Is this a special cap? Vintage yellow Chemi-Con LL

[Soul One]

Hey guys,

Can anybody shed some light on whether these capacitors are low leakage ones? I am asking because of the LL writing.

These are used in the channelstrips of my vintage (1984) Tascam M-520 console (see schematics).

I was planning to use Elna Silmic II here. Or would a specific low leakage cap be the better option?

Thanks!

 

[moamps]

It is a coupling capacitor and should not be low leakage, IMO, so you can use Elna Silmic instead. Although I would rather put Nichicon fine gold UFG type there.

 

[Bo Deadly]

It's not clear to me why low leakage would be important in this case. Just about any electrolytic should work fine there.

Although it looks like they could be slightly reverse biased but for now I will assume they're oriented that way for a reason.

 

[Brian Roth]

I found an owner's manual online. That desk had 48V phantom, so those caps would be phantom power blockers. 50V DC rating seems low to me. I would also believe that a low leakage part would be a good choice.

Bri

 

[Bo Deadly]

I would also believe that a low leakage part would be a good choice.

Why? Even if it were 100uA (which is crazy high), that would be a 0.3V drop in phantom V.

 

[Brian Roth]

I guess my main point was that the "+" lead orientation in the schematic is correct for blocking the phantom. I may be over thinking the possible effects of leakage which could (in theory) throw off the bias point of the input transistors.

Bri

 

[JohnRoberts]

Gross leakage current could be an issue feeding the input to a 60dB gain stage.

I have only seen problems with one batch of electrolytic caps in that circuit node over a few decades. I didn't bother to figure out why, I just black balled them for use in production (it was a new cap series, from an existing vendor).

Decades later thinking about it they might not have been properly formed in by the cap mfr (WAG).

JR

 

[moamps]

For the electrolytic capacitors produced today it could be said that they are all "low leakage" compared to capacitors from 40 years ago or more. For the types mentioned here (Elna Silmic and Nichicon FG), as well as for the relatively popular Panasonic FC series, the leakage current is given according to the formula

Ileak(uA)=0.01CV (C-uF, V-volts),

which means that for the capacitor 22/50, the leakage current is 11uA.

On the other hand, low leakage promoted capacitors (Nichicon KL, Elna RLB for example) give that current according to the formula

Ileak(uA)=0.002CV (C-uF, V-volts),

which is "only" 5 times better than standard ones.

In this Tascam M520 mixer example, the 11uA leakage current will offset the input transistors in the differential pair by 0.1V, but both equally. With the addition that the transistors are powered by a current source, I don't see that there could be any problems in the functioning of the preamplifier.

 

[user 130681]

I found an owner's manual online. That desk had 48V phantom, so those caps would be phantom power blockers. 50V DC rating seems low to me. I would also believe that a low leakage part would be a good choice.

Bri

How can they be for blocking the 48v, there is 3k4 to ground on each leg. Phantom power will be killed with that impedance! They are for erronious DC blocking.
I would leave them alone unless they are low value.

 

[abbey road d enfer]

How can they be for blocking the 48v, there is 3k4 to ground on each leg.

Huh?

 

[abbey road d enfer]

In this Tascam M520 mixer example, the 11uA leakage current will offset the input transistors in the differential pair by 0.1V, but both equally. With the addition that the transistors are powered by a current source, I don't see that there could be any problems in the functioning of the preamplifier.

The 0.01CV is given as a maximum, so it's quite possible that one side gets only half of the other, resulting in 50mV offset, which would try to push 50V output offset with 60dB gain. Of course it's a quite extreme situation, but I think some attention must be put there.

 

[gyraf]

Jon, those attenuator resistors are between hot+cold, not in p48's way...

 

[JohnRoberts]

How can they be for blocking the 48v, there is 3k4 to ground on each leg. Phantom power will be killed with that impedance! They are for erronious DC blocking.
I would leave them alone unless they are low value.

22uF is a popular value for phantom DC voltage blocking. In use these caps will see well less than 50V DC.

JR

 

[user 130681]

Just noticed, the 48v line is not to ground, Doh!
That makes sense now.
By the time the 48v is dropped by the supply load resistors, it will probably only be around 35v or so.
Either way I agree, the voltage of 50 is a bit close.
A&H use 47u 63v everywhere, so no issues with their pre amps.

 

[moamps]

The 0.01CV is given as a maximum, so it's quite possible that one side gets only half of the other, resulting in 50mV offset, which would try to push 50V output offset with 60dB gain. Of course it's a quite extreme situation, but I think some attention must be put there.

You lost me here. Which microphone amplifier (without a servo) has a DC gain of 60dB?

 

[Soul One]

Thanks a ton for the replies so far, very helpful!

 

[abbey road d enfer]

You lost me here. Which microphone amplifier (without a servo) has a DC gain of 60dB?

IDK. There are "purists" who use a Cohen-style circuit without the dreaded emitter capacitor...

 

[JohnRoberts]

IDK. There are "purists" who use a Cohen-style circuit without the dreaded emitter capacitor...

I am cautious about operating electrolytic caps into low single digit ohms for audio... DC servos can manage Vbe errors.
===

I even know of mic preamps that avoid the phantom DC blocking capacitors too... I killed a lot of brain cells soaking that idea in beer, Wayne actually made it work.

JR