ZX81 memory troubles
-
- Posts: 41
- Joined: Sun Jul 08, 2012 11:53 pm
- Location: Moscow, Russia
Re: ZX81 memory troubles
gozzo, i alreay consider this zeddy fully working until rampack is plugged, after that it takes time to recover and again is fully working, well, 95% of the time i guess.
i have already checked power ripple down to a few nanoseconds/div with my digital scope. Though, i am limited to 150mhz sampling and 60mhz bandwidth.
I have check the clock already and posted the picture of it. Is there anything else needs checking with the clock buffer or it can be considered okay?
i have already checked power ripple down to a few nanoseconds/div with my digital scope. Though, i am limited to 150mhz sampling and 60mhz bandwidth.
I have check the clock already and posted the picture of it. Is there anything else needs checking with the clock buffer or it can be considered okay?
Artem
-
- Posts: 41
- Joined: Sun Jul 08, 2012 11:53 pm
- Location: Moscow, Russia
Re: ZX81 memory troubles
Ladies (if any) and Gentlemen!
Behold: Da phaquing shit!
Literally.
So, before going to bed before tough monday i decided to torture my poor zeddy once again so i can gain a lot more data on the problem with my o-scope.
Basically, i decided to replug a ram pack to provoke the problem and then check all signals.
So, before doing it i've noticed that edge connector is kind of dirty. So, i took my bottle of IPA and cleaned it a bit with a paper towel moitened in ipa.
Plugged a ram pack (which is presumable a working one) and turn it on. IT worked! K cursors, peek 16387 = 127.
But since i needed the glitch, i plugged another one one (which is supposed to be dead), but it WORK! K cursor, peek 16387 = 127. I turned on and off, switched the packs several times. It just works. If you think it was dirty connector - don't be so sure.
So, i reassembled the computer, screwed the case and tested again. Different packs, several times on/off, and got a glitch again! However, no noise this time, only white screen. It restored to normal operation very quickly, in about 2 minutes again K cursor. I tried again with a pack and after about 5 tried sent it into a knock out again. Then i decided, whatta heck, if IPA made it better maybe the cause of all this is the dirt!
So, i dissembled it again, picked all ICs and replugged them, cleaned the whole PCB with a antistatic brush with IPA, and, just for kicks, desoldered +5V and signal connection from TV modulator. Assembled the case again.
Since then i tried like 15 times to repeat the problem but i cannot. It just works.
So, let it sit for a day and i will try to make it go crazy again.
Behold: Da phaquing shit!
Literally.
So, before going to bed before tough monday i decided to torture my poor zeddy once again so i can gain a lot more data on the problem with my o-scope.
Basically, i decided to replug a ram pack to provoke the problem and then check all signals.
So, before doing it i've noticed that edge connector is kind of dirty. So, i took my bottle of IPA and cleaned it a bit with a paper towel moitened in ipa.
Plugged a ram pack (which is presumable a working one) and turn it on. IT worked! K cursors, peek 16387 = 127.
But since i needed the glitch, i plugged another one one (which is supposed to be dead), but it WORK! K cursor, peek 16387 = 127. I turned on and off, switched the packs several times. It just works. If you think it was dirty connector - don't be so sure.
So, i reassembled the computer, screwed the case and tested again. Different packs, several times on/off, and got a glitch again! However, no noise this time, only white screen. It restored to normal operation very quickly, in about 2 minutes again K cursor. I tried again with a pack and after about 5 tried sent it into a knock out again. Then i decided, whatta heck, if IPA made it better maybe the cause of all this is the dirt!
So, i dissembled it again, picked all ICs and replugged them, cleaned the whole PCB with a antistatic brush with IPA, and, just for kicks, desoldered +5V and signal connection from TV modulator. Assembled the case again.
Since then i tried like 15 times to repeat the problem but i cannot. It just works.
So, let it sit for a day and i will try to make it go crazy again.
Last edited by ArtemKuchin on Mon Aug 13, 2012 7:25 am, edited 1 time in total.
Artem
- 1024MAK
- Posts: 5534
- Joined: Mon Sep 26, 2011 10:56 am
- Location: Looking forward to summer in Somerset, UK...
- Contact:
Re: ZX81 memory troubles
Well there you go, it just shows what a little partly conductive dirt can do...
I have heard of, and seen what leakage from electrolytic capacitors and battery leakage can do (always check and if needed, clean under such components). "Waste products" from animals also causes no end of problems (don't ask!).
Glad that you are making progress
Mark




