That'll be fantastic! Thank you.

Evert wrote:

Care to elaborate on these other mods? I've never recapped a power supply in my life.

That kind of PSU would be an excellent one to learn on.

Is there anything I can do to protect myself from the large capacitors? Apart from not being stupid?

keropi wrote:

^ see this PDF that is the schematic this thing is made of: http://www.mikrocontroller.net/attachment/164 … ter_eng104k.pdf , at page 40 , chapter 5.3.4 it goes into detail about how it measures ESR . I can't tell if it's a good way or not.

Near the bottom of page 40 he says the test signal is "less than 680 Hz" which is both vague and a weird frequency to choose.
I would have to say that the ESR function of that meter is not too useful for checking caps against datasheets.
It's still useful for ESR comparisons (like cap A verses cap B) or to find caps that are 100% failed.

I don't want to talk down the whole meter though. The other things it does make it a pretty good meter to have.

ESR is a tough one to pack into a multi-meter device because it must generate a very low voltage very stable AC of some kind (often a square wave) for a test signal.
It's even tougher if you want a 100kHz test signal.
BK Precision has a multi-meter that can do ESR at 100kHz but last I looked it cost something over $800.
(They do have some that do ESR but that can not do it at 100kHz that cost less.)
It makes more sense to me (and my wallet) to get a dedicated ESR meter (ESR only meter) that uses a 100kHz test signal.
As new those typically cost $75-$125. Some are in kit form and some are pre-built.

I did (on eBay) see a Chinese one not to long ago that cost about 1/2 what those do and did have 100kHz signal.
They were advertizing it as suitable for in-circuit testing but (as per a review I saw on YouTube) it's test voltage was a wee bit too high for that.
It would be fine for out of circuit testing.

I would go find it and link but I'm getting behind. Maybe later I can do that.
.

Evert just let the psu discharge, I don't think you'll have a problem

I also uncovered the ICs under the heatsink on the jamma psu:

Regarding the meter, it's a simple and cheapish one - not the cheapest but nowhere near the price of the good 800+$ ones... more like 1/20th of the price 🤣 it's a simple tool to have around - especially handy with resistors 🤣
I am keeping the original 220uf/400v cap btw 😎

This is what I mean about the the Ripple multiplier for Frequency.
(And why 680 Hz is a weird choice for a meter. It doesn't match up with any normally used multiplier frequencies.)
The meter is not why I brought it up though.
You need to use this when comparing caps with Ripple rated at 100kHz to those rated at 120Hz.

Also note that Ripple for 105°C caps is not directly comparable to Ripple for 85°C caps in datasheets.
It can be done with math but I'm usually too lazy.
A value for Ripple rated at 105°C will be higher (better) at 85°C so I just use 105°C caps whenever possible and skip the math.
.

Thanks for the extra explanation, it's good to have an idea of how things work. So ESR on my meter is a bogus value not to be taken seriously...
Any advice on the simple jamma psu? Should I use my chemicons or order FCs along with the monitor caps? 😀

Evert wrote:

Is there anything I can do to protect myself from the large capacitors? Apart from not being stupid?

It depends if you want the white smock & pocket protector answer or the old jeans & beer belly answer.

With PC PSUs the vast majority of the time you don't need to do anything particularly special.
There are discharge paths inherent in the design so simply being disconnected from AC for 5 or 10 minutes is more than enough.
IT IS STILL BEST TO CHECK FOR VOLTAGE WITH A METER BEFORE YOU TOUCH THE CAPS. - Because....
In rare cases something in the discharge path is broken so the cap(s) can't discharge.
To the best of my recollection I've only personally seen that happen once in 20 years. (Boy do I remember it though!)

So: First check the caps for voltage. - Always.
About the time you get lazy and don't check one will bite you.

If you find a charged cap you need to find a way to discharge it.
How much voltage will matter for what you do about it. (Another reason to measure it.)
- Sometimes time is all you need. Leave it sit for a day or three and check the voltage again.
- A common safe way is to use the voltage and ohms law to find a resistor that will discharge it slow enough to not fry the resistor - based on the resistor's ohm value and wattage rating. Any conductor (wire) you use also has to handle the amps. - And ideally wear insulated gloves when making the connections. - The configuration is simple. Two gator clips connected by wire with a resistor in the middle.
- The happy-go-lucky red-neck way (that most people actually do but won't admit it in public) is to find a big honkin' screwdriver or wrench and short it whilst hoping the cap doesn't explode. If you have ever accidentally shorted something in a car you already know that 12v is A LOT when a dead short is involved. Chunks can blow out of tools and fried terminals (cap leads) is a distinct possibility.

A BIGGER MATTER. - Bigger because you are more likely to see it.
The heatsinks in PSUs are very often energized when the unit has AC applied.
Those near the AC end can have almost 400 volts on them depending on the design.
At some point you may have to do a voltage check or something that requires powering it up with the cover off (or disassembled on the workbench).
If and when that ever comes up remember to stay clear of the heatsinks. It's not just the heat 'hot' you need to worry about.
.

keropi wrote:

So ESR on my meter is a bogus value not to be taken seriously...

