[Hulting74]

Hi

Anyone seen this issue?
Starts with a green light but do not boot and no chime, and if i press VOL Down or VOL Up it turns off.

Tried to reseat GPU, and memory. Removed dallas but no luck :-(
Also tested to replace the eeprom (firmware 5.0 and 5.3)

Suggestions?

/S

[dexter1]

There could be many things going wrong but the usual suspects are PSU, memory or maybe the motherboard

First thing to try is to hook up a serial console and check if there is any activity. If that stays dead, start stripping the machine leaving only the bare essentials: one 4-dimm memory bank, no gfx, no disks.

[Hulting74]

I dug up another working Indy and tested the parts individually, seems like the motherboard is broken... :-(

Anyone have any information if there is any known or "common" faults on Indy motherboards?

[robespierre]

Anyone have any information if there is any known or "common" faults on Indy motherboards?

I believe that I have heard more than once about tantalum cap failures on the Indy. Those are the large-ish (6 to 10mm) square yellow ones with a stripe. They fail by going short.
To test you should remove all possible users of power from the board (CPU, memory, cards) and then check the resistance through one of the caps. If it is very low (< 10 ohm? This needs to be checked on a good board) then there is a short, although not necessarily the same one you checked, as they can be in parallel in the circuit.

[Hulting74]

Anyone have any information if there is any known or "common" faults on Indy motherboards?

I believe that I have heard more than once about tantalum cap failures on the Indy. Those are the large-ish (6 to 10mm) square yellow ones with a stripe. They fail by going short.
To test you should remove all possible users of power from the board (CPU, memory, cards) and then check the resistance through one of the caps. If it is very low (< 10 ohm? This needs to be checked on a good board) then there is a short, although not necessarily the same one you checked, as they can be in parallel in the circuit.

Hi

Thanks for the input :-)

I went through all yellow caps and they seem to test ok, also tested if there was any capacitors that had "ground" on both sides but all seemed ok.

Finally found the serialcable, and got some nice messages.

### SGI Indy Errormessage ###
Data path test                            *FAILED*
        WD SCSI0 path test                *FAILED*
        RTC path test                      *FAILED*

Exception: <vector=Normal>
Status register: 0x30004803<CU1,CU0,IM7,IM4,IPL=???,MODE=KERNEL,EXL,IE>
Cause register: 0xc000<CE=0,IP8,IP7,EXC=INT>
Exception PC: 0xbfc3eebc, Exception RA: 0xbfc02138
Interrupt exception
Bus Error ?
Local I/O interrupt register 2: 0x12 <KBD/MOUSE>

NESTED EXCEPTION #1 at EPC: 9fc3eebc; first exception at PC: bfc3eebc
### End SGI Indy Errormessage ###

Could it be that my dallas battery is finished? Is it really that easy..?

Time to dig in the spareparts drawer :-)

[weblacky]

It's more likely have you a short on your 5v line after the circuitry that combines the RTC with main PSU standby 5V power. I had a PC (not SGI) do this and finally fixed it this year with a voltage injection test. But MISSING stuff usually means - either disconnected or unpowered (low to zero volts).

I'd suspect unpowered, a localized short that's behind a voltage regulator will still output MAX current for that voltage (feeding the short) and still be near 0 volts after the nearest limiter/resistor. A thermal imager will spot this quickly, failing that...your finger...depending on the max current (somes it's not enough to burn you but just be warm).

But based on your symptom (and not saying anything about your PSU being OK or not), I think you'll find the RTC signal may not be present. I'd check the RTC socket for 5V from PSU, if not...you're in the short's geographic region. If you are getting 5V at the RTC socket, then I'd check the SCSI controller for voltage.

I'm not sure what "Path test" means...but I'd assume it's saying...I don't hear them. So that's why I'd go with unpowered. If those two ICs share a voltage source (regulator) then that could be your area that has been shorted (if it's a short). It's pretty easy to tell my going to check the ICs themselves first for a short on their VCC pins.