[Ampersand]

So I'm in the process of modding a Music Hall cd25.2 and naturally I need/want to replace the output caps. But the output cap configuration has me a little confused. Below is a schematic





With the two electrolytics in series, and then in parallel with the film, I calculated the overall capacitance to be ~111uF.

The amp that I'm using has an Input Z of 100kohms.

Using the equation 1/(2pi * R* C), i find the unmodded unit to have a cutoff frequency of .014Hz

Here's my question: I'd like to completely remove the electrolytics and just use a nice film cap. If I used a 0.47uF film in place of the above configuration, I calculate a cutoff freq of ~3.3Hz.

Does anyone see a problem with removing these electrolytics? Other than boosting capacitance, is there a reason they were used in the original circuit?

edit: Don't know how I missed it earlier, but the electrolytics are oriented so that their cathodes touch each other. Also I accidentally omitted a 4.7 Ohm resistor just before the output. Pretend its there

Thanks for the help.
mike

 

[synaesthetic]

If there are big electrolytics in the output it means that there's probably a lot of DC being pushed out of that circuit.

 

[Pars]

Well, the best thing to do is try it. Why they would put two electros in series is beyond me, unless they were fresh out of 100u caps. Do they use 220u elsewhere in the player, and just happen to be same brand, etc.? And I wouldn't suspect that there is ANY current in this application.

3.4 Hz is a bit high as the filter will still be having an effect an octave above this. Obviously increasing the cap (1u, etc.) will have an effect.

Running this into something with a lower input impedance than 100K will also have an effect.

 

[Ampersand]

Hey Pars, as a matter of fact they do use 220uF caps of same voltage and brand elsewhere.

I guess if I plan on using this cd player with other amps in the future, I should consider that as well. Also, I'll try and shoot for a 2Hz cutoff frequency.

 

[Ampersand]

Had to edit the first post. Electrolytics were not in the correct orientation earlier.

 

[amb]

Using two polarized electrolytics end-to-end like that is equivalent to a single non-polar ("bipolar") electrolytic of half the capacitance. IMHO its best to use a non-polar one if you need that much capacitance, or to use a good film cap if you don't. Music Hall probably did it that way because they use the same value polarized cap elsewhere in the circuit and it's one less item in the BOM, and cheaper too.

 

[Ampersand]

Wow, I never looked at it like a non-polar. Pretty cool. I'll probably go ahead and use a 110uF non-polar here then. Thanks!

 

[Pars]

You could just grab a couple of 47uf BlackGates such as:

Code:

[left]6.3 NX/47 6.3 x 7 2.50 16 FK/47 10 x 12.5 5.25[/left]

and be done with it. Or Nichicon Muse or another "audiophile" electro. In PP such as Solen, etc., you can get 4.7uf or so in a reasonable size for not alot of money.

 

[Ampersand]

Quote:

Originally Posted by Pars /img/forum/go_quote.gif You could just grab a couple of 47uf BlackGates such as: Code: [left]6.3 NX/47 6.3 x 7 2.50 16 FK/47 10 x 12.5 5.25[/left] and be done with it. Or Nichicon Muse or another "audiophile" electro. In PP such as Solen, etc., you can get 4.7uf or so in a reasonable size for not alot of money.

I'd like to use a single NX/100uF and then bypass with a 1uF film, but the NX is only rated 6.3V. The original caps are 25V. Am I correct in assuming their should be little DC at this point and 6.3V will suffice?

edit: Maybe I should just eliminate the film altogether if I'm using Blackgates. But I was soooo looking forward to a nice big round film cap

edit: I did find a 4uF Dynamicap that isn't too large. Could just use that and no electrolytics in output. Too many choices, I'm going insane.

 

[Pars]

Yeah, if you go BlackGate, I probably would not bypass them. 47uf is plenty, not sure why you would go to 100u. 6.3V is used often for DAC coupling caps, and there shouldn't be much DC here, maybe a volt or so (measure before you order though just to get an idea). My Rotel CDP had 3.3V at this point off the opamps, just because of how they were configured.

In general, I would think film caps would sound better than any BG, though the BG FK is reputed to be the best for coupling.

I am about to order some Dynamicaps myself for my tube pre...

 

[Ampersand]

Quote:

Originally Posted by Pars /img/forum/go_quote.gif I am about to order some Dynamicaps myself for my tube pre...

In my endless capacitor searches, when i come upon a Dynamicap review - it's usually positive. And they're very affordable (compared to others I considered - Mundorf,Relcap,Multicap)

 

[Roam]

Quote:

Originally Posted by Ampersand /img/forum/go_quote.gif edit: Maybe I should just eliminate the film altogether if I'm using Blackgates. But I was soooo looking forward to a nice big round film cap edit: I did find a 4uF Dynamicap that isn't too large. Could just use that and no electrolytics in output. Too many choices, I'm going insane.

Junk the Blackgates, no electrolytic on Earth is going to come anywhere near the performance of a decent film capacitor. All electrolytics have large amounts of hysteresis & non-linearities at low voltage levels, this results in all the low-level details in the signal being wiped out. Electrolytics have absolutely no place in the signal path, put them in the power supply where they belong.

Use a good polypropylene film cap in the 1-5uF range and you're done.

 

[Ampersand]

Quote:

Originally Posted by Roam /img/forum/go_quote.gif Use a good polypropylene film cap in the 1-5uF range and you're done.

Well, I went the Dynamicap route with my order today, so we shall see....

 

[Budgie]

I agree with Roam. That being said,if you have to use a "lytic" in a coupling application use a non-polarized.