Dell Optiplex GX110- Error 8602 and No PS/2 Controller

[TechGeek]

I recused this Dell Optiplex GX110 several months ago. It had sat for months or years outside.

Powered it up, PS/2 controller reports keyboard failure with no keyboard plugged in. No keyboard failure reported with a keyboard plugged in, but unresponsive to all inputs.

What should I check first? New CMOS battery installed, machine taken to bare essentials. Have not tried shorting the chassis intrusion switch.

[momaka]

Re: Dell Optiplex GX110- Error 8602 and No PS/2 Controller

For a machine that has sat outside for a long time, I think the first few checks to do is see if all traces are intact on the motherboard. Contaminated water and dust can cause tracks to corrode or become conductive, so watch out for that.

Also, if it was really cold outside, make sure to give the PC adequate time to "climatize" to your room temperature. Bringing cold stuff inside a hotter environment tends to cause condensation, and that can throw things off (or even cause damage in some cases) when the device is powered on.

Finally, when things sit outside for that long, some of the contacts in the various ports and I/O are bound to get at least a little bit of oxidation on there. Removing and inserting all parts with contacts (CPU, RAM, add-on cards, and even peripherals such as mouse and keyboard) may help strip some of that away and make whatever is misbehaving work. So give that a try. Better yet, first give the MB a wash with dish detergent and water, then let dry for a few days (at least), then try removing and inserting everything several times.

Oh, and this should go without saying but... lookout for bad caps!
Especially Nichicon HM and HN from that era (2001-2004), as they are very prone to failure and quite common in Dell PCs. Even if they look "OK", if they are from that period, chances are extremely high they are bad.

I have a Gateway desktop here as an example of that. Tried it last year after acquiring it from a PC repair shop. It booted up the first time I tried it, so I thought all was OK. A month or so later, I tried booting it up, and it would refuse, unless I used only one stick of RAM. I eventually "tricked it" to accept two sticks of RAM by making it boot with one, then shutting down and inserting the second stick. But even then, I still got random crashes every once in a rare while. When I looked at the motherboard, it had 3x Nichicon HM caps near the CPU. They looked good and they were inter-mixed with poly caps, so I was suspicious that would be the cause of my issues. But then I decided to remove one, and sure enough, it read over 9900 uF (it was supposed to be 3300 uF) - clearly out of spec! The other two had similar readings. I capped the motherboard, and these issues I was having were gone.

So yeah, lookout for bad caps.

[Wester547]

Re: Dell Optiplex GX110- Error 8602 and No PS/2 Controller

Oh, and this should go without saying but... lookout for bad caps!
Especially Nichicon HM and HN from that era (2001-2004), as they are very prone to failure and quite common in Dell PCs. Even if they look "OK", if they are from that period, chances are extremely high they are bad.

I googled the GX110 motherboard and looked for one on eBay. It seems these boards only use solid capacitors around the CPU (Sanyo OS-CON SP, TCNQ complex salt / organic semiconductor). I guess those may not last forever either despite using epoxy bungs, their leakage current may increase over time and capacitance may decrease (and ESR may increase), but I doubt they are the source of the problem. This is more likely the problem:

Finally, when things sit outside for that long, some of the contacts in the various ports and I/O are bound to get at least a little bit of oxidation on there. Removing and inserting all parts with contacts (CPU, RAM, add-on cards, and even peripherals such as mouse and keyboard) may help strip some of that away and make whatever is misbehaving work. So give that a try. Better yet, first give the MB a wash with dish detergent and water, then let dry for a few days (at least), then try removing and inserting everything several times.

All that said...

I have a Gateway desktop here as an example of that. Tried it last year after acquiring it from a PC repair shop. It booted up the first time I tried it, so I thought all was OK. A month or so later, I tried booting it up, and it would refuse, unless I used only one stick of RAM. I eventually "tricked it" to accept two sticks of RAM by making it boot with one, then shutting down and inserting the second stick. But even then, I still got random crashes every once in a rare while. When I looked at the motherboard, it had 3x Nichicon HM caps near the CPU. They looked good and they were inter-mixed with poly caps, so I was suspicious that would be the cause of my issues. But then I decided to remove one, and sure enough, it read over 9900 uF (it was supposed to be 3300 uF) - clearly out of spec! The other two had similar readings. I capped the motherboard, and these issues I was having were gone.

So yeah, lookout for bad caps.

Yep.

I have lost count of how many bad era Nichicon HM and HN (those made before the second quarter of 2005) I've seen fail this way, with or without bulging or leaking. The electrolyte attacks the aluminum oxide layer and leakage current increases dramatically, causing a partial (or full) short in the dielectric, leading to the aforementioned issues.

KZG and KZJ (made before 2008) are affected by this problem as well, I believe, but to a lesser extent, and in KZG's case, usually they bulge when they give up just sitting in storage (from what I've seen, especially the 3300uF 6.3V).

Some bad era Nichicon HZ (those made before 2006) fail this way too. Sadly, the older Sanyo WG (those made before 2005) have a tendency to fail this way too and the early Sanyo WX die in that manner as well (those made before 2002), in my experience, regardless of physical appearance. Sanyo WF will also fail like that no matter when they were made.