Console PSU recap - What about those tantalums?

[zamproject]

I have been servicing a vintage Neve console for 11 years now, and have never had a single issue with tants in signal path, in my opinion they are part of the neve sound, which you may or may not enjoy. But absolutely tants don't belong in PSUs, only in signal path in my experience. Electrolytics for everything else.

I'll say exactly the same with my 70' Studer console, mostly daily use and self serviced since years.
Thousand tans (in fact everything up to 100u...) I barely replace 1 or 2 per year.
BUT replaced with alu for anything related to power supply (PSU, or modules onboard regulations)

Also servicing sometime same era/design consoles (089) for pple
One of them have been completely recapped with alu (not by me), I service the console to fix faulty modules as full alignment/calibration, I run REW on everything after that to check all is ok, and it was bad regarding frequency response (like -3@ 60Hz...) for all channels, so maybe it was on masters or line out cards.
I have no time left to investigate further before the owner take back the console, I rapidly check caps value, see nothing bad or error...so still frustrated and puzzled about this since...

 

[Dave Obergoenner]

OH, and that lovely aroma when they short across the PS rails!!!
I only leave them in for timing circuits. They do maintain high accuracy for that, and nearly never fail.
Never never never leave them in across the PS rails. Never ever.
Dave O.

 

[Whoops]

Yeah - but I have (literally) seen relatively modern electrolytics in PSU type apps burst at random for no good reason on occasion.

I have been using Panasonic Electrolytic capacitors rated at 105degres for 16 years in hundreds of recap jobs and never ever had 1 one them fail.
My colleague that does this professionally and even does much more recaps than me also uses the same capacitors and never had 1 fail also.

Lesson is, only buy a good quality Brand

 

[Whoops]

Have in mind also that the by using Tantalum you are also supporting wars and deaths








https://www.mining-technology.com/features/featureon-the-trail-of-tantalum-4831288/
https://www.dissentmagazine.org/online_articles/the-problem-with-conflict-minerals/

 

[JohnRoberts]

I did try the nulls with a music source as well and got marginal very thin HF pass of audio in both cases at very low level.

'That sounds like a discrepancy between the two channel's high frequency roll off. Not clear how an electrolytic DC blocking cap would impact that?

Did the two channels deliver a deep null before changing the capacitors in one channel? After that result I'd be tempted to put new caps in a second channel and see if they null better after that.

JR

 

[metalockpick]

Back to Electrolytic capacitors-
105 degrees C is a minimum high temp spec, and the low temp spec should be at least -40. The cold spec is so when the product the capacitors are mounted in is shipped by air to the northern hemisphere in January (or the southern hemisphere in July) in an unpressurized, unheated cargo hold it won't fail as soon as it is plugged in.
Something I found important is to not buy capacitors made in China. This is because air pollution will deposit acids on the foil sections of the capacitor during construction of the capacitor. The acid tends to slowly eat away at foil (and the capacitance), the capacitance value dropping more than 50% in 3-8 years.
Clean rooms are apparently not being used in China for capacitor manufacture.
This was first noticed in the computer industry, where quality motherboards are made in Taiwan and advertise "all Japanese capacitors".
This problem seems to be confined to capacitors made in China, (not Japan, Taiwan, Vietnam, etc).

 

[RoadrunnerOZ]

'That sounds like a discrepancy between the two channel's high frequency roll off. Not clear how an electrolytic DC blocking cap would impact that?

Did the two channels deliver a deep null before changing the capacitors in one channel? After that result I'd be tempted to put new caps in a second channel and see if they null better after that.

JR

Actually they didn’t. I think it’s due to minor gain difference on the switched input gain pot when testing just the line module solo.
But with the faders in circuit when you achieved max null there was very little breakthrough but tiniest breakthrough there was - before recapping more than after with a full range sweep or music. Bit of bottom end breakthrough with uncapped vs recapped maybe due to dried out electrolytics. Very slight wispy top end stuff after recap which made me wonder what else was out of whack - maybe difference in frequency response of IC’s, transistors or the mosfet direct out stage - we ran it through that, pre and post EQ rather than through the output transformers. There may have been some filtering caps for HF control but I can’t remember now, also we ran it through the EQ set to null and in bypass - not much change there in either mode.

 

[JohnRoberts]

Actually they didn’t. I think it’s due to minor gain difference on the switched input gain pot when testing just the line module solo.

a gain difference would result in a wide band null product output.

But with the faders in circuit when you achieved max null there was very little breakthrough but tiniest breakthrough there was - before recapping more than after with a full range sweep or music.

Not sure what breakthrough is, perhaps one of the two strips clipping at some internal stage might sound like that.

[edit- current limiting or spurious instability might be characterized as "breakthrough". /edit]

JR

Bit of bottom end breakthrough with uncapped vs recapped maybe due to dried out electrolytics. Very slight wispy top end stuff after recap which made me wonder what else was out of whack - maybe difference in frequency response of IC’s, transistors or the mosfet direct out stage - we ran it through that, pre and post EQ rather than through the output transformers. There may have been some filtering caps for HF control but I can’t remember now, also we ran it through the EQ set to null and in bypass - not much change there in either mode.