Just no good for checking against datasheet info.
It's not useless for ESR.
If a cap is supposed to be 0.02 ohms and the meter shows 0.80 the cap is obviously bad.
And at 680Hz a cap's ESR will be higher than at 100kHz so if it's okay at 680Hz then it's okay at 100kHz.
And you can tell if cap A is better or worse than cap B or not.

I wouldn't dis that meter for having one weak point.
It looks like a great meter to have for other things (like capacitance measurements).

keropi wrote:

Any advice on the simple jamma psu?

I'm trying to work back to that. You are much faster than me today.
.

I only recapped one, an old Seasonic SS-300FS, and it was the easiest soldering job i have ever done. These Seasonics are really well designed and have -5V line.

PCBONEZ wrote:

[...]
I'm trying to work back to that. You are much faster than me today.

ah sorry, didn't mean to sound rude or anything, take your time - whenever you can. I thought that either I found all by myself or the matter was buried 🤣 🤣

keropi wrote:

PCBONEZ wrote:

[...]
I'm trying to work back to that. You are much faster than me today.

ah sorry, didn't mean to sound rude or anything, take your time - whenever you can. I thought that either I found all by myself or the matter was buried 🤣 🤣

Not to worry. I did not take it as rude.

Just letting you know I'm spread a little thin ATM but I haven't forgotten.

Did you resolve the issue of finding the bipolar cap?
.

Yeah I know what you mean ... Some days are just crazy...
Don't worry about the bipolar cap - I'll put it on backorder on rs and get it with the next batch of caps 😀

keropi wrote:

....and here is something PSU related, a JAMMA psu 🤣

I had to search JAMMA because I didn't know what it was. 🤣 🤣

keropi wrote:

It has a mix of (what I assume) crap caps and Panasonic SU series:

That looks to me to be exactly what is going on there.

The first thing to do is figure out what grade (so far as ESR) the original caps are and so far that's been a problem.
That's why this is taking so long.
When this can't be determined it's better to default up to better grades of caps.

"ESR grades" are my own concept. A way to compare one brand/series to another.
There is no Industry Standard that defines ESR grades.
I discussed it in some detail twice here in the last couple of weeks and I'd rather not write it all again.
If I can find it again I'll put a link to it.

Some companies try to use terms like "ultra low ESR" or "very low ESR" to distinguish grades but there is no accepted Standard for those terms either so company A's "low ESR" might be exactly equivalent to company B's "ultra low ESR".

The only I've found so far are MIEC and you didn't mention the series so I dunno which to look at.
http://www.miec-cap.com/products/index.php?gr … 8&title_id=1966
[[Edit]] Actually it looks like they don't publish datasheets anyway. What a crap company!!

Searching 'giant' capacitors of course brings up all sorts or useless things (big, huge, large).
And so far Elgen appears to be defunct. (Or just no website.)

As to the SU caps.
One option is to for now just replace the crap brand caps then if there are still problems replace the SU's also.
Kind of depends on how fast you want the part back in service.

I am going to look a little more for Elgen.
.

thanks again for looking into this 😀

the MIEC caps don't have any series on them, here is the only text they have printed:

I do believe at this point that they put Panasonics in crucial parts and whatever crap they could find for other parts... AFAIK the giant/elgen caps are filtering ones (who cares if they are crap, jamma gameboards also have filtering on them as well so lets save some $ 🤣 ) and the MIEC ones don't get stressed...

The PSU in question was working fine - all this is done in advance so there are no failures in the future. It powers mainly 680x0 gameboards of the 90s, 5v "heavy" stuff.

Regarding ESR, it's all very interesting info. I don't fully understand it yet that's why I keep coming to this thread for advice. I fear that I will use a "too good" capacitor for an old board and it will cause problems like oscillation that was mentioned before... 😒

started measuring the jamma psu caps for fun this morning...

MIEC 220uf caps measure ~195uf
ELGEN 3300uf measures ~3230uf
GIANT 1000uf measures 726uf
Panasonic 2200uf measures 1977uf
Panasonic 470uf measures 408uf
Panasonic 100uf measures 91uf

more or less all are within the 20% tolerance - the thing is that most lean towards the outer extreme limit with less uf

Panasonic caps can stay

keropi wrote:

started measuring the jamma psu caps for fun this morning... […]
Show full quote

started measuring the jamma psu caps for fun this morning...

MIEC 220uf caps measure ~195uf
ELGEN 3300uf measures ~3230uf
GIANT 1000uf measures 726uf
Panasonic 2200uf measures 1977uf
Panasonic 470uf measures 408uf
Panasonic 100uf measures 91uf

more or less all are within the 20% tolerance - the thing is that most lean towards the outer extreme limit with less uf

How long since the board has had power to it?

^ 10~12 days , is it enough to discharge?

keropi wrote:

^ 10~12 days , is it enough to discharge?

Yes but that's not what I was thinking.

It would be normal for capacitance to read low on parts that have been stored long term.

Once it's had power to it for, oh, 30 minutes to an hour the capacitance(s) should come back to their actual values.
That is part of why they used to refer to Aluminum Lytic caps as "Self Healing" caps.
Since it's had power the low values are their actual values.

Since these are still on the low side after having power and since I know one of your big concerns is long life after the recap I think maybe you should replace all of them.

Realistically (except for the Giant) you probably don't -need- to replace any before 5 more years. - But that doesn't fit with your stated objective.
.