I have heard of, and seen what leakage from electrolytic capacitors and battery leakage can do (always check and if needed, clean under such components). "Waste products" from animals also causes no end of problems (don't ask!).
Glad that you are making progress

Mark
ZX81 Variations
ZX81 Chip Pin-outs
ZX81 Video Transistor Amp
Standby alert 
There are four lights!
Step up to red alert. Sir, are you absolutely sure? It does mean changing the bulb
Spring approaching...
ZX81 Chip Pin-outs
ZX81 Video Transistor Amp


There are four lights!
Step up to red alert. Sir, are you absolutely sure? It does mean changing the bulb

Spring approaching...
Re: ZX81 memory troubles
You lucky b****rd !!! I wish I could cure my dodgy ZX's just by giving them a good cleanup!!! 

Re: ZX81 memory troubles
'Waste products from animals...." ?! Possibly worse is semi-dried "waste products" from a young child in a sega game console...YEUUCCHH!!
Had to clean one out years ago! Not pleasant ! Can't remember if it worked or not !

Last edited by gozzo on Mon Aug 13, 2012 2:54 am, edited 2 times in total.
Re: ZX81 memory troubles
ps. "conductive dust" used to be a radio engineers favourite reason/cause (apart from "it's a valve gone") of trouble in radios in the 1930's...!!!
- RetroTechie
- Posts: 379
- Joined: Tue Nov 01, 2011 12:16 am
- Location: Hengelo, NL
- Contact:
Re: ZX81 memory troubles
I'd put my money on that... cleaning the edge connector may have helped, cleaning the board did nothing, chances are it was corrosion / oxidation on IC pin(s) and/or socket contact(s). This is why turned pin IC sockets are preferred for stuff that's meant to last. Not that same thing won't happen with those, just a *lot* less likely. Of course downside is the cost. No socket is best (& cheapest!) but of course makes diagnosis & replacement more difficult.ArtemKuchin wrote:(..) picked all ICs and replugged them (..)
Note that there are low cost IC sockets that touch IC pins on 1 side, and some that touch IC pins on both sides. Also there may be differences in plating materials, layer thickness, mechanical construction / accuracy etc (same for IC pins & turned pin sockets).
-
- Posts: 41
- Joined: Sun Jul 08, 2012 11:53 pm
- Location: Moscow, Russia
- RetroTechie
- Posts: 379
- Joined: Tue Nov 01, 2011 12:16 am
- Location: Hengelo, NL
- Contact:
Re: ZX81 memory troubles
Those with circular, machined contacts (usually gold plated), like: http://media.digikey.com/photos/Mill-Ma ... 001000.jpg
As opposed to the low cost ones with flat contacts, touching IC pins only from the side (usually nickel plated), like: http://media.digikey.com/photos/3M%20Ph ... 000-CP.jpg
The turned pin variety is often fabricated to tighter mechanical tolerances, and it's easier to solder to a socket pin on component side of the board. With low cost IC sockets that is very difficult.
As opposed to the low cost ones with flat contacts, touching IC pins only from the side (usually nickel plated), like: http://media.digikey.com/photos/3M%20Ph ... 000-CP.jpg
The turned pin variety is often fabricated to tighter mechanical tolerances, and it's easier to solder to a socket pin on component side of the board. With low cost IC sockets that is very difficult.
-
- Posts: 41
- Joined: Sun Jul 08, 2012 11:53 pm
- Location: Moscow, Russia
Re: ZX81 memory troubles
ahh. aka the zero force socket
well, i think they have one drawback - the contact area with flat contacts socket is a lot more.
well, i think they have one drawback - the contact area with flat contacts socket is a lot more.
Artem