If not using transformer coupling, a direct-coupled CF or SF driver for a PP output stage would seem to offer a number of advantages. It can be used to prevent blocking distortion of pentodes, to supply sufficient drive current to overcome miller capacitance of triodes, or to serve as a low impedance source to drive Class AB2.
Seems like a good idea to use this type of driver in every PP design, but is there any downside to doing this?
Seems like a good idea to use this type of driver in every PP design, but is there any downside to doing this?
Blocking distortion AFAIK is caused when the grid of the OP tube (in this case) is fed by a coupling cap and goes positive, becoming effectively a diode. It can happen on transients and it takes a large current from the coupling cap, which then takes some time to recover. It's not exclusive to pentodes but it is more severe, because they take much longer to recover than triodes (I think).
This kind of distortion can be counter-acted, to some extent, by using a high value grid stopper on the OP tube or it can be completely avoided by using DC coupling (i.e. no coupling cap at all).
I wasn't aware of the recovery time difference. As far as I know, they're pretty much the same- it's the RC time constant which determines recovery time.
OK, you know the advantages of a follower (I think source followers are even better in that application, but I'd probably be disinvited from the European Triode Festival this year for saying that). The feedback question Douglas raised is probably not an issue- follower bandwidths are ridiculously good. One more advantage is the ability to drive appropriate tubes in AB2, which is good for another dB or two of power.
Disadvantage is parts count and heat. For a fixed-bias triode output stage, I'd use a follower driver whenever possible. For a fixed-bias pentode that won't be driven AB2, I'd probably just use a big grid stopper instead.
OK, you know the advantages of a follower (I think source followers are even better in that application, but I'd probably be disinvited from the European Triode Festival this year for saying that). The feedback question Douglas raised is probably not an issue- follower bandwidths are ridiculously good. One more advantage is the ability to drive appropriate tubes in AB2, which is good for another dB or two of power.
Disadvantage is parts count and heat. For a fixed-bias triode output stage, I'd use a follower driver whenever possible. For a fixed-bias pentode that won't be driven AB2, I'd probably just use a big grid stopper instead.
Ray,
I thoroughly RECOMMEND a Source Follower drive of output tube grids.
Advantages are, as you state, freedom from blocking distortion.
It also allows use of low Rg1 resistor values which helps with bias stability too.
Waveborn has obviously been reading Allen Wright. He is showing a MOSFET version of Allens SLCF. With triodes it is good to current source load the cathode follower (Constant Current mode operation) and Bootstrap the anode voltage (Constant Voltage mode operation).
With a MOSFET Source Follower I might consider the current source load (although I'm using simple resitive load with good results) but as long as you keep minimum Vgd to say +25V with maximum signal swing, then the bootstrap is totally un-necessary.
If you want to go the "No Sand" route, then use a pentode cathode follower with the screen bootstrapped to the output - with or without a current source load.
Cheers,
Ian
I thoroughly RECOMMEND a Source Follower drive of output tube grids.
Advantages are, as you state, freedom from blocking distortion.
It also allows use of low Rg1 resistor values which helps with bias stability too.
Waveborn has obviously been reading Allen Wright. He is showing a MOSFET version of Allens SLCF. With triodes it is good to current source load the cathode follower (Constant Current mode operation) and Bootstrap the anode voltage (Constant Voltage mode operation).
With a MOSFET Source Follower I might consider the current source load (although I'm using simple resitive load with good results) but as long as you keep minimum Vgd to say +25V with maximum signal swing, then the bootstrap is totally un-necessary.
If you want to go the "No Sand" route, then use a pentode cathode follower with the screen bootstrapped to the output - with or without a current source load.
Cheers,
Ian
gingertube said:
No. Wavebourn has obviously a diploma of Engineer Designer and Technologists of Radio Electronics Equipment from famous TIASUR.
I just loaded a source follower on a CCS and bootstrapped it with one more source follower so the main SF sees nearly constant SD voltage and current. As the result, distortions caused by variations of transconductance and capacitances are minimal.
Obviously Allen Wright had similar purpose in mind...
Also, if I remember right, George (Tubelab.com) used something similar. Sometimes very similar tunes may be made of 12 notes available for composers. 12 notes. How many elements are available for electronics designers?
SY, I should hace said that a pentode OP stage has a much harder time recovering from blocking than a triode stage because of the much greater amount of NFB used in the former case. The triode will clip but only momentarily.
I suppose what I really want to know is if there is any reason not to use direct coupled cf/sf drivers, apart from economy.
I suppose what I really want to know is if there is any reason not to use direct coupled cf/sf drivers, apart from economy.
ray_moth said:
Screen grid is a good sensor of overload: I usually put Vactrol there to sense current peaks by it's LED, and reduce amplification factor by it's opto-resistor. Also, bright red LED's grab my attention to too much input level. Triodes don't offer such a convenience, but I drive them with 1'st grid current, and a source follower does that fine.
hey-Hey!!!,
I have gotten good results with using a cascode pair of MOSFET's instead of a single every time I have needed such a thing. I even tried them as grounded source amplifiers. It seems to deliver the near-constant capacitance of a pentode without the need( and trouble ) of g2 current.
cheers,
Douglas
I have gotten good results with using a cascode pair of MOSFET's instead of a single every time I have needed such a thing. I even tried them as grounded source amplifiers. It seems to deliver the near-constant capacitance of a pentode without the need( and trouble ) of g2 current.
cheers,
Douglas
#1
Hello ray_moth
This is one advantage of push pull against single ended.
You'll get the blocking distortion at the grid of the cathode follower.
Thus it is not preventing blocking distortion.
There are a lot of good reasons not to go to AB2 mode, especially for triodes.
If you can live with the crossover distortion from AB1 to AB2 the follower is a good solution.
Kind regards,
Darius
Hello ray_moth
A transformer does not solve these problems and Miller capacitance does not cause any problem in a PP design.
This is one advantage of push pull against single ended.
You'll get the blocking distortion at the grid of the cathode follower.
Thus it is not preventing blocking distortion.
There are a lot of good reasons not to go to AB2 mode, especially for triodes.
If you can live with the crossover distortion from AB1 to AB2 the follower is a good solution.
Kind regards,
Darius
Hi Darius,
I had the opposite impression. Ithought that that triodes had more to offer by going into AB2, in terms of extra power, than pentodes.
In any case, I wouldn't want to run in AB2 for any length of time, but it does seem to offer a bit of useful headroom for transients. Better than ugly clipping, anyway.
I don't follow that. Triodes have significant Miller AFAIK, whether they're SE or PP. It's pretty well negligible for pentodes, though.
That can happen, true, but it's easy to avoid.
I had the opposite impression. Ithought that that triodes had more to offer by going into AB2, in terms of extra power, than pentodes.
In any case, I wouldn't want to run in AB2 for any length of time, but it does seem to offer a bit of useful headroom for transients. Better than ugly clipping, anyway.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.