advertisement


Wear and tear on amplifiers

richards13 said:
Rebuilding my valve amp recently it was great to have a choice of 105c 7000hour rated electrolytic capacitors. Solar panel and data centre usage was the rational mentioned in the components application notes. Anyone know the life of electrolytic capacitors unloaded at room temperature?
Click to expand...
This one says 16 000 hours at 85c :

https://www.partsconnexion.com/media/pdfs/mundorf_ag.pdf

So I would guess 20 years if left on a shelf uncharged ?
 
It's audio grade so lifetime if left on shelf probably about 5yrs then someone will get twitchy and bin it I case it affects sound quality.....

but in reality......
 
My own IT experience was relatively tangential, but in an HE environment, older equipment was retained because it 'did the job', particularly when the job in question was mission-critical (eg payroll, student admissions, timetabling). The very last thing the bods in charge of such systems wanted was something shiny and new and unreliable.

See also the Post Office for an example of a shiny new system that didn't actually work properly.
 
H20GNA said:
Womack, Jones and Roos
Click to expand...

Lean? Surely you don't expect Defense manufacturers to actually care about the costs etc? (Seriously though we did run Six Sigma but in reality it was about reducing cost rather than improving quality).
 
H20GNA said:
You would be surprised many enterprises still use very old equipment, for the reasons you state. Banks tend to be the worst historically, as they refuse to pay the cost of model office.
Click to expand...

The very last IT contract I had was at a very posh trading bank in That London. The type of place that trades £millions a day, dealers with multiple screens for up to date market information etc. There was an absolutely mission-critical ancient DOS machine in the server room, a real crusty old thing, maybe a Compaq 386 I can’t remember, but if that failed the whole system wouldn’t come up. I wasn’t the IT manager so I just pointed and laughed at it, but if it was my server room that thing would have had multiple levels of redundancy!

PS This was 2002!
 
It'll still be running!

What on earth was on it? I here that Cobol is now a black art and nobody touches those in case every bank system in the UK falls over.
 
Heat being a problem seems logical but my Sugden A21SE is 18 years old, before that I had an A21a and before that I had an A48II bought in 1978. I have never had anything go pop other than a volume pot and when, after reading stories of recapping Naims etc. after 10 years, I changed my 65º output caps for some panasonics at 105º I was rebuked by Patrick and his dad for changing "summat that wasn´t broke". So not so sure.
 
JimmyB said:
I here that Cobol is now a black art and nobody touches those in case every bank system in the UK falls over.
Click to expand...

I used to teach people how to program in COBOL. I suspect, if I wanted to, I could make a pretty ridiculous salary or contract rate these days as I’m really young for a COBOL programmer! It is a 1960s language. Not many of us still about!

PS I can actually remember most of it too, unlike all the others (Basic, C, Pascal, dBase, PL1 etc) which have all blurred together to the point I can’t recall which command or syntax belongs to which language. COBOL is unlike anything else and as such quite memorable.
 
Joe Hutch said:
My own IT experience was relatively tangential, but in an HE environment, older equipment was retained because it 'did the job', particularly when the job in question was mission-critical (eg payroll, student admissions, timetabling). The very last thing the bods in charge of such systems wanted was something shiny and new and unreliable.

See also the Post Office for an example of a shiny new system that didn't actually work properly.
Click to expand...

The shiny new equipment worked perfectly, the software was at fault, and the designers knew it and lied in court, The Post Office believed the software developers instead of a huge number of Postmasters

Gary
 
JimmyB said:
Lean? Surely you don't expect Defense manufacturers to actually care about the costs etc? (Seriously though we did run Six Sigma but in reality it was about reducing cost rather than improving quality).
Click to expand...

I was at MIT same time as these guys.

Gary
 
mercalia said:
so what did you do? try some switch cleaner on it?
Click to expand...

No delivered to Firebottle mid morning on the day after it came, he returned it working mid afternoon, what service, we are friends though :). I think it was a poorly soldered joint that had corroded.

Gary
 
Gervais Cote said:
This one says 16 000 hours at 85c :

https://www.partsconnexion.com/media/pdfs/mundorf_ag.pdf

So I would guess 20 years if left on a shelf uncharged ?
Click to expand...
Nope! its a bit more complex than that. I did a back-of-fag-packet calculation for those caps working at their full rated voltage and if you keep the ambient temp at around 30 degrees (instead of 85) the expected life is over 80 years. Temperature really is the killer of electrolytic capacitors.

Electrolytic caps rely on chemistry to work and the rate of chemical reactions was discovered by Arrhenius. Basically the rate of a chemical reaction doubles for each 10 degree rise in temp so even without doing any more big complicated sums we can determine that if life at 30 degrees is 80 years then at 40 its 40 years at 50 its 20 years at 60 its 10 years - get the picture?

Cheers,

DV
 
IME commercial servers/farms are generally on 24/7 ... but not for reasons of hardware longevity.

Is the whole Internet turned off? Again no, but hardware longevity is not the reason.
 
Things are often left on 24/7 in circumstances where they are "mission critical" etc. IE longevity is not the requirement but doing everything possible to make sure nothing goes wrong within say the next 6 months is the requirement....even if it means everything is worn out and must be replaced in a fairly short time.
 
In my experience, which is not much. I found out it was the inrush current into cap from zero volts, the first part of the curve is sharp so on the M110,s WB, M50 and the new M40A? they are all pre charged to about 0.1 or 10% of there working voltage, also noted the caps with low ESR draw more current on start up and after a few failures, discovered the ultra sonic aluminium welds were damaged to the terminal, this reduced there current operation and increased the current into those remaining welded. So the failure got worse until nothing of use, this also cause small hot spots and destroyed the dielectric chemical, this caused leaks, so bye bye cap. Now what I have found with soft pre-charge is the cap last up to 10,000Hrs longer, and this was with a 50V PSU at 100Amps.
Next I tried it o SMPS i.e. soft start extreme, life went on for those as well.

Conclusion:- On all the time not good as they heat up life is shortened, switch on with massive inrush, not good. Pre-charge better so I go for pre-charge if possible followed by soft start, as we do, or use crap diode with a Vf of 2V plus or so he he.

Have fun guys.

Just to change the subject I had my first day out this week in over a year .well out side a pub with my step son, it was great after being locked in for a year.

The Redheaded waitress was good but the IPA and grub was amazing. The Barge Inn.
 
You must log in or register to reply here.


advertisement


Back
Top