AudioAngel said:
Yes they are. The 12AT7 really isn't very good for cascodes. Its g(m) collapses too quickly with diminishing Ip. You'd be better off trying a 6BQ7A. This dual was designed for use as an RF cascode, but it has pretty decent linearity for audio as well (although it likes rather highish Vpk's for good results). It's one of those common TV t00bz, and shouldn't be too difficult to get. It has a high g(m) that doesn't collapse very fast.
Although I'd prefer a small signal pentode. The A Number One problem with cascodes is that the bottom tube operates into a very low load. For linearity, this is not how triodes operate best. You'd get less distortion from a pentode like a 6SJ7, and still avoid the Cmiller problem.
You see cascodes in guitar amps where linearity and low THD doesn't have the premium that it does for hi-fi.
If you still want to go that route, I have a "white paper" that explains it quite well. Can send via e-mail.
The real problem is that Raphaels Power Amp is an Aleph X
with balanced input, so he does need not a cascode, whatever variation.
His need is a diffamp an possibly Cathode followers as a Buffer.
For the 20 K Input impedance of the Aleph X I would design a single gain stage like the BAT preamps with 6H30, ECC 99 or something similar.
Reinhard
with balanced input, so he does need not a cascode, whatever variation.
His need is a diffamp an possibly Cathode followers as a Buffer.
For the 20 K Input impedance of the Aleph X I would design a single gain stage like the BAT preamps with 6H30, ECC 99 or something similar.
Reinhard
Hi DougL
I wrote that 2 sentences in German, explaning there is amazon.de where you can buy the book for about 40 euro, including postage costs, neither expensive or hard to get.
I was about to ask why Raphael wants a cascode: just as an exercise, or for actual line stage use? What I gather from tha cascode is that it is a very high gain stage, unnecessary for line stages. That makes Reinhard's post a good one, explaining the situation and giving hints for Raphael's linestage. It would be even possible to make the diffamp with cascodes, and than buffer the whole thing with cathode followers (and severely knock the gain down at the output)...that would be both a learning exercise as an actual line stage.
Erik
I wrote that 2 sentences in German, explaning there is amazon.de where you can buy the book for about 40 euro, including postage costs, neither expensive or hard to get.
I was about to ask why Raphael wants a cascode: just as an exercise, or for actual line stage use? What I gather from tha cascode is that it is a very high gain stage, unnecessary for line stages. That makes Reinhard's post a good one, explaining the situation and giving hints for Raphael's linestage. It would be even possible to make the diffamp with cascodes, and than buffer the whole thing with cathode followers (and severely knock the gain down at the output)...that would be both a learning exercise as an actual line stage.
Erik
Hello!
Concerning the book. I've found a bookstore. They will import a new book from UK and I've only to pay the normal price without shipping costs.
Erik's post sounds good. I'd like building a differential amp with cascodes because then I can use that, what I've tried in a theoretical way by using PSpice. I don't like it to think about something only theoretically.
I think that is a good way. Maybe later I'll build another thing but then I can reuse the tubes so there is no wasting money.
Raphael
Concerning the book. I've found a bookstore. They will import a new book from UK and I've only to pay the normal price without shipping costs.
Erik's post sounds good. I'd like building a differential amp with cascodes because then I can use that, what I've tried in a theoretical way by using PSpice. I don't like it to think about something only theoretically.
I think that is a good way. Maybe later I'll build another thing but then I can reuse the tubes so there is no wasting money.
Raphael
Hi !
To sum it up, a cascode has several Problems:
a very high ra (=rout)
the lower tube sees a very low RL, but current can be increased by a paralleling the top tube with a resistor.
Thats the main Problem, because the lower tube should be very linear and hast to work in thet low RL.
I wouldn't go with a pentode, because one advantage of the cascode is that it is free of any noise, but pentodes can be noisy, especially in sesitive preamps.
And, off the records, the ECC81 is absolutly not linear !! It's a nice tube, high mutal conductance, medium high µ, but IMHO not suitable for linear gainstages.
Its has the same system like a ECC85.
IMHO ECC81/85 belong only in vintage German tube radios.
Manta
To sum it up, a cascode has several Problems:
a very high ra (=rout)
the lower tube sees a very low RL, but current can be increased by a paralleling the top tube with a resistor.
Thats the main Problem, because the lower tube should be very linear and hast to work in thet low RL.
I wouldn't go with a pentode, because one advantage of the cascode is that it is free of any noise, but pentodes can be noisy, especially in sesitive preamps.
And, off the records, the ECC81 is absolutly not linear !! It's a nice tube, high mutal conductance, medium high µ, but IMHO not suitable for linear gainstages.
Its has the same system like a ECC85.
IMHO ECC81/85 belong only in vintage German tube radios.
Manta
As I told you before a cascoded Diffamp is the wrong decision for a line amp.
If you don´t trust me look at one of Morgan Jones books. For Linestages he is always using a plate loaded double Triode with
low gain like 6SN7 or 6BX7.
Or look at some industrial designs, BAT, ARC,
Or look at John Broskies tubecad.com.
