At first glance one would think the input capacitance of a Mosfet would be an issue but it's not. Here is why. (source follower configuration)
Cin = CGD + CGS(1 - AGS)
The only way it would be a problem is if the gain from Gate to Source was exactly 1, which it wouldn't be. It would be close to 1, lets say .99,
Cin = CGD + CGS * .01 So for a Mosfet with input C of 1200pF it would only be 12pF
Cin = CGD + CGS(1 - AGS)
The only way it would be a problem is if the gain from Gate to Source was exactly 1, which it wouldn't be. It would be close to 1, lets say .99,
Cin = CGD + CGS * .01 So for a Mosfet with input C of 1200pF it would only be 12pF
Yes, the input capacitance may be not an issue with source follower in most cases, but for an single MOSFET CCS the output capacitances (Coss + Crss) will play an role. An MOSFET with high Ciss probably will also have an high Coss+ Crss, and not very linear...
In cascode configuration, these capacitances are bootstrapped within CCS devices themselves, see CCS performance measurments
But I like an "elegant" solution with low capacitance devices, since single MOSFET CCS is less prone to RF oscillations than cascode stages, from my own experience.
PS.: if stage gain drop to 0.9 from original 0.99 in source follower applications, the 12pF becomes 120pF in an 1.2nF device, and this is common in amps with lateral MOSFETs at output.
In cascode configuration, these capacitances are bootstrapped within CCS devices themselves, see CCS performance measurments
But I like an "elegant" solution with low capacitance devices, since single MOSFET CCS is less prone to RF oscillations than cascode stages, from my own experience.
PS.: if stage gain drop to 0.9 from original 0.99 in source follower applications, the 12pF becomes 120pF in an 1.2nF device, and this is common in amps with lateral MOSFETs at output.
Since µTracer oscillated with E810F, I re-do my measurements. In pentode mode the crazy E810F don't oscillate, so I can estimate the parameters from here. 41mA/V with 19mA anode is crazy for a small tube...
Attachments
Well, what the helicoid, why not throw in this too, for posterity. Download the 1997 Valve Magazine and in the September Issue Volume 4 Number 9 on page 7 is another driver. Not one stage but one tube direct coupled.
The whole series archived is here.
The whole series archived is here.
What about a 2A3 driving a 300B ? I was thinking op point for the 2A3 around 40mA with135V on the anode
I used 2x ECC83 in // SRPP preamp stage for my ULTIMATE 300B : it's what I found working the best for a single-stage preamp driving a 300B. And it uses a very common tube...
For 0.8V RMS input, the amp gives 8W RMS output on 8R load. The gain of the preamp stage is 62.
A+!
An externally hosted image should be here but it was not working when we last tested it.
For 0.8V RMS input, the amp gives 8W RMS output on 8R load. The gain of the preamp stage is 62.
A+!
I'm thinking large triode of the dissimilar dual triode 6EM7, 35mA or so, 1:1+1 interstage to PP 300b
The small triode section doesn't get a lot of love, so I am thinking of just leaving it unused and maybe using a 6J5 . If it gives me enough gain.
TBH I symphatize more with the low power section than with the power section of a 6EM7. See the curves, is very good for the low power section. Trouble is, has high rp and mu, so gain will be excessive if using it.
And will be at a DHT style! Like with 4P1L. I'm using the 4P1L now, instead of E810F. I compensated the low gain (comparing with the E810F) with a 1:4 input trafo, so the net gain is close.
Scary low THD with low level high order H (the important part from a HD measurement). Like Ale Mogliaa published in his site, is a very clean driver.
TBH I preferred the 4P1L sound (and saved the E810F for another project), due to the cleanliness (this time, technical and sound results are "in phase" 😉😀 )
I believe the 2A3 sound in same cleanliness ballpark, but results and preference may vary
Why not? Is there any way to get around the shortcomings? I just happen to have a pair of 26 and am wondering what to do with them.
John
John
DIYbras, the problem with going 1:4 with an insterstage is that the driving impedance the 300B sees is 16 times bigger. Instead of (to assume) 1,2k it will be 19,2k. How is your frequency response?
If I had to use the 4P1L, I'd bet on a three stage design with a 1:1 ITT.
If I had to use the 4P1L, I'd bet on a three stage design with a 1:1 ITT.