cap couple or direct couple class A SE amp?

[AudioGeek]

I am in the process of building a Class A SE amp.

supply is either 30v and cap coupled output or +-15v and direct couple output.

output stage is simple source follower with a CCS source.

Is it "better" to cap couple the output with some High quality caps (8000uF 50v) bypassed with films, or use a DC servo to prevent any DC with the split supply?

If I "were" going to cap couple it.... any recommendations for make or type of cap?

Thanks!

[dhaen]

The power supply caps are in series with the signal whichever way you do it.
Will the servo introduce LF noise?

[AudioGeek]

a dc servo is not the way i wanted to go.

cap coupling is probably what i will do.

any recommendations for a output cap?

[Chris Jennings]

Hi.

Try a cheapish industrial grade electrolytic, 5000-10000uF and bypass with a good quality polypropylene 10-50uF (the audio type used in speaker crossovers) work well.

This combination gives a remarkabely good performance for the money, it's cheap to proyotype and easy to upgrade with better parts if required later. But from my experience even reasonably priced electrolytics work really well if the by-pass is of good quality.

Chris

[AudioGeek]

should I use the same value / brand of caps in my power supply as my output cap?

how about 6800uF cap either Nichicon or BG?

then bypassed with 10uF film, bypassed with a 0.33 oil cap..

maybe just use the Nichicon in PS...

[AudioGeek]

if my output coupling cap and my power supply caps are all the same... and bypassed the same....

then I should have no problems...

or it is a waste of time to spend a ton of $$$$ in the output coupling cap and not much $ in the PS?

[dhaen]

I think you may be answering my deleted post. I deleted it because I had read your original post wrong - sorry.so the detail in my post was incorrect for your system. However it was correct in principle, so yes, PS component quality is of exactly the same importance as signal components, in SE.

below is my deleted post for the record:

IMO a reasonable quality cap of not too high a value. So why aren't I saying to use the latest "Silver-Gold-Super-Hyperbola" cap? Because of the reasons stated in my previous post.
Consider: The anode (plate) of the driver valve is cap coupled to the grid of the output valve. Let's called this signal send. Every signal has a send and a return. The return path is from the cathode of the output valve through the cathode components to the ground line, then through the smoothing cap or local decoupling cap to the top of the anode resistor of the driver valve. That's a torturous path. What is the quality of those components? If is up to the standard of the "Silver-Gold-Super-Hyperbola" cap?
There are ways of avoiding this, but all use transformers for coupling.

[AudioGeek]

so how far should I go at bypassing?

once?
twice?

and then I assume the same in the PS?

6,800uF 50v Nichicon cap in output... with a 10uF then a 0.33?