[Noris]

Completely stumped here... VCAM POKA is failing!

The voltages on backplane seem fine: 48v is stable, albeit slightly high at 48.8v) and the 5v and 12v rails come up fine during power on but get powered down as soon as controller hits the bad POK from the VCAM board. The power boards are all ok, but io4 error LEDs indicate bad -12v regulator (-5.2v one shows no error). I have tried another IO4 and it is exactly the same so thought it could be dust, bent pins, poorly seated card or something but everything seems good.

The diagnostics document I'm using has a figure suggesting to disable VCAM to find error, which I dont knowhow to do... is this worth trying? Is it just a case of removing the vcam or is there more to it? Will this get me anywhere anyways because that seems to just be removing the troublesome device!

The other curious thing is that the diagnostics roadmap indicates that the io4 draws 48v straight of the backplane but there doesnt appear to be any significant converter blocks on it... is my interpretation of the diagram correct? If not I'm thinking the 512 could not be providing correct voltage, thought LEDs are all amber which I think indicates power outputs are ok.

Also with no cards installed except for controller the system gets through power on but fails with no system clock. Is this referring to one of the dallas timekeeper chips on the controller/io4 or is the system clock elsewhere? It is highly likely the batteries are dead but I thought they were only for nvram etc and noncritical for booting up the system.

This fascinating world has quite a steep learning curve so any advice / experience would really help!

Regards

[mosiniak]

"No system clock" shows up probably because You removed IP board. I'll confirm it this evening.

There is also possibility that backplane is damaged. You can inspect all connectors for bent pins, obstacles, dust, etc. -12V leaves VCAM and is available on the backplane (VME bus) - maybe there is short circuit.

[jan-jaap]

(I moved the tech bits over from the other discussion thread)

Jan-jaap, is the VCAM board necessary to power on the Onyx into PROM? VCAM if I'm right is only the "bridge" between everest bus and VMEbus. What is the minimal Onyx configuration allowing entering the PROM (system controller, IP, MC3, IO4 for console, 512?)? 303 and 505 boards are needed for powering the graphics (303 for IR and 505 for RE), right?

I think you're right. VCAM also bridges the graphics to the everest bus, btw. You need one IPnn, one MC3 and one IO4. You need a 512 to power the IO4 (MC3 and IPnn have their own 48V power bricks). In my experience the -5, -12V inverters (powered from the +12V rail) die quite often, so they may simply both be dead. I think the boot master IPnn provides clock to the backplane, I've seen that clock message myself at some point, but it's been years since I stripped down an Onyx that far. I think you can even get some output from the system controller serial port with only a CPU board installed and the sysco "debug menu" correctly set up, but this will obviously not complete POST but reach POD at best.

When swapping out bits, beware that IR graphics only work with the last VCAM rev installed. Older IO4 boards may lack PROM support for IP21 or IP25 CPU boards.

The Challenge/Onyx series are temperamental beasts, and their complex power subsystem can be a real pain to debug. The IR or RE2 graphics board sets can be extremely fragile, especially the RM boards.

[Noris]

I have thoroughly checked for bent pins and it all looks fine and really clean but I might try some isopropyl on the relevant cards.

It's useful to know system clock is on CPU board. So can I just remove vcam and try booting with just mem, CPU and io4?

[mosiniak]

So can I just remove vcam and try booting with just mem, CPU and io4?

and 512 and system controller

You can leave other cards disconnected in their slots and run the system with doors open. For short time it will not overheat. Leaving boards in vertical position is, in my opinion, the safest for them.

If there is a problem with VCAM You can monitor voltages generated by the converters. There are test-points (see attached photos), You can connect cables to them and install board in the system.

Btw. the converters LT1074 cannot be powered from 48V (max 45V) so they must be powered from lower voltage (12V from 512 power board).

Try using 512 board from second rack.

[Noris]

So can I just remove vcam and try booting with just mem, CPU and io4?

and 512 and system controller

You can leave other cards disconnected in their slots and run the system with doors open. For short time it will not overheat. Leaving boards in vertical position is, in my opinion, the safest for them.

