coupling caps?

[Ty_Bower]

What do you mean by "corner frequency?"

These guys explain it better than I ever could. Scroll down to the graph titled "Frequency Response of a 1st-order Low Pass Filter". It's a little less than halfway down the page.

http://www.electronics-tutorials.ws/fil ... ter_2.html

Our filter in this case is actually a high pass. While the phase shift is different for the high pass vs. the low pass, the frequency response works pretty much the same (just turned around). Basically, at the corner frequency of a first order filter you are 3dB down. Rolloff actually begins earlier than that, so you want the corner well outside the audible range.

[fran604g]

Thanks Ty, for the link,

I won't pretend to understand the math part, but, like a crossover network RC circuit...I'm assuming that at some point the 0.1uF value was used in the prototypes and it must have been discovered later that the value could be changed to .22uF to allow even lower frequencies to pass and still control the undesirable frequencies that needed to be blocked, maybe hum for instance? Or am I way off base? Maybe the .22uF was chosen for other reasons?

Fran

[Ty_Bower]

You need a capacitor to block the high voltage DC found at the plate of the 6SL7 from getting to the grid of the output tube. There are some designs which direct couple one stage to the next, but the GSG is not one of them. The voltage rating of this cap must be at least as high as the voltage normally found at the plate of the 6SL7 while operating. Keep in mind there will also be some voltage swing (audio signal) present there. It would be wise to rate the cap as high as the maximum voltage that might be present when the power supply is fully unloaded (cold output tubes not conducting).

In a design such as the GSG, you want the cap large enough to get good low frequency extension. If you try to go too large, the cap may be unnecessarily expensive, or may not physically fit on the board. You might also want to size the cap small enough that you avoid delivering low frequencies that your output transformers cannot handle. Otherwise, I don't think you can hurt anything by making the cap "too big". Making it too small will cause the bass response to suffer, but I'm confident that even a 0.1uF part is more than adequate.

In other designs where substantial feedback is utilized, it is even more important that you avoid saturating the output transformers with excessive low frequency material. This is especially true if feedback is taken from the secondary winding of the OPT. You may also need to consider "the dominant pole" in the amplifier. Each cap in the signal path presents a phase shift, and too much shift in the wrong place will turn your nice amplifier into an oscillator. I don't understand how one calculates which point in the circuit is the dominant pole, nor do I even comprehend how one chooses which pole should be dominant. As far as I am concerned, that is an issue I hope the designer addressed while he was evaluating the stability of the circuit.

[fran604g]

Thanks for the explanation.

I hope I'm not giving you the impression that I don't like this amp, I do, very much. I expect that it will be my one amp for a very long time. I would love to have the capacity to understand the complexities of my hobby more than I do, and because of people like yourself, I am learning.

The one thing that intrigued me the most about this particular amp is it's flexibility. I love the idea of running 300B's, 45's, any of the "old" amp tubes. To me the sound is exactly what I hoped it would be with my Altec 604-8G's.

I just plain got lucky to see this amp on A'Gon at a price point that was, with all the tubes that came with it, more economical than buying the PCB and building one myself, which is exactly what I had planned early in March.

The .1 caps that are in it certainly work well enough that I probably wouldn't have ever noticed if it weren't for my incurable curiousity...LOL

Thanks again for your help, it's greatly appreciated!
Fran

[lth1]

Fran, here is a three way switch that I frequently find myself using; it alows you to install two different caps. The choices are, cap 1, cap 2, and then cap 1 + 2. Perfect for comparing three different values, or having one cap as a bypass and evaluating its effect.

Cordially,
Lee Hankins

http://glass-ware.stores.yahoo.net/ouseswandpcb.html

[lth1]

Also, picture of my installation of large teflon caps in the GSG. The GSG can use 6A3's or 6B4G's.

Lee

https://picasaweb.google.com/108890459358226980635/20110511?authkey=Gv1sRgCLX20dyO3o6glgE#5605670056751957842

https://picasaweb.google.com/108890459358226980635/20110511?authkey=Gv1sRgCLX20dyO3o6glgE#5605670097839242562

[Shannon Parks]

Since the GSG is zero feedback, one can use about any value coupling cap they have (with a 400V DC voltage rating) and not worry about screwing up any stability. For example, you could use a .047uF and the LF -3dB would still be 7Hz (LF -3dB = 1/(6.28*.000000047*475000). So that means you can plop in all kinds of different caps, particularly the nifty Russian teflons. Have fun!

Shannon

[fran604g]

Fran, here is a three way switch that I frequently find myself using; it alows you to install two different caps. The choices are, cap 1, cap 2, and then cap 1 + 2. Perfect for comparing three different values, or having one cap as a bypass and evaluating its effect.

