[rickcr42]

maybe these ?

http://www.keyelco.com/products/prod20.asp

 

[n_maher]

Quote:

Originally Posted by NeilR If I were going to mod that board, I would make C2 the same size as the other caps. I think that was a strategical error. I agree with you- they are just as important [just as much in the signal path] as the output couplers.

I'm not sure I agree with this statement, but I won't fight it. When I look at the schematic I don't see how C2 is in the signal path. Quote:

I hate to ruin the aesthetics with some big caps air wired onto the board, but I guess it ought to be done. I've been thinking about that for the "educational builds". All I used for C7 was a Cerafine. I'm haven't gotten into boutique caps [yet].

So wire the caps under the board

. And the uber-millett starts construction soon. All PRP/Holco/Caddock resistors, all BG caps (and I do mean all) and the output and cathode bypass will be configured as Super-E caps (requires 3 caps at each position). Volume control will be handled by a stepped attenuator, buffers will be DBs, and that's about it for audio-bling. Should be fun to total up the cost of the board components since someone issued the challenge of seeing just how high it could go so I figured I'd search for the ceiling.

Nate

 

[NeilR]

Quote:

Originally Posted by rickcr42 Check the rest of the Keystone catalog.There are so many pcb mount options from available them that if not there it likely is not availabvle anywhere ! (have you considerd a simple hard metal "hook" type post ?) Good company with great service,unique parts and great prices.One of the true DIY Supply good guys

Do you buy direct from Keystone? I have only bought their parts through Mouser.

It was irrelevent to me because after I couldn't get the Keystones (which I had laying in my bin) in, I decided to source the probe tip sockets for the back panel. I just pointed that out to keep someone from ordering those parts you illustrated, thinking they will fit. The pad holes for the probe tips are very small. Anyone trying to solder in some sort of probe helper needs to check the pad size and the pin size of whatever they try to put in there. Going panel mount eliminated that issue since I soldered 22ga wire into them.

 

[NeilR]

Quote:

Originally Posted by n_maher I'm not sure I agree with this statement, but I won't fight it. When I look at the schematic I don't see how C2 is in the signal path. So wire the caps under the board . And the uber-millett starts construction soon. All PRP/Holco/Caddock resistors, all BG caps (and I do mean all) and the output and cathode bypass will be configured as Super-E caps (requires 3 caps at each position). Volume control will be handled by a stepped attenuator, buffers will be DBs, and that's about it for audio-bling. Should be fun to total up the cost of the board components since someone issued the challenge of seeing just how high it could go so I figured I'd search for the ceiling. Nate

Nate- no room under board for caps if it is going in a Hammond, as mine did.

C2 - this is an AC bypass for the signal. Without it, the audio would pass through the pot and that would actually be a regenerative feedback system, which is not desired here. The cap has an impedence measured in single digit ohms or less (I calculated the corner once but I forget the exact value- it is small- maybe a fraction of an ohm). If you were an electron asssociated with an audio signal, would you go through the roughly 1K resistor (trim pot) or the 1 or 2 ohm (or less) capacitor? To AC, that cap is a straight wire (with a little phase shift for flavoring).

Some people say- no, that connects the cathode (more or less) directly to ground at AC. Not true or the wrong perspective. The signal starts at ground, runs up through the two bypass caps, through the tube and up the plate circuit. From the AC perspective C2 is a coupling cap.

 

[rickcr42]

Quote:

Do you buy direct from Keystone?

Yes unless I am ordering a bunch of stuff from digikey or mouser and adding in the Keystone parts I go direct to the source and they are cool to work with.

 

[NeilR]

Nate - don't feel bad about the "mistake" with C2. Pete Millet got it "wrong" too, if the current line of thinking is correct. This was discussed in depth in a recent thread that I could locate if anyone cares. I guess thinking has evolved over the years. It would be interesting to broach this to Pete for his thoughts. He is the Tube Maven.

 

[n_maher]

I'll ask Pete about it at the National Meet

.

And like I said, to be on the safe side my C2's will be of the same quality as C7, which is to say ludicrous overkill.

Nate

 

[NeilR]

Here is some homework...

Note his reference to ck as an ac bypass and the discussion of shelving frequency if the value is too small. If that cap will shelve the signal, it is in the path.

Or maybe it is just one more example of internet misinformation

 

[NeilR]

Nate- in the above link, note Rg and the ensuing discussion of that part. In the Millet, the volume pot acts as RG, which is why the tube will not bias properly without the pot installed, or more accurately, at least a temporary resistor or even a wire. Without the pot, the grid is just floating in air. I think this issue has come up from time to time in trouble shooting threads.

 

[dsavitsk]

Quote:

Originally Posted by NeilR Here is some homework... Note his reference to ck as an ac bypass and the discussion of shelving frequency if the value is too small. If that cap will shelve the signal, it is in the path. Or maybe it is just one more example of internet misinformation

Nope, the information is good. You might also spend an afternoon with Morgan Jones' book if you trust it more. This is the achilles heal of this topolgy and the reason why the ultrapath is so clever, and why it performs so well. See: http://www.electra-print.com/ultrapath.html

Quote:

Originally Posted by NeilR Nate - don't feel bad about the "mistake" with C2. Pete Millet got it "wrong" too, if the current line of thinking is correct.

Not quite sure why this is the case. You always need an electrolytic there, and Pete used one of an appropriate size and quality. Further, it can be smaller because it never sees a high voltage.

 

[rickcr42]

Quote:

This is the achilles heal of this topolgy and the reason why the ultrapath is so clever, and why it performs so well.

......and why it is also called the "Western Electric" connection because they used it way back in the thirties in their amps and is another example of something "new" just rediscovering something the old timers already had figured out

 

[NeilR]

dsavitsk,

By "Pete's mistake" I referred to his reference to (our) C2 as a "bypass cap" and not stressing the fact that it is in the signal path (but keep reading).

Here is the thread I referred to. In it I calculated a value of only 75uF needed for a low shelf frequency and low phase shift, using a formula provided by Steinchen which he used to compute the value of C7 for various headphone impendances. My thinking was possibly to reduce the capacitance (from the spec'd 200uF) and trade for a bigger piece of jewelry.

I think Pete was trying to dissuade people from "Boutiqueing" the amp so he did not get into fancy caps. So he may not have made a "mistake", it may just have been his perspective on things. That's why it would be interesting to ask him his thoughts, which Nate has taken on for us.

UPDATE: In his article, page 3, he says "Bias for the tube stage is developed across an adjustable resistor... which is paralleled by both an electrolytic ... and film capacitor..... The capacitors provide a low impendance path for the audio signal"

So there you have it. He got it right and I got him wrong

But then he goes on to say "Note that the exact value of these capacitors is not at all critical". I don't think that most people here would agree with that statement (value not critical) in the context of signal couplers. We would all get out slide rules and compute corner frequencies. Maybe that is where I got it in my head that he had something "wrong" in relation to typical thinking here.