If there is a problem with VCAM You can monitor voltages generated by the converters. There are test-points (see attached photos), You can connect cables to them and install board in the system.

Btw. the converters LT1074 cannot be powered from 48V (max 45V) so they must be powered from lower voltage (12V from 512 power board).

Try using 512 board from second rack.

Wow that's brilliant thanks mosiniak you're a star. Well i'll try minimal config as suggested. Can i even try it without the whole IO4?

It is strange that both VCAM boards were failing with same error but knowing about those test points gives me some hope of moving forward.

The 512 seems ok as i have already tested the +12v and +5v lines on backplane and they are good. There is only a single 12v + 5v power board in the other rack (no number like others but it clearly has 12V and 5V printed on the back of board) so i can try that one if suitable.

Incidentally i did wonder whether the various power boards need to be in any particular power slot but since there is nothing in the docs saying the 512 needs to be in slot 1 or whatnot i assume position is not important. In any case i didnt fancy shifting them around!

[mosiniak]

You must be very careful when playing inside working Onyx. The power bricks are very powerful (512 can deliver up to 40Amps at 5V and up to 17Amps at 12V). Any short circuit can be very painful. 40Amps can melt down or burn wires / pcb tracks in seconds.

[Noris]

You must be very careful when playing inside working Onyx. The power bricks are very powerful (512 can deliver up to 40Amps at 5V and up to 17Amps at 12V). Any short circuit can be very painful. 40Amps can melt down or burn wires / pcb tracks in seconds.

Eek you could probably weld up a car panel with that let alone pcb tracks! thanks for heads up.

[Noris]

Well it seems a bit of contact cleaner and reseating cards and / or cleaning and retinning the earth points on the vcam has done the trick!

I did try booting with only controller mc3 and ip25 and that produced no system clock error. Adding the 512 power card caused under voltage -12 and -5.2 errors even without the io4 installed which is slightly surprising. Io4 with no vcam produced same under voltage error..

So it seems that the system clock is on the io4 board? Not sure this is sensible conclusion though as it doesnt feel right... The io4 Dallas is soldered to board, where as on the system controller it is in a socket. There is also what looks like an oscillator chip on the back plane which could be it. Then again maybe it is on the CPU and it was the contact cleaner that fixed it. Did you manage to check where this system clock resides mosiniak?

Anyhow Vcam is putting out -13v on -12v according to my cheapo multimeter which must be on the edge of an over voltage failure but I guess this would come down a bit with some vme cards installed

Anyhow this screen made me feel a lot more positive than I was yesterday

[jan-jaap]

So it seems that the system clock is on the io4 board? Not sure this is sensible conclusion though as it doesnt feel right... The io4 Dallas is soldered to board, where as on the system controller it is in a socket.

There are two Dallas chips in the Onyx (more if you have multiple IO4s). Both of them store the SSN (system serial number). I believe the one on the system controller keeps the system time and the one on the IO4 keeps the environment variables.

I tried to desolder a Dallas from an IO4 once and gave up on it for fear of destroying the PCB which is very thick. There are probably some ground planes inside which are very effective at absorbing and spreading the heat. I ended up modifying the Dallas while it was installed. You can read about it here: https://www.vdheijden-messerli.net/sgist...io4-nvram/

I did a similar hack on the Dallas on the system controller, but I think the post went down with Nekochan... still have the photos somewhere. For now there's just the attached photo.

Regarding the SSN: in SGI systems, the serial number or MAC address is normally "attached" to the chassis, not the CPU or mainboard. In this case it's stored in the system controller, with the IO4 as a backup. If the serial numbers mismatch, the system controller "wins" and updates the IO4. A blank Dallas in the system controller will be initialized using the IO4.

Only if both are empty will a special procedure work to set the SSN. it is described here: http://www.futuretech.blinkenlights.nl/c...procs.html in sec 6.3.3.5