I agree that it is important to support things like AX if they provide a source of good information. Unfortunately, that is not usually the case. I had a subscription for about 5 years, and found that I wasn't interested in many of the articles they were writing and the ones that were interesting were things I already knew or were so far over my head without sufficient explanation that I found that the reading was not of much use to me. It is not that I don't want to support AX, but if I find that this board is a better source for information, inspiration and support in seeing projects through than AX is.
I just figured I would ask. I do have a friend who lives in Virginia who will buy this issue and scan me the article. I just wanted to see what SY's design looked like before having him go through the hassle.
I just figured I would ask. I do have a friend who lives in Virginia who will buy this issue and scan me the article. I just wanted to see what SY's design looked like before having him go through the hassle.
OK, let's look at that step network in some detail. We have 170-180V on the plate of the first tube, we want 118V on the grid of the second tube for reasons I explained in the article. So we know that the ratio R9/(R8 + R9) = 118/180 = 0.65. Now, this is one equation in two unknowns, so there's an infinite number of possible combinations. We would like to pick the resistors to be high enough that the plate of the first tube isn't loaded down very much, but not so large that stray capacitances and any grid current from the second tube will upset things.
Assume we're at midband where the capacitor has effectively bypassed R8. The first tube's plate load, then, is R9. The plate resistance of the first tube is something like 7k. So if we pick a value for R9 that's at least a hundred times greater, the load on the tube is negligible, and we will still get nearly the full mu as gain, i.e., we'll still be operating the first tube in nearly constant current loading.
1M is a nice, standard value that is known-good for the second tube (ECC88). It's a very common value for a grid leak, and since there's effectively 100% feedback in the second tube, any odd grid current is taken care of. So... we arbitrarily set R9 at 1M, which is more than 100x the plate resistance of the first tube. Plug 1M into the voltage divider equation and solve for R8. 1/(R8 + R9) = 0.65, so R8 = (1/0.65) - R9 = (1/0.65) - 1 = 0.52M. The closest standard value is 510k (0.510M).
Now, that value would work just fine. But in the 45 year tradition of my coffee-can engineering, I didn't have any good quality 510ks, but I did have some 680k. That would put the grid of the second tube at 108V, which is not very far from our target and barely changes the operating point of the second tube. So that's what I used.
In retrospect, I should have been a little clearer... In any case, the resistor values aren't very critical, and 510, 560, 620, or 680k will all work fine.
Assume we're at midband where the capacitor has effectively bypassed R8. The first tube's plate load, then, is R9. The plate resistance of the first tube is something like 7k. So if we pick a value for R9 that's at least a hundred times greater, the load on the tube is negligible, and we will still get nearly the full mu as gain, i.e., we'll still be operating the first tube in nearly constant current loading.
1M is a nice, standard value that is known-good for the second tube (ECC88). It's a very common value for a grid leak, and since there's effectively 100% feedback in the second tube, any odd grid current is taken care of. So... we arbitrarily set R9 at 1M, which is more than 100x the plate resistance of the first tube. Plug 1M into the voltage divider equation and solve for R8. 1/(R8 + R9) = 0.65, so R8 = (1/0.65) - R9 = (1/0.65) - 1 = 0.52M. The closest standard value is 510k (0.510M).
Now, that value would work just fine. But in the 45 year tradition of my coffee-can engineering, I didn't have any good quality 510ks, but I did have some 680k. That would put the grid of the second tube at 108V, which is not very far from our target and barely changes the operating point of the second tube. So that's what I used.
In retrospect, I should have been a little clearer... In any case, the resistor values aren't very critical, and 510, 560, 620, or 680k will all work fine.
Nelson Pass said:
For future reference, I make a point of retaining copyright, and I never
accept money for submissions.
That said, people need to buy a subscription to AX. It's the least they
can do for the efforts made for their benefit.
I became a subscriber over a year ago based on Nelson's recommendation. I've learned a lot, and the 2 part tutorial in 2008 on speaker measurements was worth it alone for the year.
-David
DaveM said:
I understand what you mean. I subscribe to aX and there is a lot of stuff over my head. And some articles just seem like it's information I'd never use. But, there is just a ton of good information in there, and I may not appreciate it at first, but I pick up things as I read more. Like when they published Dennis Colin's article "A Low Noise Measurement Preamp," in my extreme naivety, I thought "Who cares about that?"
But several months later I was finishing up a preamp I designed and it had a noise problem that was very irritating. And this was a well filtered supply. Of course, I go digging through Dennis Colin's article and learn about noise measurement, and build his Low Noise Measurement Preamp. And I found the noise source, and learned about noise measurements.
So for me at least, aX is an invaluable supplement to what I learn on the forums (or is it the other way around?). If you have some questions you can always post it on John Curl's Blowtorch thread like I did.
They answered some questions about noise for me. Or anywhere on the forum for that matter. I'm a just a big fan of aX, and I think it's one of best resources out there, for learning and expanding your knowledge about audio.
After re-reading my last post, I'd like to rephrase some things that I thought may have come across wrong. First of all I think everything I've read in aX is interesting and relevent to high end audio, and even though some stuff is above my understanding, that's the whole point - to learn something new! And they have lots of great audio projects to build, music reviews, and I really enjoy the letters section.
aX is my favorite audio publication, period! And I really don't understand why anyone on this forum would not want to subscribe. I'm also a huge fan of Dennis Colin and he has certainly opened up my eyes about many interesting things in high end audio and how to design high quality quality audio gear. Not to mention the outstanding audio designs he's published in aX, like the LP797 phono preamp. And there's all the great designs Pass has published in aX as well.
I mean the regular contributers to aX (Gary Gallo, Chuck Hansen, Erno Borblely, Joseph D'Appolito, Vance Dickason, Jann Didden, Reg Williamson and others including Nelson Pass!!!) are legends in high end audio. Who wouldn't want to learn from them?!
I'm not trying to advertise aX, so much, as I'm trying to defend an audio publication that I really enjoy and care about. I just thought some of my comments could have been taken negatively and that's most certainly not how I mean it.
aX is my favorite audio publication, period! And I really don't understand why anyone on this forum would not want to subscribe. I'm also a huge fan of Dennis Colin and he has certainly opened up my eyes about many interesting things in high end audio and how to design high quality quality audio gear. Not to mention the outstanding audio designs he's published in aX, like the LP797 phono preamp. And there's all the great designs Pass has published in aX as well.
I mean the regular contributers to aX (Gary Gallo, Chuck Hansen, Erno Borblely, Joseph D'Appolito, Vance Dickason, Jann Didden, Reg Williamson and others including Nelson Pass!!!) are legends in high end audio. Who wouldn't want to learn from them?!
I'm not trying to advertise aX, so much, as I'm trying to defend an audio publication that I really enjoy and care about. I just thought some of my comments could have been taken negatively and that's most certainly not how I mean it.
Good catch.