processor gets very hot. For god's sake, What's going on here?

My men, if you think there is no way out, I stop the investigation right away.
Not to mention, seeing a foreigner talking in English must be a PITA!

FWIW I have a VL4200 and a clean BIOS dump of it. I'm working on modding it to add real cache and missing jumpers. It performs well without cache, but obviously could be better. Such a nice clean late model board.