Choice of caps

[dbelousov]

I use the Pana FR almost exclusively and yes, increasing the size helps with distortion.

I found that it's hard to "kill" line level signals nowadays. I have a monitor controller (LM4562-based, direct coupled, servos on the outputs, good relays) so I can switch from the direct outputs of my DAC to the monitoring through my recording console. That means the signal is routed through the stereo input channel to the master and then to the controller. This path contains lots of the LM4562s and 3 pairs of dramatically oversized Pana FR caps (faders and ouputs). And, with careful level matching, there is almost no difference in the audible sound quality. I need a very decent orchestral record to distinguish the paths, and with any "-6 pop mix" it's just impossible in a blind test.

On the other hand, with low level signals (mics), lytics, in my experience, degrade signal a lot, so I use the polypropylene types only.

So yes, I've experienced the "harshness" of Pana FR myself when I used them as the phantom blocking caps. It was unexpected, but it's true. It's especially pronounced with fast uber op-amps. Sibilant, harsh, piercing sound. I've tried other lytics and they all sucked in this position in different ways. So, I use MKPs only as phantom blockers now. Nothing special, though, just regular TDK/Wima industrial staff. The major problem is the SIZE, obviously. I would be happy not to spend such amount of money and real estate for them, but I've found any other types are inferior in this position. And I can match the polypropylene ones.

 

[Rob Flinn]

Having recapped quite a few large desks and other pieces of gear, in my experience it is mainly low value caps that go bad.  I have bulk tested the caps I pull out on several occasions & most of the ones displaying an high ESR reading are 47u and below in value, then maybe a few 100u are bad, but generally speaking the 100u & above test ok.

 

[spunky]

Fascinating thread this became.

As far as I understood, the choice between the brands I've mentioned is more like a personal preference, than an actual significant difference. The board now is filled with Jamicon. Like said in this thread, most modern caps from these ages will do a real good job.

I will go with one of the Panasonic series. Like said before, I keep their original values. But perhaps I will higher the voltage of the caps, if the space on the board will allow it.

I want to thank you all for your valuable opinions and replies. From me to all of you, all the best for 2021

 

[soapfoot]

Going higher on the voltage is almost always a good idea when space allows (and it usually does when replacing old parts with modern parts).

My method of choosing electrolytic caps begins with measuring the old parts and/or the space in which the old parts were installed.

Then I look at the datasheets for the new parts in which I'm interested, and get the highest voltage (and temperature-rated) cap that will fit. I can often double or even triple the voltage rating, which will usually result in a longer life.

The way it was explained to me by people better-informed than I: There's sort of a 3-way relationship between voltage rating, temperature rating, and service life. The more margin you can allow in the first two areas, the longer your service life is likely to be.

 

[Khron]

Can't say about voltage, but the temperature-lifetime dependency is an official spec. The basic lifetime is spec'd at the max operating temperature and max rated ripple current.

For every 10C under the rated temperature, expected lifetime is doubled. 2000h cap rated for 85C should be good for 4000h at 75C, or 8000h at 65C and so on.

Then I look at the datasheets for the new parts in which I'm interested, and get the highest voltage (and temperature-rated) cap that will fit. I can often double or even triple the voltage rating, which will usually result in a longer life.

The way it was explained to me by people better-informed than I: There's sort of a 3-way relationship between voltage rating, temperature rating, and service life. The more margin you can allow in the first two areas, the longer your service life is likely to be.