Regards Reinhard
Not to speak about the difference in sound of a cascode or a plate loaded triode.
If you don´t trust me look at one of Morgan Jones books. For Linestages he is always using a plate loaded double Triode with
low gain like 6SN7 or 6BX7.
Or look at some industrial designs, BAT, ARC,
Or look at John Broskies tubecad.com.
Regards Reinhard
Not to speak about the difference in sound of a cascode or a plate loaded triode.
12AT7 in cascode, etc.
Mein Herr!
With "textbook" resistive loading, the 12AT7/ECC81 is not linear. CCS loading changes the situation radically. Also, the distortion spectrum of the 'T7 is an advantage in a PP amp. The 'T7 makes a wonderful LTP phase splitter, as the net distortion spectrum is a VERY pleasing "waterfall".
IMO, the 12AT7 is a poor choice in line stages, where it has too much gain, and a poor choice in SE circuits, where its distortion spectrum is unfavorable.
----------------------------------------------------------------------------------
All,
Cascoding the 12AT7/ECC81 is of interest to me. As a H/K Cit. 2 owner, the looming demise of the 12BY7 is most important. A hybrid cascode with a FET on top and a 'T7 with its sections in parallel on the bottom is a possibility to replace the 12BY7 as the voltage gain stage.
I'd appreciate feedback on the hybrid cascode idea.
Mein Herr!
With "textbook" resistive loading, the 12AT7/ECC81 is not linear. CCS loading changes the situation radically. Also, the distortion spectrum of the 'T7 is an advantage in a PP amp. The 'T7 makes a wonderful LTP phase splitter, as the net distortion spectrum is a VERY pleasing "waterfall".
IMO, the 12AT7 is a poor choice in line stages, where it has too much gain, and a poor choice in SE circuits, where its distortion spectrum is unfavorable.
----------------------------------------------------------------------------------
All,
Cascoding the 12AT7/ECC81 is of interest to me. As a H/K Cit. 2 owner, the looming demise of the 12BY7 is most important. A hybrid cascode with a FET on top and a 'T7 with its sections in parallel on the bottom is a possibility to replace the 12BY7 as the voltage gain stage.
I'd appreciate feedback on the hybrid cascode idea.
On the record, absolutely not true. With a 120k load and 1.2mA idle current, as used in my power amp project, the ECC81 clocks in at about 0.03%THD at 15VRMS output. That's pretty darn linear.
Eli, why not use a FET on the bottom, and 12AT7 on top? I would think something like an NTE458 with some source degeneration ought to work very well.
Re: 12AT7 in cascode, etc.
Eli,
I went to a hamfest yesterday and nearly tripped over all the "NOS" 12BY7s for sale. My usual tube source has NOS for about $8 each. Why sully this classic amp with a FET? For the cost of a night out to a good restaurant and a movie you could stock up on enough 12BY7s to keep your Cit. II cooking for the rest of your life and maybe enough to allow your grandkids to enjoy this amp too. Of course, they'll have to recap the amp
Agree on your comments about the 12AT7.
Eli Duttman said:
Eli,
I went to a hamfest yesterday and nearly tripped over all the "NOS" 12BY7s for sale. My usual tube source has NOS for about $8 each. Why sully this classic amp with a FET? For the cost of a night out to a good restaurant and a movie you could stock up on enough 12BY7s to keep your Cit. II cooking for the rest of your life and maybe enough to allow your grandkids to enjoy this amp too. Of course, they'll have to recap the amp
Agree on your comments about the 12AT7.
Hi !
Grmpf.....
My experiences are different wir the ECC81
I used it as driver/gainstage for my EL38 amp, and it sounded worse.
I swapped to the ECC88 and the sound was awesome, as usual.
Of course, the resistor values were changed, I didn't simply swap the tubes.
When I had a look an the curves, the situations was obvious.
ECC81:
And
ECC88
That was my experience with the ECC81. I'm glad about your comments about that tube. Maybe I just used it at the wrong place in my circuit...
Manta
Grmpf.....
My experiences are different wir the ECC81
I used it as driver/gainstage for my EL38 amp, and it sounded worse.
I swapped to the ECC88 and the sound was awesome, as usual.
Of course, the resistor values were changed, I didn't simply swap the tubes.
When I had a look an the curves, the situations was obvious.
ECC81:
An externally hosted image should be here but it was not working when we last tested it.
And
ECC88
An externally hosted image should be here but it was not working when we last tested it.
That was my experience with the ECC81. I'm glad about your comments about that tube. Maybe I just used it at the wrong place in my circuit...
Manta
SY said:
Hey, I like Allen. I'd go easy on him. No groin kicks or Adam's apple punches, or anything like that. He does use a FET in the one ALMOST excusable place: at the base of a cascode (with an ECC88 upstairs) for MC preamplification. Here the nasty gate capacitance is being driven by a stump-pulling 4 or 10 ohm source (the MC), and the Miller effect is ALMOST eliminated by the drain being held quasi-fixed by the upper cathode. Low noise is the goal and it works well for that.
Wishing I could join your guys at ETF...
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.