What is value of Electrolytic Cap with legend 15/ on Vintage TAC Mixer

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

tomheavybeats

Well-known member
Joined
Dec 11, 2020
Messages
56
Hi, I bought a TAC 12:4 mixing desk recently and plan to re-cap as its from around 1976. It has a whole load of Electrolytic Caps with the legend 15/ and 40V. I looked up 15uf, 40VDC electrolytic caps online and they seem to be rare so I am wondering if I've got the value wrong. I have attached picture.
I am new here so apologies if I am asking in the wrong place. Google searching a forward slash doesnt seem to be allowed I think so thought I'd ask here. Thanks in advance!

 
15uF was quite common back in the day so although not usual now, it was usual when that cap was made.

Caps now are generally made in a much smaller range of values than 50 years ago.

Cheers

Ian
 
ruffrecords said:
15uF was quite common back in the day so although not usual now, it was usual when that cap was made.

Many thanks Ian

squarewave said:
It is virtually guaranteed that you can use 22uF there. If you post a schem, we can confirm.

I wish I had schematic  :) Would be great if I could put in 22uf. I asked Cemal at Total Audio Control and he got it confirmed from Graham Langley that this was an early custom build desk. I have attached some pictures, excited about the Sowter transformers but cannot find info on the 2300 series either.





 
jensenmann said:
Make sure to recap the tantalum caps, too. They degrade with age and turn into a resistor.

Yes thanks, just been reading about this and whether to replace with new tantalums but I think I will replace with Panasonic or Nichicon electrolytics
 
analogguru said:
Yes, yes..... therefore all the "satellites fail"..... every day another comes down....

So you suggest not to replace Tantalum Caps? There is a lot of conflicting info on this site, from what I can gather, if Tantalum Cap is connected to power rail then replace, sound like good advice?
 
Here's a replacement cap for the 15uf ones:

https://www.pollin.de/p/elko-axial-210159


Re the tantalums: It depends on their function in the circuit. If they were in the signal path, the sound might change quite a bit with electrolytic caps. Modern drop type tantalums can last a long time. There are other (ethical) problems with them, though.
 
tomheavybeats said:
So you suggest not to replace Tantalum Caps? There is a lot of conflicting info on this site, from what I can gather, if Tantalum Cap is connected to power rail then replace, sound like good advice?
A tatalum normally ONLY fails due to overvoltage (and then they short) what means that if you have a 15V power supply and only 16V tantalums they can fail.  But if you have 15V supply voltage and 25V or 35V tantalums they will probably last forever.

And don't replace any capacitor before checking if tey are faulty - most of the time it will make things worse.  Blindly recapping doesn't cure a fault - FIRST LOCATETHE FAULT WITH MEASUREMENTS.  Yes, there are some japanese junk electrolytics from the 70's and the first panasonic SMD-electrolytics from the early 90's where all will fail, but if it is a Frako, Siemens, Philips etc. they normally will not be bad.  Here is the proof - at the left a 22uF Philips electrolytic from 1976 and at the right a new "low ESR" 22uF electrolytic from 2019:

index.php

 
analogguru said:
And don't replace any capacitor before checking if tey are faulty - most of the time it will make things worse.  Blindly recapping doesn't cure a fault - FIRST LOCATETHE FAULT WITH MEASUREMENTS.
Electrolytics dry out, loose capacitance, get increased ESR and leakage after a while. Especially when exposed to heat. So the device could be loosing some significant low-end and picking up noise over time. Recapping would definitely cure that. I have heard your argument before about diagnosing before blindly replacing things and I really don't understand the logic of it. Many really hard faults are actually multiple independent problems masking each other. If you're going to eventually replace stuff anyway, then unless the unknown fault is fatal to the device as a whole (like heads on a tape machine), then logic would suggest replacing those parts and thereby minimize the complexity of your problems. If I have some old 70's gear that I want to have confidence in for the future, I would definitely "shotgun" it and just replace all of the electrolytics and probably the tantalums (with film if the gear is really fancy and they fit and aren't too expensive for the required size).

analogguru said:
but if it is a Frako, Siemens, Philips etc. they normally will not be bad.
People on tapeheads.net claim the Frakos are notorious for failing short. Maybe they're just talking. The Frakos pulled from my B77 measured ok but the machine wasn't used much.

analogguru said:
Here is the proof - at the left a 22uF Philips electrolytic from 1976 and at the right a new "low ESR" 22uF electrolytic from 2019:
One data point is not proof. It could be that that old cap is actually leaking more and it's throwing off the ESR measurement. I don't have a lot of confidence in some of those no-name LCR meters. If you run the same measurement 10 times, how close are the results? It is not abnormal for ESR of electrolytics to be higher at low frequencies. Maybe that particular series of what looks like Panasonic has higher ESR in exchange for some other properties like life, heat resistance, cost, etc.
 
analogguru said:
A tatalum normally ONLY fails due to overvoltage (and then they short) what means that if you have a 15V power supply and only 16V tantalums they can fail.  But if you have 15V supply voltage and 25V or 35V tantalums they will probably last forever.

