Indigo 150mhz Static Ram Error

legodude

Member
Apr 2, 2024
62
12
8
My dead MIPS CPU woes continue... I was lucky enough to pick up a cheap 150mhz CPU for my Indigo a while ago and all was good until I went to use it recently. It failed partway through the boot process and I was at a loss until I hooked up a serial console:

1767193296765.png


The error is quite helpful at pinpointing the exact problem, and U4 is helpfully labeled as the third chip from the bottom on the right:

1767193331450.png


An Ebay seller had the replacement chips in stock so I broke out the hot air station and went to town:
1767193486653.jpeg


1767193552347.jpeg


1767193654512.jpeg


Sadly, the error is exactly the same:
1767193682601.png


I doubt both chips failed in exactly the same way. I can't see any other damage to the CPU board. The system works fine with the old 100mhz CPU. Any ideas?
 
Okay, now the system won't start with the 100mhz CPU. I'm worried the system board is on the way out.
Sigh
 
Yeah I was actually going to say don't remove the chip yet and actually check the traces for the connection to the chip. SRAM really-ish doesn't go bad, unless something external acts on it.

Broken solder joints are real thing, you see how you got a bunch of ffff, that normally doesn't indicate a bad cell it indicates that there's a total lack of communication to it. Given that there was no communication my guess would be a power, ground, or lack of enable signal. Chips have have a chip select line that tells them to pay attention or not, and often there's another chip that tells them to do this, some form of multiplexer.

My gut tells me that you either have a poor power connection to the chip or the enable signal never comes to the chip from the thing that controls it. If you look up the data sheet for the SRAM chip you'll find the CS line which should be the chip select line. Use a multimeter and find where that leg goes on the other SRAM chips. Because you'll probably find that either the multiplexer is damaged or you don't have a connection on that chip enable pin/leg for just U4, but the other ones do. So use the other ones to track down the Multiplexer (for that bank), likely that's where you're fix needs to be. Either a physical trace/solder joint cracking at the Multiplex to U4 CS line/pin or the Multiplex itself, since the rest of the bank seems to work.
 
Last edited:
Thanks for the comments. I don't see a multiplexer on the PCB. I think the smaller chips are resistor packs and the other IC is an octal flip flop, plus the two 8pin packages which I haven't really investigated.

I did some probing with the multimeter. Ground, VCC, and address lines all seem okay, as do some other control signals. CS line is shared with the chip on the left side of the PCB, so that seems good to me as well. I'm probing leg-leg so that should take into account soldering deficits as well.
 
One trick I'll divulge to you, when you probe a chip leg to leg don't touch the leg. This can give a false positive reading from the downward pressure. One way you locate bad solder joints is to make sure that you're not putting any force on the board that would cause flexing but also don't put your test lead on the solder joint which could momentarily push it together to complete the circuit. Make sure you're always probing just away from the solder joint on the pad area of the track so hopefully you catch any separation that way.

Also make sure that you are performing both readings for ohms and a reading in diode mode when performing your tests. What I would do is also compare bank to bank readings. That is if you assume positionally that U4 as a doppelgänger in the other bank, use ohms mode and diode mode to measure all the pins of the suspect chip and compare those readings to its doppelgänger in the other bank. You would assume incredibly similar to same values. One of the great things about troubleshooting circuits that have replication in them is that multiple channels or multiple grouping of things often have duplicate designs in the same circuit with the same parts that are the same vintage from the same supplier. So you can compare parts of the circuit with itself and any differences can lead you down a new path.
 

About us

  • Silicon Graphics User Group (SGUG) is a community for users, developers, and admirers of Silicon Graphics (SGI) products. We aim to be a friendly hobbyist community for discussing all aspects of SGIs, including use, software development, the IRIX Operating System, and troubleshooting, as well as facilitating hardware exchange.

User Menu