• Please review our updated Terms and Rules here
  • Exhibitor application for VCF West 2022 is now open! If you are interested in exhibiting, please fill out the form here.
  • Here are the results of the VCF East 2022 Post Event Survey: Survey Results

PS/2 model 70 - unreliable power on

modem7

Veteran Member
Joined
May 29, 2006
Messages
7,785
Location
Melbourne, Australia
Recently there was a thread somewhere in the forums that deviated onto the subject of PS/2 model 70 reliability. At the time, I looked for my old notes from the 80s/90s because I remembered that at one time, IBM gave us a new motherboard for every PS/2 model 70 that we had, because of a particular problem.

I've relocated my notes. In the 1992 section, there is an entry about the problem. I'll provide the detail here in case someone can use it.

SYMPTOM: The PS/2 model 70 sometimes does not power on.

DIAGNOSIS: From the power supply there are 2 signals produced. On the motherboard, there is a Tiawonese made chip which it turned out doesn't always meet specs - this affects one of the 2 signals.

RECTIFICATION: Once IBM discovered the issue, future runs of the motherboard were corrected by compensation. The compensation is that the U24 delay line on the motherboard (next to the 386) was increased in value from 10nS (p/n 23121) to 12nS (p/n 23125).

IDENTIFICATION: The affected motherboards can be identified by their barcode. There will be a white number of the form 0XXXXXX90. Affected motherboards have a XXXXXX number of less than 352362.
 

IBMMuseum

Veteran Member
Joined
Aug 28, 2006
Messages
1,393
Long ("Type 1") or Short ("Type 2") 16 and 20MHz planars, or later -Axx (386DX-25) and -Bxx (486DX-25 and 33MHz)? I wouldn't be surprised if it was early on, despite the year of your notes. The initial PS/2 55SX planars and risers were known for their errata too.
 

IBMMuseum

Veteran Member
Joined
Aug 28, 2006
Messages
1,393
U24 "next to the CPU" sounds like the shorter "Type 2" planar (where the CPU is at U20). The long "Type 1" planar had the 386DX at U143. And on the -Axx 25MHz planar, the CPU was on a daughtercard with the NPU socket, cache, and 80385 cache controller.

EDIT: U24 on the -Axx 386DX daughtercard is an oscillator for the CPU, so my declaration stands...
 

modem7

Veteran Member
Joined
May 29, 2006
Messages
7,785
Location
Melbourne, Australia
That's all the info I have on the subject - something written into my notes. 1992 would have been when we were contacted by IBM about the issue. This was IBM approaching us (a very large and important customer) about the problem, not initiated by communications in the other direction. Maybe the "Tiawonese made chip" took a while to deteriate.
 
Top