DC Blocking caps on BPA200. Electrolytic?

There are LOTS of strong opinions floating around, here, about input coupling capacitors, and how different types might affect the sound. So I'll mostly only address the "practical" aspects.

Basically, you can probably use any type of cap, for those (although I'd stay away from the old-style tantalum types). A polarized aluminum electrolytic should even be fine, at least electrically. The DC and AC levels should be pretty small, at those points in the circuit. But you'd probably still want to use a voltage rating that's at or above the difference between the power supply rail voltages, "just in case".

If you're worried about using a polarized electrolytic type, you can use two in series (with the two "-" terminals connected together), with each being about twice the needed value, to form a bipolar electrolytic. So, say, two 47uF in series for each 22uF.

If you're going to try electrolytics, at least at first, you could use something like the "low-ESR, high-reliability" Nichicon UHE1E470MDD 47uF/25V radial caps (two in series), which are part # 647-UHE1E470MDD at mouser.com . Or you could try the 22uF/50V, 647-UHE1H220MDD at mouser.com . Both types are about $0.15 each, qty 1.

If you do use electyrolytics, it would probably be a very good idea to parallel each 22uF with a smaller-valued film cap, maybe something between 0.47 and 2.2 uF.

I have also heard that you can get better sonic performance when using two electrolytics in series by applying a DC voltage between them. You'd probably have to experiment, or do some research, to find out how to best do that.

For the best sound, and if you have enough space and money for them, you would get some 22uF film caps, i.e. with polyester, or, better yet, polypropylene dielectric. [I won't even mention polystyrene, polycarbonate, or teflon, for these sizes (But if you find some cheap 22uF Teflon or polystyrene caps, please do let me know!).] But 22uF film caps could be a bit expensive, maybe, and will be quite large in size, compared to electrolytics. You can also parallel smaller values (which makes their values add), if that happens to work out better for you.

AVX makes some large-valued polyester and polypropylene caps. Their FFB series has a 20 uF/100V polyester for $6.33 each (Mouser.com # 581-FFB14E0206K). They also have a 33uF/75V for $7.95 each. The FFB caps with voltage ratings at or above 525V have polypropylene dielectric, which is better. There's a 10 uF/525V model for $10.58 each. But it would get too expensive, paralleling those, I guess. There are cheaper ones on ebay.com, anyway.

You also might be able to find a large quantity of dirt-cheap 2.2uF polypropylenes (or polyesters) on ebay, or somewhere, and parallel ten of them for each 22uF cap. (The 2.2uF polyesters can be quite small, in the 63V and 100V sizes).

Cornell Dubilier has a 100V 10 uF polyester model in their DME-series that mouser.com has for $4.22 each (mouser.com # 5989-100V10.0F). You could parallel two of those and one 2.2 uF, to retain the bass response of the original circuit.

Cornell Dubilier also has a 20uF 200V polypropylene in their 935-series, mouser.com's 5984-200V10-F, for $11.16 each.

I have also heard that "motor run" caps, which are usually polypropylene, work well, and might be cheaper.

If I'm wrong about any of this, or have left something out, I'm pretty sure that someone else will chime-in.

Good luck.

- Tom Gootee

http://www.fullnet.com/~tomg/index.html

-

Nick,

That all sounds reasonable. Although, the electrolytics in those positions won't be likely to fail from overheating, since there will be virtually no self-heating, due to the low currents. And the caps I mentioned are 105degC parts, anyway. But, they might still fail after 20 years or so, if they dry out, and maybe much sooner if a seal is bad, or there is some other defect.

At any rate, polypropylenes would be wonderful. You could even go higher than 22 uF. This is just a guess, basically, but, I think I'd still also parallel the big polypropylene input-coupling caps with some much smaller polypropylenes; say 0.47u or less, just to make doubly sure that fast-rising edges get through as well as possible, for better "punch" on fast attacks. You could even add another smaller size, too, maybe something around .047uF. (It might be nice if the smallest one were polystyrene. But I haven't seen many of those above about .01uF.)

- Tom Gootee

http://www.fullnet.com/~tomg/index.html

You might also want to consider using DC Servos, instead of blocking caps. I posted a link to a schematic for a pretty-good one, in this thread:

http://www.diyaudio.com/forums/showthread.php?s=&threadid=103308&perpage=10&pagenumber=1

- Tom Gootee

http://www.fullnet.com/~tomg/index.html