And don't replace any capacitor before checking if tey are faulty - most of the time it will make things worse.  Blindly recapping doesn't cure a fault - FIRST LOCATETHE FAULT WITH MEASUREMENTS.  Yes, there are some japanese junk electrolytics from the 70's and the first panasonic SMD-electrolytics from the early 90's where all will fail, but if it is a Frako, Siemens, Philips etc. they normally will not be bad.  Here is the proof - at the left a 22uF Philips electrolytic from 1976 and at the right a new "low ESR" 22uF electrolytic from 2019:

index.php

I´ve had a shitload of blown tantalums in my gear and other gear I had to repair. But maybe I´m wrong and they still worked because they cannot fail as you say. I´ve heard the theory that tantalums never fail too often. It´s bad advice. Check the reality.
 
analogguru said:
A tatalum normally ONLY fails due to overvoltage (and then they short) what means that if you have a 15V power supply and only 16V tantalums they can fail. 
Tantalums are known to fail when reverse voltage is applied to them. In a well designed product it shouldn't happen, but actually it can happen, particularly at turn-off.
Replacing tants with Al types is said to change the sound, however I have no proof that the tant sound is better than the Al sound.
Anyway a well-dimensioned AL electrolytic should have no sound at all, if the circuit is decent.
The paper by Ciryl Bateman shows that, in order to get readable measurements, he had to subject the capacitors to such a current that creates a 3V rms voltage across it. This is far from being a practical case.
Typically, designers choose for coupling caps a value that is 10 times what's necessary for -3dB at the lowest frequency of operation. This results in the worst-case voltage across the cap 1/10th the output voltage.
In order to produce 3V across the capacitor, the output voltage must be 30Vrms.
A practical example would be a 130uF cap loaded with 600r. 30Vrms results in 1.5W.
I'm not saying it's impossible, but one who designs such a beast must be aware of the consequences and choose the value of the cap accordingly.
We are not anymore in a 600r world, so let's take a more current example.
With 100uF and a double load of 10k, the LF-3 dB point is at 0.3Hz. At 20Hz, the voltage across the cap is 1.5% the output voltage. In order to get 3 Vrms there, the output level would be 200V!
And the distortion levels measured by Bateman were all in the 0.002% region worst case.
He noticed that the distortion percentage varied more or less linearly with level.
That means that our hypothetical circuit of 100uF/5k, would produce 0.0001 THD at 10Vrms output.
I am in awe of people who can discern such low levels of distortion.
Particularly considering the quasi-total absence of energy at 20 Hz in most programs.
 
squarewave said:
People on tapeheads.net claim the Frakos are notorious for failing short. Maybe they're just talking. The Frakos pulled from my B77 measured ok but the machine wasn't used much.

Several years ago I was working on a Studer A800 24 track with several (maybe six???) channels all having a similar problem.  Shorted Frakos (with the gold/yellow plastic wrap).  I recall other bad Frakos in other gear as well.

Funky Frakos....is the the name of an old dance tune?  <g>

Bri
 
The older tantalum caps which look like a little blob of candy (TicTac or a jelly bean) across PSU rails seem to be aging poorly.  Ampex had them scattered around on the ATR-100 machine PCBs.  Now that I'm maintaining a fleet of eight of those machines, I've been seeing failures (shorts) of those decades-old tants.  Most recently, all was OK on one audio card for a minute or two, then the PSU rail was dragged down from 15VDC to 5 VDC or so,  turn off the deck for several minutes, rail was again OK then sagged again.

Bri
 
Spoke to an audio engineer I trust and have decided to leave Tants in as it constitutes the sound of the desk. He advised to run signal sweeps to see if old electrolytics causing frequency loss. I have bought replacement Nichicon radial caps just for one channel as a test too, I can always put old caps back in if I don’t like what new caps do
 
Back
Top