Output capacitors, is there better than Electrolytic?

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

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

calaverasgrande

Well-known member
Joined
Jan 15, 2015
Messages
91
I'm refurbishing an old synth. Just noticed I'm getting DC voltage on the output.
Apparently the output cap is not really a cap anymore.
So before I order the exact same part which was specced 40 years ago, is there a better part?
I know this is an asinine question to ask a priori.
However I'm just curious if materials science has progressed, or standard practice has changed since the late 70's.
Right now its a 220uf 16v lytic. Not specified as anything special in the service manual. Tolerance is +50/-10.
Curious if a film cap might be better?

BTW I've searched for info on this for days, but seem to have had a hard time coming up with search terms that don't run me aground of audiophile gibberish on one side or unrelated stuff about microphones. Seems odd, lots of things use DC blocking output caps?
 
The capacitor technology has improved dramatically in 4 decades (while I have some in my back lab older than that).

Replace with the same capacitance and voltage and you will be fine. That will not be your weakest link.

[edit- be sure to observe proper voltage polarity /edit]

JR
 
Last edited:
Technically a similarly rated film cap would be better. But you might want to check the physical dimensions and cost 😳 And tbh no one is going to hear a difference on a synth output.
Apart from that you can probably get a 'lytic with greater capacitance and/or voltage/temp rating to fit same space envelope and footprint.
Last time I bought reservoir caps for an old amp head PSU I had to put foam round them to fit the existing cap holder clips !
 
Last edited:
Yeah, I spent some more time on Mouser a couple days ago and while I did find some film caps and a couple others that were in the neighborhood of the original cap, they were all prohibitively expensive for capacitance that size.
As far as physical dimensions, It's kind of hilarious how much smaller caps are now than the giant things they used for 16v (output) or 36V (power supply) 40 years ago.
Might revisit these caps after I get the other problems sorted, like the all important Moog filter.
 
Note if you don't have an output load to bleed off leakage, it's normal to measure some DC on the output side. Not to say it shouldn't be replaced. Could be worth comparing for educational purposes.
 
Yeah I was just looking into that.
Turns out it originally had a little PCB with a resistor network that terminated in RCA jacks.

Now there is just a 1/4" jack with no resistors.
Which is a common mod a lot of people do to these synths.

Gonna strap 47k on it and a 1k to ground in an approximation of the original output.
Not sure the 47k makes a ton of difference, may try 22k or 10k.
But the 1k to ground certainly does.

Update, that fixed the voltage offset quite well.
Now gotta figure out why the filter sounds like it's full of rocks.

Screen Shot 2021-07-03 at 6.53.27 PM.png
 
Last edited:
If you need to eliminate more DC leakage, you can use the 16V/220uF electrolytic cap followed by a 220uF/6V cap since the second one doesn't need to withstand any DC voltage. But it sounds like you're more worried about major DC leakage due to aging. Yes an electrolytic cap will dry out eventually but we're talking probably decades depending on the cap specs
 
I still struggle with the whole DC blocking output cap thing.
So much that I read early on in my journey into audio electronics railed against electrolytics anywhere in the audio path.
I feel like I'm putting a big ole booger on a plate fresh out the dishwasher. Lol.
Maybe I'll start a trend of galvanically isolated monosynths.
 
Note that most electrolytics these days are on the low end of their tolerance spec, so a 100 uF +/- 20% part will usually be 80-90 uF. At least the price sensitive stuff. In critical applications it's a good idea to go one or two sizes higher i.e. 330 uF for a 220 uF part. Although I think in this case 220 uF is overspec'd enough not to worry, if it's only 180 uF.

Also, for line outs you may want to consider a 63V rating, if space permits. Just in case it ever gets plugged into a mic pre with phantom power activated.

And if the piece of gear generates any heat, make sure to use 105°C parts. Even if the temperature is much lower, they will last longer.
 
Note that most electrolytics these days are on the low end of their tolerance spec, so a 100 uF +/- 20% part will usually be 80-90 uF. At least the price sensitive stuff. In critical applications it's a good idea to go one or two sizes higher i.e. 330 uF for a 220 uF part. Although I think in this case 220 uF is overspec'd enough not to worry, if it's only 180 uF.
Production design engineering 101.... design for worst case tolerance variation. In large volume production ignoring tolerance can and will come back to bite you given enough time. I knew one amplifier design engineer who ASSumed all capacitor manufacturer's voltage specs were conservative and pushed the envelope in his amp designs to save on BOM cost (higher voltage caps cost more and take up more space). That burned him.
Also, for line outs you may want to consider a 63V rating, if space permits. Just in case it ever gets plugged into a mic pre with phantom power activated.
Capacitor coupled line outputs were commonly used to block DC from the internal circuitry corrupting the output***. A polar electrolytic cap in that case would be facing the wrong direction to block positive phantom voltage feeding back into the output. Of course you could use non-polar blocking caps (its only money).
And if the piece of gear generates any heat, make sure to use 105°C parts. Even if the temperature is much lower, they will last longer.
+1 higher heat reduces MTBF (mean time between failure).... cooler is cool. 105°C caps will tolerate heat better than 85°C parts, better is better.

JR

PS: Hi Rossi

*** single supply (class A?) circuits are typically internally biased at a +V/2 voltage. Of course some single supply circuits operate from negative supply voltages. Output blocking caps in split supply circuits could be randomly polarized.
 
I remember the Neve VR (post Rupert, he would have never...) choosing a cheaper cap, with a lower temp, and the entire VR console line had to have them replace ALL the coupling caps in every module, which took an 8 hour day for each one. That was a mess. Don't be that guy... For audio coupling, since the swing isn't rail to rail, in usually pick 25 -35 volts when using +-16V. For the size, I calculate my lowest frequency I want and then make it one size larger, for phantom, the MINIMUM would be 63 volts, I try 100 if they are available, I have seen 50V... That scares me.
 
The capacitor technology has improved dramatically in 4 decades (while I have some in my back lab older than that).

Replace with the same capacitance and voltage and you will be fine. That will not be your weakest link.
Very true, although it is never a mistake to recheck if everything is working as expected. Some years ago I recapped a mic pre and it turned out that the (supposedly high quality) replacement cap I'd chosen for the input was actually leaking some remains of the 48V phantom power into the preamp input. Replaced that cap with a different brand and all was fine. Parts have become a lot better, but nothing is 100% safe...

Michael
 
Simple rules of troubleshooting:

If you built it, and made a mistake, you will overlook the mistake. Have someone else check it.
If you don't think it's that part, it is.
Humility is success.
 

Latest posts

Back
Top