[dude_500]

I have a DIY Strain amp with $11 output caps from mouser 935C2W10K-F

Would it be worth upgrading those to say, 15uf 200v sonicaps for a total of $50? Seems like a lot of money for two capacitors. Will I actually notice a difference?

Listening on HD600's and in the process of building a gamma 1

 

[nikongod]

Not worth it.

Although the sonicaps are no doubt nicer than the c(r)aps you are using now, you are missing out on a bunch of bass with 10uF caps on a 300-ohm headphone.

You will probably find great improvements by improving the QUALITY of c2, and the SIZE of the output cap: 10uf is too small for a 300 ohm load.

Im too lazy and unmotivated to look up what other people like in HV electrolytic caps, but you can. 10:1 odds says someone here thinks my favorite electrolytic coupling caps sound like crap anyways. 10:1 odds says I think their favorite output cap sounds like crap (If I have even heard it) right back at them.

100:1 odds says we are both right, and have successfully built our systems to our own tastes.

10:1 odds on top of that says other people say that a transformer is the only output for a tube stage, they are right too! CRIKEY! 1000 different opinions from 1000 different people listening to 1000 different pieces of music with 1000 different pairs of ears...

 

[dsavitsk]

Quote:

Originally Posted by nikongod /img/forum/go_quote.gif Not worth it. Although the sonicaps are no doubt nicer than the c(r)aps you are using now, you are missing out on a bunch of bass with 10uF caps on a 300-ohm headphone. You will probably find great improvements by improving the QUALITY of c2, and the SIZE of the output cap: 10uf is too small for a 300 ohm load.

It's not quite as bad as you think as you need to add the Zout to the Zin. Usually you just use Zin to calculate cap size because you figure nobody would design an amp with a high enough Zout to matter. But, here they did.

So, C = 1/(2 * pi * F * Z), and F is about equal to 800. That said, 10uF, or even 15uF is still a little low, though perhaps tolerable. Moreover, this design has enough problems that while the cap upgrade might make it sound better than it does, it will never make it sound good. Finally, since failure on the part of these caps is a safety issue, 200V caps are too low. Their voltage rating ahould exceed B+.

 

[dude_500]

ok, I'll look at getting a better quality on C2 and upgrade the size and voltage rating of the output cap.

You said the "Design has enough problems"... are you refering to the strain amp schematic? Is that not a very good one to build? oops ha ha


Where do you even get a 20+uf 300+v capacitor for less than $100 a piece? Sonicap doesn't even go that high

 

[pabbi1]

Quote:

Originally Posted by dude_500 /img/forum/go_quote.gif ok, I'll look at getting a better quality on C2 and upgrade the size and voltage rating of the output cap. You said the "Design has enough problems"... are you refering to the strain amp schematic? Is that not a very good one to build? oops ha ha Where do you even get a 20+uf 300+v capacitor for less than $100 a piece? Sonicap doesn't even go that high

Mundorf M-Caps?

 

[cetoole]

Quote:

Originally Posted by dude_500 /img/forum/go_quote.gif Where do you even get a 20+uf 300+v capacitor for less than $100 a piece? Sonicap doesn't even go that high

Motor Run caps. I got a box of 10 100uf 370VAC ones right here, paid about $100 for the whole thing, shipped. Metallized Polypropylene in Oil. Of course, as Doug said, that schematic really doesnt look like something I would want to drive my headphones with, and I would see about upgrading/replacing the whole thing instead of just replacing output caps.

 

[dude_500]

Can I go too big on the capacitors or should I just get the biggest one I can find/want to pay for that is atleast B+? I'm not sure what F and Z are in your equation or how to find Zin and Zout if that's what they are...

 

[cetoole]

F=frequency, Z=impedance, Zin=headphone impedance, Zout=amplifier output impedance, and this you need to calculate, but considering that the output stage is basically parallel 12au7 cathode follower, it is gonna be high. Dont feel like running any numbers right now, and I dont have any spice models for these (or any other) tubes, but I would guess Doug isnt too far off.

 

[dsavitsk]

Quote:

Originally Posted by dude_500 /img/forum/go_quote.gif Can I go too big on the capacitors or should I just get the biggest one I can find/want to pay for that is atleast B+? I'm not sure what F and Z are in your equation or how to find Zin and Zout if that's what they are...

No reason to go bigger than you need. And, bigger can draw current in the form of a pop that you might not like.

F is the 3db corner frequency (explained here) which many choose to be about 2Hz, but up to 20Hz is probably fine. Z is the impedance (here Zin plus Zout). Zin is your phones impedance which you can get from Headroom's website if nowhere else. Zout is given in the description of the amp. (any my last post has a typo -- Z is 800, not F). C is the capacitance.

Anyway, the design is not my cup of tea as it does not really do what I think a headphone amp should do to sound good. I could be wrong, though.

 

[dsavitsk]

Quote:

Originally Posted by cetoole /img/forum/go_quote.gif Dont feel like running any numbers right now, and I dont have any spice models for these (or any other) tubes, but I would guess Doug isnt too far off.

Zout of a cathode follower is 1/Gm or ra/mu. It is a little more complicated here, but i just read the value from the description. Maybe it's wrong.

 

[cetoole]

Quote:

Originally Posted by dsavitsk /img/forum/go_quote.gif Zout of a cathode follower is 1/Gm or ra/mu. It is a little more complicated here, but i just read the value from the description. Maybe it's wrong.

Yes, but you still need to know the op point. That may be in the article somewhere, but if it is, I must have missed it, along with the stated output impedance.

 

[nikongod]

Quote:

Originally Posted by cetoole /img/forum/go_quote.gif Yes, but you still need to know the op point. That may be in the article somewhere, but if it is, I must have missed it, along with the stated output impedance.

The output impedance is in the paragraph immediately below the schematic. It is roughly reinforced by the simulations CHU did at the bottom, and fricken high! maybe only 200-250 ohms in reality though, still high for an OTL tube amp.

Anyways, working backwards from the info in the text (the author says about 20ma somewhere in there) and the listed values I found an operating point that I think checks out.

An idle current of about 18ma allows all voltages in the amp to line up nice and clean with the specified resistor values. The voltage at the top of R2 is 140V, the plate voltage for the bottom pair is 118V it may not be right, I dunno, see if they make sense to you. Its late, and im tired. The triodes are paralleled, so the voltages line up with 9ish mA on the plate curves....

What does anyone need 18mA of plate current with a 250-ohm output impedance for anyways? it cant possibly put out enough voltage to put 18ma into a grado considering the Zo, and not a chance it can do it to a senn...

There must be something about this amp I dont understand.

 

[dude_500]

Not sure, it sounds dang good to me but I haven't heard one of the bijou or other "great ones". It sounds infinitely better than my stereo headphone port.

It runs my sennheisers very very loud, running on extremely low volume. Sounds like this is the "mystery schematic" ha ha

I finally got my parts from mouser today and got the 120hz ripple out I was talking about 2 weeks ago, and the bass got much stronger and it overall sounds even better than it did.