48k Rubber Key Spectrum Problem

mrmessy · June 2013

I'm testing my mate's spectrum that hasn't been used since the late 80's. Unfortunately, it seems to keep randomly resetting itself (like pressing the reset button on a Spectrum +). Anyone have any ideas what might be likely to be wrong with it?

Pierre · June 2013

Try replacing all the electrolytic caps first. There are many topics in the forum describing how. The caps dry up due to age. If it starts up ok and then reset, it is normally a cap problem.

mrmessy · June 2013

I tried it again yesterday and it now seems to be fine. Is it possible the circuits were simply "playing up" because they had not been used for so many years, and now they have got that glitch out of the system? Is such a situation possible with old electronics?

roko · June 2013

mrmessy wrote: »

I tried it again yesterday and it now seems to be fine. Is it possible the circuits were simply "playing up" because they had not been used for so many years, and now they have got that glitch out of the system? Is such a situation possible with old electronics?

I noticed this phenomenon several times, and found for myself two possible explanations. Still I am not satisfied.

1. Bringing electronics from a cold attic into a warm living room may result in condensation on the circuits. This damp layer usually dissappears after a number of hours.
2. It might be possible that some kind of recuperation takes place as soon as a voltage is supplied to an electrolytic cap.
Details:
http://www.vcomp.co.uk/tech_tips/reform_caps/reform_caps.htm

Pierre · June 2013

Strangely enough I found that the lower 16K (4116 chips) sometimes need a few power-ups to start working if the unit was not switched on for a long time. When you switch on for the first time it has the same symptoms as with a faulty chip, which disappear after a few power-ups. I have not taken a close look at the design specs of the 4116 chips, but as the RAM matrix is just a number of elements that needs to be charged, it is quite possible that a few power-ups will get the chips to a working state. The same will be for dried up electrolytic caps, which will get up some charge after usage.

balford · June 2013

I'd vouch for this theory - I used 4164 chips from a long dead +2 Spectrum to upgrade a Issue 2 16K Spectrum to 48K.

For the first couple of power ups the machine would initialise as a 16K machine, even though any subsequent tests from basic didn't throw up any errors in upper RAM. After a few hours of repeated boots, the machine started powering up with RAMTOP where a 48K machine would have it.

Said machine has been reliably operating ever since. Weird, huh?

mrmessy · June 2013

On third session (5 mins) it went wonky again. I think I need to leave it on for a good while...

48k Rubber Key Spectrum Problem

Comments