 

[RoadrunnerOZ]

a gain difference would result in a wide band null product output.

Not sure what breakthrough is, perhaps one of the two strips clipping at some internal stage might sound like that.

JR

Just to clarify - with the line modules only, no fader, only switchpot input levels, the difference was broadband but low level - with full channel card, faders in place trimmed to maximise null there was the HF difference.

 

[Mr. Moscode]

What is the most used value/voltage for interstage coupling caps in the boards we're talking about?
I'm guessing 10mfd.

 

[living sounds]

Crosstalk in a console is usually so high that null tests are of limited usefullness there in my experience.

 

[RoadrunnerOZ]

a gain difference would result in a wide band null product output.

Not sure what breakthrough is, perhaps one of the two strips clipping at some internal stage might sound like that.

[edit- current limiting or spurious instability might be characterized as "breakthrough". /edit]

JR

Could also have been channel crosstalk - either backplane cards or all cabling coming to connector blocks in the centre section of the console - we did have the monitor gain really cranked for null tests. Ran out of time to really get to the bottom of it. Client didn’t want to re-chip the whole console either - more than it would be worth.

 

[Dave Obergoenner]

Back to Electrolytic capacitors-
105 degrees C is a minimum high temp spec, and the low temp spec should be at least -40. The cold spec is so when the product the capacitors are mounted in is shipped by air to the northern hemisphere in January (or the southern hemisphere in July) in an unpressurized, unheated cargo hold it won't fail as soon as it is plugged in.
Something I found important is to not buy capacitors made in China. This is because air pollution will deposit acids on the foil sections of the capacitor during construction of the capacitor. The acid tends to slowly eat away at foil (and the capacitance), the capacitance value dropping more than 50% in 3-8 years.
Clean rooms are apparently not being used in China for capacitor manufacture.
This was first noticed in the computer industry, where quality motherboards are made in Taiwan and advertise "all Japanese capacitors".
This problem seems to be confined to capacitors made in China, (not Japan, Taiwan, Vietnam,

I have been using Panasonic Electrolytic capacitors rated at 105degres for 16 years in hundreds of recap jobs and never ever had 1 one them fail.
My colleague that does this professionally and even does much more recaps than me also uses the same capacitors and never had 1 fail also.

Lesson is, only buy a good quality Brand

I'd second the vote for Panasonic aluminum electrolytics. I've been installing them almost exclusively for at least 20 years now. Thousands, perhaps 10's of thousands of them by now. Usually the FM series, but if the value or size I need isn't available in the FM line, then it usually is in the FC line. Both are 105 deg, and low ESR. Very good quality. These are great sounding caps in any level of pro gear. And the price is reasonable. Lately supply chain issues have made some difficult to source however. 470uf @ 35v has been a problem in the FM line for Mouser to acquire. But that seems to be improving lately.

I've never had a single piece of gear that I've recapped with them ever come back to me with a failed cap...ever. I check each cap prior to installation for value and ESR, and I've never even taken one out of the package (usually from Mouser) that was not good.

Panasonic caps never fell victim to the "exploding cap syndrome" that was VERY common back about 20 years ago, and here's why... Panasonic has always made their own electrolytic paste, in-house. Usually in Japan. They never bought the (deliberately?) contaminated, but cheap paste from china inc. That's the reason I started using them thousands of caps replacements ago, and I've never been let down by them.

Buy good quality (Panasonic) from a reliable supplier (Mouser or Digikey) and you'll probably never need to recap again.

Dave O.

 

[FIX]

deep nulls are difficult because even tiny frequency or phase response errors will compromise the depth of null.

I would focus on small errors in the strip eqs to get the null deeper.

Listening to the null product will generally tell you if there is something nasty going on or not, probably not.

JR

After hanging with all the "engineers: in DC for way too long, we always had listening tests for everything... I was one of 3 or 4 that could hear and identify tape, els, mics etc. After getting board with that kind of comparison, as when doing two items, you start to hallucinate and hear satan's cry. So I employed John's trick, and you could null them enough to hear the differences, once you identified what it sounded like, you could then listed to each and identify which one had it. Today, it's still the best way to compare.

I also use that to match mono compressors when I want to use them in stereo. Set one where you want it, then feed both with the same signal and sum out of polarity. When doing that, you adjust the new one until it nulls. They will be matched.

 

[FIX]

My option of Tants. For audio they are good, though they sometimes short for no reason, maybe contamination when made, or something. In power supplies, they are very low ESR, BUT when they short, they burn like a bastard, and if not stopped, they WILL burn through everything all the way to China. They aren't worth it. You can "simulate" their excellent capabilities by putting an electrolytic cap in place of them that is between 10X and 100X the size, as it will mimic-ish the lower ESR.

 

[Whoops]

when they short, they burn like a bastard, and if not stopped, they WILL burn through everything all the way to China.

 

[RoadrunnerOZ]

Where there’s smoke…….