[marmotta]

Hi! I want to try to repair a ip28 motherboard, I think is a overvoltage the cause of damage (I didn't damage it)

I have the working ip22 (it's almost the same) I have tested the power lines and is all ok... except the pin 5 on j10, on ip28 is present more resistance, I presume the issue is here. (5v standby)

Not exist any schematics of ip22/ip28? Multilayer pcb is very difficult to reverse engineering...

[robespierre]

As mentioned here, most of the internal information (schematics, etc) appears to have been destroyed, although a few bits leaked out here and there. Reconstructing information about multilayer boards is possible, but requires either destructive analysis (sanding) or very clever tricks with NDT techniques.

[weblacky]

This, there won’t ever be schematics.

I recently managed to haggle for just under half price for a Huntron 3200s for Non-Destructive Testing.

I’m clearing old projects right now but when starting it I plan to focus on Power Supplies first as that’s that’s the greatest natural need.

You’re not the first person I’ve heard about wanting to tackle mainboards and eventually things will come full circle and we’ll have that need.

While I applaud your initiative, you say you have a similar board and found a difference but it’s higher?

I’d say, without more information from yourself, and just going by your description, that sounds like a design difference between revisions. Not a suspected failure.

Normally a short has a short in it, in that it’s much less resistance. There are other cases involving semiconductors that don’t quite show up unless you use diode mode to find shorts.

It’s not a hard rule but most common. If you have two identical boards (same model number and same rev) then I’d say you’re on to something. But you have a similar board with a different processor family on it, so this isn’t exactly comparing to identical boards.


Perhaps you should tell us the symptoms of what about the board isn’t working?

Also resistance is an ok place to start, but on semiconductors and caps, there are better modes and ways. A low voltage bench PSU injection along with diode mode on the PSU connector will likely tell you a lot more about actual shorts in that area and the main rails.

[marmotta]

Now I have much time... I misured exactly: ip22 is 5,95kohm ip28 is 3,73kohm.

Now I try to put a 5v on standby and read the current absorbing and eventually heat on ic close the dallas clock (I suspect this ic’s).

I have only this bare motherboard, I not have the r10000 cpu, if I repair it, I buy the cpu 😉

[Irinikus]

Another useful way to detect failed components is the view the board through an infrared camera while powered up. Failed IC’s tend to show up as hotspots.

[marmotta]

I think I'm off track... ip22 consumption is 75mA ip28 78mA
The difference is irrelevant

---

This is the boards... is almost the same, the only difference is close the cpu slot.

If I insert the r4400 cpu on ip28... can be dengerous? 

[robespierre]

If I insert the r4400 cpu on ip28... can be dengerous? 

I don't think the processor module socket is the same, but it's been many years since I compared them.
I suggest you don't try it anyway, the CPUs have different voltages and for sure will not work, since the address bus on IP28 is 64 bits compared to 32 bits on IP22

[marmotta]

If I insert the r4400 cpu on ip28... can be dengerous? 

I don't think the processor module socket is the same, but it's been many years since I compared them.
I suggest you don't try it anyway, the CPUs have different voltages and for sure will not work, since the address bus on IP28 is 64 bits compared to 32 bits on IP22

Yes, try is not a good idea 
But the connector is the same (look a picture)

[weblacky]

What is the chip that's getting hot and do you measure any issues with diode mode between it's VCC and GND and it's legs?

[marmotta]

What is the chip that's getting hot and do you measure any issues with diode mode between it's VCC and GND and it's legs?

I searched for hot chip... but not found (is all cold). The story of the motherboard is psu fried and consequently motherboard damaged (words of previous owner). My first idea is testing the power connector and the power lines... but seems to be all ok 

Remain to test with cpu, when I find one...