Sinclair reliability

[angus]

In the last two months I have "killed" four modern computers.

My Samsung P510 laptop - screen has died.
My new (August 2012) quad core desktop PC - memory fried
My work desktop PC - PSU died
My work webmate - memory errors.


Although my work PC has been fixed due to the excellent and rapid efforts of our on-site IT team (who have just been made redundant...grrr) there was a period of three days where the only working computers I had access to were my two Zeddies and my Speccy. 30+ years old and still going strong.

I know much was said about the build quality of Clive's creations, but I have never had any serious issues with Sinclair computer reliability.

[gozzo]

The 'old stuff' definitely seems to be more reliable than modern things...I have many old computers, including some not-so old(mid 1990's PC's and laptops) that are good condition and fully working, and my 'modern' (now 6 years old,though) Dell laptop seem to need re-installs every so often or else it plays up and starts getting erratic! Plus of course there was(still is??) that problem with substandard/counterfeit electrolytic capacitors used in a lot of equipment including PC's! Seems strange that most of the electrolytic capacitors I've replaced have been these high temperature 105c 'long-life' things, and the bog standard 85c ones rarely seem to give trouble...!

[gozzo]

Angus - if you computer is 'frying' your memory there must be something wrong with whatever regulates its supply voltage, either something on the motherboard or the PSU itself. I had a PC that its processor would get hot enough to fry eggs on, although otherwise working OK,all voltages checked OK on a test meter,I saw there was a capacitor that had a bulge in it and when this was replaced ran nice and cool, the faulty capacitor was probably causing HF ripple on the processors supply line! I still have another motherboard to service when I get round to it....bulging capacitors, again!!

[RetroTechie]

In a way, much multi-decade old electronics (read: those that survived so far!) are like the military-spec parts of today. Over-engineered, inefficient, clunky, but more likely to survive 'smaller' events like an electrostatic discharge, partly degraded materials etc. Like a steam engine that still runs after a layer of rust is removed and the parts have been cleaned & greased up.

Seems strange that most of the electrolytic capacitors I've replaced have been these high temperature 105c 'long-life' things, and the bog standard 85c ones rarely seem to give trouble...!

Labels may not say much anymore, and a lot depends on where you're looking. I suppose most places where you'd find a 105c long-life cap, are places that get hot, and where those caps have to work hard. For example a modern (switching) psu, versus the linear supply in most 80's home computers. Read: places where a capacitor is more likely to die anyway. And/or you've mostly been sticking your nose in cheapo China crap lately...

Angus - if you computer is 'frying' your memory there must be something wrong with whatever regulates its supply voltage, either something on the motherboard or the PSU itself.

Granted, power supply is a biggie. But don't forget cooling (read: airflow). Which is an art of sorts, just sticking in an extra fan doesn't always improve the situation. And with a few RAM sticks next to each other, there's good potential for slow-moving air in between those RAM sticks.

[gozzo]

I suppose most places where you'd find a 105c long-life cap, are places that get hot, and where those caps have to work hard. For example a modern (switching) psu, versus the linear supply in most 80's home computers.

... not all the duff caps were in warm places or passing high ripple current..(probably some of those dodgy ones!)....

And/or you've mostly been sticking your nose in cheapo China crap lately...

....some...(yuck).. ......but not all... As for Sinclair reliability..pretty good, apart from the early ZX81 ULA, as all the duff ones I've so far come across have been....and problems with the oscillator transistor in a few spectrums destroying themselves within a few seconds (still not yet sorted them out..I have suspicions that one or more 4116 RAM chips are drawing heavy current..damn things!) - will need to make a test rig using external -5v, +12v and +9v so I can power the thing up without killing all my limited supply of ZTX 651's...

[1024MAK]

I have suspicions that one or more 4116 RAM chips are drawing heavy current..damn things!) - will need to make a test rig using external -5v, +12v and +9v so I can power the thing up without killing all my limited supply of ZTX 651's...

Ahh, if you have a suitable PC PSU or even a Sinclair +3 PSU, plus 7809 and 7905 voltage regulators, you should be sorted

[gozzo]

That's what I was thinking of , a PC PSU, (plus a 7812 with a few series diodes for the 9v, could use a 7809 but means buying one, I have a few 7812's lurking around!), got a good few, but most are at the back of my barn and impossible to get at , at the moment, I have 2 nearby but they are faulty awaiting possible repair....and the faults..guess what.....bulging/leaking capacitors!!!

[PokeMon]

Today the big companies have obsolescence managers which keep there eyes on a planned limited lifetime of products. Really crazy but true.

[gozzo]

Hmm, yeah, like a Ferguson video recorder I tried to repair some years back, it was only a few months out of guarantee so it couldn't be done for free, and as it was a French Thomson made model, the part had to come from them...no luck..for several years I kept trying, the part was supposedly constantly 'out of stock, awaiting supplies'...eventually found another supplier that listed them as in stock, but no, 'very limited supply'....eventually the poor thing had to be dumped!
It was said back then that the estimated average life of a TV was about 7 years and a video 5 at the most . More recently I had a combined VCR/DVD thing that suddenly made a clunking sound and although it seemed to still work, would screw tapes up randomly when first inserting them - just out of guarantee again - the insides were horrible, everything had to be taken out to get to the mechanism, and even then it was impossible to separate it from the PCB without desoldering the delicate flat flexicables between the two - they didn't use removable connectors..I didn't dare do it in case I couldn't reconnect them...checked it as best I could but didn't find anything obviously wrong...had to abandon it and get another... These days your lucky if it lasts out the guarantee period!!

[1024MAK]

If I don't have a 7809 regulator handy, I use a 7805 with a BZX55C4V3 zener diode in the common leg.

Mark