October 12, 2000
Seeing ever more 2400s bite the dust on the Duolist, I'll come in from shadows in the form of a generic reply to various death symptom problems posted on the Duolist, based on some empirical discoveries I've made over the years. I use a primary G3/320/112/30GB and a backup G3/240/80/20GB, both on OS8.6 (I'm definitely not a fan of OS9). Since I don't live in the Apple delimited PB2400 service areas of USA or Japan, I must have some way to keep at least one active. I have, on several occasions, thought one or the other of my 2400s was dead and beyond repair. Each time I've brought them back (so far).
Each 2400 has been apart numerous times. In fact I left my G3/240 in pieces on the shelf for more than 6 months before "fixing" it some 18 months ago, just by re-assembly including a totally drained PRAM battery (which cleared the corrupted Power Manager which repeated resets of every duration failed to resolve).
It's clear that the 2400 is an EXTREMELY sensitive machine. It takes very little to have a 2400 go into a "terminal" "green light of death" or "no light & click of death" or other troubling and increasingly terminal symptoms. Once, surfing with Explorer resulted in a freeze and the subsequent reboot led to increasing corruption until, eventually, I got the "green light of death". Even transporting my computer in a car in sleep mode has resulted in a "no light & click of death" situation--more than once. I'm puzzled as to why. Another time, just a routine reboot under AC power on my G3/320 resulted in ever deteriorating symptoms of terminal death, giving every sign of a power board failure, or worse, and quite different from the aforementioned Explorer problem (see below #3). Here's my latest on dealing with "death symptoms".
1. GREEN LIGHT (constantly lit)
2. NO LIGHT (hitting Power key results in faint click)
3. MOTHERBOARD SWAPS
I would contend that probably over 50% of the 2400s shipped off for repair of motherboards probably don't actually need a motherboard repair/replacement. The troublesome PM reset, unsuccessfully executed, seems terminal. A disassembly of the 2400 actually drains the PRAM battery and the PM is usually easily reset after that when out for service. For those of you whose boards have been returned as repaired, without procedural documentation even if requested, this is a probable explanation.
Now, because I've troubleshot my own 2400s, back to back, I want to relate one other curious but important anomaly which has happened twice (that I know of in Europe, verified independently). A motherboard swap must be done with caution, apart from the usual precautions. Each 2400 motherboard appears to be matched to its originally installed power board (possibly both were matched tech revisions, but I haven't looked that closely). I intentionally swapped my "freshest" components into my preferred case. After about 3 weeks, my G3/320-2400 "failed", terminally, or so it seemed. I pulled my backup G3/240 apart and put its MB into my 320. Everything worked fine. BUT, installing the "defective" motherboard back into the 240 also worked fine. Both, restored to their original boards, have been working fine for the past 6 months. Now, I know of another case where two 2400 users had the same experience in swapping motherboards for testing their 2400s. The swap worked only for a few weeks before "failing" whereupon a swap back "fixed" the problem. Go figure!
So, after countless "deaths" both of my 2400s actually have been working "fine" since new, but test me every so often with death symptoms. If I didn't live Switzerland, I, too, probably would have given up and sent my computer in for "repair" several times or unwittingly parts it out or parted with it. Necessity remains the mother of invention. YMMV.