Cordially,
Lee Hankins

http://glass-ware.stores.yahoo.net/ouseswandpcb.html

I like this idea very much,I can see doing this to mine.
Thanks,
Fran

[fran604g]

Also, picture of my installation of large teflon caps in the GSG. The GSG can use 6A3's or 6B4G's.

Lee

https://picasaweb.google.com/108890459358226980635/20110511?authkey=Gv1sRgCLX20dyO3o6glgE#5605670056751957842

https://picasaweb.google.com/108890459358226980635/20110511?authkey=Gv1sRgCLX20dyO3o6glgE#5605670097839242562

I came across your post of this at some point. how do you like them?

Also, I was just telling my wife the other day how I would love to set mine up for multiple tubes...she sneared at me and didn't want to hear about it LOL
How do you switch tubes, just by removing and changing cold? I don't see a switch in the pics? I see you have the sockets wired directly wired to the board.
Oh, and by the way VERY IMPRESSIVE work!

[fran604g]

Since the GSG is zero feedback, one can use about any value coupling cap they have (with a 400V DC voltage rating) and not worry about screwing up any stability. For example, you could use a .047uF and the LF -3dB would still be 7Hz (LF -3dB = 1/(6.28*.000000047*475000). So that means you can plop in all kinds of different caps, particularly the nifty Russian teflons. Have fun!

Shannon

Thanks Shannon,

I can see having a lot of fun with this amp, as I mentioned somewhere, I can see buying a second board and doing a different output like Ed and others have done. The sky's the limit! (well that, and $$$ lol)

Fran

[lth1]

Third try at posting this response, frustrating being old and not very computer savvy, much shorter this time and no preview.

Fran, I do like the Russian teflon caps, the .22 is the only value that I have used in the GSG. V-Caps and other teflons do sound better, but they are 20-30 times more expensive and I would rather spend the difference in areas where I can actually hear and result in more bang for the buck.

You are correct that the tube sockets are directly connected to the board and that the tubes are changed cold. A switch could be put on the 4 pin socket so that a 2A3 could be added to the mix. I have not heard much difference in the 6A3's or 6B4G's that I have tried, but seemed to have settled on a pair of GE 6A3's or Tung-Sol 6B4G's.

In the GSG I like the sound of solid state rectification via the Weber SS rectifier over the 6AX5 tube. The greatest change IMHO results from tube rolling the driver tube. In new builds I have gone to the 12 volt version of the SN7/SL7 family of tubes, again a bigger bang for the buck. They are suspose to be identical except for the filament voltage, but to these old ears the 12SN7/12SL7's sound better, probably due the subconsious reasoning that I have save a considerable amound of money by switching to the 12 volt family, tube rolling can get rather expensive.

Hope this helps
Lee Hankins

[Ty_Bower]

Third try at posting this response, frusting being old and not very computer savvy, much shorter this time and no preview.

I've been frustrated myself sometimes while trying to post on some forums. I've developed the habit of first composing my message in Notepad or Wordpad or something, then copy & pasting it into the forum editor when I am finished. It spares the agony of when your browser crashes or freezes up while trying to make a post. All your work typing is not lost in the forum editor.

[fran604g]

Hi Folks,

Well, be it a waste of time, energy and money or not;

[elbinster]

I set up a G*S*G to run multiple tubes through a switch instead of tube removal. Looking back, just wiring in multiple circuits and removing the unused tubes would be about seven hundred million times easier. It's cool to be able to switch, but I've only done a hot switch once (actually, my wife did it before she saw my look of shock and horror). Circuit survived, but it sounded painful.

Here's my P2P wiring first draft (note the yellow-electrical-taped K40y coupling caps) near the top. The three-deck switches are Soviet and seem up to the task of switching B+ voltage and filament current.



Not-so surprisingly, some portions of this wiring nest have started to cause problems, but luckily the 2A3 output stage is independent from the 6B4G/300B, so if one gives me trouble, I can switch to the other in a few minutes time.

[fran604g]

elbinster,
I hope that someday I can put together something on my own...I guess at least I finally started messing around.

My Dad worked at General Dynamics here in Rochester before they moved to the west coast in 1968. He was a military radio systems test engineer. I think I was taught how to solder when I was 5 or 6 yrs old, built a Morris code Oscillator circuit with speaker and key for my first project. Problem was, he was a pretty cruel teacher that liked to drink, so, by the time I was 15 or 16 I stopped helping him build things and had lost interest in listening to him long before then.

Needless to say, I would be much further ahead had things been a little different, but I've always had the spirit to build...

I like your style...