JoshK said:
hey-Hey!!!,
The pentode-like construction is governed by gm*R_load=gain. realized gm at idle point of the lower element, and the loading resistor, usually termed 'plate load', even if it is connected to the drain...
As far as delivering current, its output Z is ~R_load.
cheers,
Douglas
jon_010101 said:
There is also the 12AU7>6AU6>6V6 version, both PP and PPP (SP20 by RCA perhaps?). They've only got the plate to driver cathode FB loop, and not plate-grid around the finals like Jon's includes. I'd be fairly confident that we could elimnate the gNFB loop and still maintain excellent performance; the SP20 is the best off-the-rack commercial amp I have ever heard.
cheers,
Douglas
JoshK said:
It is a definite no. The A Number One problem with cascodes is the unusually high output impedance. This requires that the load be a Hi-Z, Lo-C load if you are to realize the cascode's bandwidth. I did a project that used a hollow state cascode. When connecting the o'scope directly to the output, it looked like the slowest thing I'd ever seen. 1.0KHz square waves had badly tilted tops, and the Fh barely made it to 19KHz (3.0db(v) point).
I knew that was ridiculous, and was caused by the capacitance of the o'scope probe, cable and input capacitance of the vertical deflection amps. That required a cathode follower with its much friendlier input. O'scoping the output of the cathode follower revealed the true picture: No tilted square waves, and an Fh= 117KHz (3.0db(v) point).
Connecting directly to the finals is likely to be a big disappointment since the grids of most finals isn't a very friendly load.
SY said:
That was my second thought (inserting a SF). Thanks for confirming my suspicions.
So let's take for way of example SY's screen driven 6LF6 amp. SY hinted earlier that if he were to do it all over, he would probably use something like a triode strapped D3A > 6S4A > SF > 6LF6's screens. Now instead let's consider, ts'd D3A/FET cascode > SF > 6LF6's screens.
My questions would be could you acheive enough gain with the cascode to do away with the extra gain stage? Which would likely perform better (guess)? The latter seems more ameniable to plate to plate feedback rather than a global loop.
Miles Prower said:
I'd say this is the perfect example of attempting to squeeze out gain in favour of bandwidth. The triode output Z is dominated by plate Z( with load a few multiples of it), and pentodes ( and cascode)being dominated by plate load( a small fraction of 'plate' Z ).
So...drop the plate load to deliver acceptable bandwidth and get gain by delivering enough gm.
cheers,
Douglas
Hi all.
Since this is about 'modern' tube circuits, there is nothing wrong in pointing out that there are continuing elements. The use of DC feedback I can also think of Stu Hegeman's Citation II. Personally I am no longer in favour of a technique I used to use, feedback tends to turn tube amplifiers into voltage mode and I have moved away from that.
But if I may use an analogy, cars used to have four wheels, they still do. But there is a huge gulf between modern cars and cars of yore. Execution of design and components show new levels of innovation. I believe that helps us to define what is modern.
With that in mind, there was a reason I posted this schematic:
Let's concentrate on the input stage:
Now compare this to a certain preamp:
I wonder if you can see the point I am making? Look carefully at the left side of the schematic.
The Vacuum State's Phono Stage is Moving Coil capable, able to take the smallest signals in audio and yet that same input stage can also, with minor changes in DC bias and increased HT also swing many hundred volts cleanly (despite some of the above comments). The fet and tube types are identical, 2SK147 (or equivalent) and 6922 - no other change necessarry.
This is an extraordinary building block. Did anything like this exist in the Golden Age of the 50's or earlier?
Joe R.
PS: Look at the Line Stage gain stage, it is the all-tube version of same. So that answers some other comments?
Since this is about 'modern' tube circuits, there is nothing wrong in pointing out that there are continuing elements. The use of DC feedback I can also think of Stu Hegeman's Citation II. Personally I am no longer in favour of a technique I used to use, feedback tends to turn tube amplifiers into voltage mode and I have moved away from that.
But if I may use an analogy, cars used to have four wheels, they still do. But there is a huge gulf between modern cars and cars of yore. Execution of design and components show new levels of innovation. I believe that helps us to define what is modern.
With that in mind, there was a reason I posted this schematic:
An externally hosted image should be here but it was not working when we last tested it.
Let's concentrate on the input stage:
An externally hosted image should be here but it was not working when we last tested it.
Now compare this to a certain preamp:
An externally hosted image should be here but it was not working when we last tested it.
I wonder if you can see the point I am making? Look carefully at the left side of the schematic.
The Vacuum State's Phono Stage is Moving Coil capable, able to take the smallest signals in audio and yet that same input stage can also, with minor changes in DC bias and increased HT also swing many hundred volts cleanly (despite some of the above comments). The fet and tube types are identical, 2SK147 (or equivalent) and 6922 - no other change necessarry.
This is an extraordinary building block. Did anything like this exist in the Golden Age of the 50's or earlier?
Joe R.
PS: Look at the Line Stage gain stage, it is the all-tube version of same. So that answers some other comments?
Josh, I think the D3a is wasted on the top of a cascode. On the plus side, cascoding gets rid of the D3a's huge Miller capacitance. On the minus side, the cascode's performance is almost entirely determined by the bottom device- a 12AT7 would probably work just as well and not need a lot of current.
jon_010101 said:
And thanks to SY for posting it (I said I'd like to look at it) and also your link Pete's site PDF file - I have in on my screen right now. There is no question that we can learn from the past, especially as an amateur historian (yeah, I like the History Channel
).But do we simply copy these ideas and live in the past, copy those schematics verbatim or do we listen to what the professor says and use that to move on and do new things that we could not have done if we had ignored the past?
If I have to think of one single modern tube development, it has to be the CCS.
Joe R.
It look similar to my Pyramid-VII: I come to conclusion that Edcor transformers need to be driven by output stages with local feedback, so pentode drivers are used. Some details are different, like I have a direct coupling from Concertina to LTP, and a SS screen VR, but main principles are similar: pentode input, Concertina, LTP drivers, local feedback in output stage, regulated screen grid voltage.
"Everything new is well forgotten old"
"Everything new is well forgotten old"
Joe Rasmussen said:
Kill joy!
What I should have said is the prevalance of CCS.
Since they are basically power supply in most cases (especially as long tail diff circuits) I have no problems using them with tubes and still call it such.
I'm just old and cynical.
Ike's idea was right, but the implementation wasn't. Once I started using better CCS circuits (mostly with Morgan's inspiration), it became evident that solid state CCS was clearly the way to go.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.