Another 300B SE

[minhaj]

Here is 300B Se again but in a different flavor. Since I have Aikido
pre with 6CG7 input tube I have gain of 10 (or so) at pre-stage. So I am looking for a linear low z out single ended (loaded with a resistor/SS CCS) input/driver tube with low gain in the range of 3-10.

300B runs with 360V and older version of James as OT

So far I have tried as input/driver tubes, resistor loaded, they are:

2A3- with 5K plate and 2000 ohms cathod resistor, supply voltage 350

12B4A- with 8K plate and 1500 ohms cathod resistor, supply voltage 350

EL86T -with 8K plate and 560 ohms cathod resistor, supply voltage 350

So far 2A3 gives the best sound, NOW THE BIG question WHICH TUBE TO TRY NEXT??

NB. I use CMC fitted DC filament supply for 2A3 and 300B, AC filament for non DHT.

Let us share our ideas and experiences.

Regards

[kmaier]

Maybe try a 45 DHT? Bias it around 30-34 ma.

I've decided my next 300B prototype will be transformer coupled from the driver to the 300B.

Regards, KM

[minhaj]

I live in a country where tubes are in a short supply I have not seen a 45 as yet, whenever I can lay my hands on a pair I will try. 30-34 mA current for 45 as 300B driver seems a bit more though.


Another point, I have a pair of valvo Aa DHT Rp is too high as 300B driver which is 30K. I am thinking of using these tubes with SS mu follower. Can someone help me to fulfill my dream vide directing me to a simple mu follower with 1-3 mA current for Aa plate. In my country DN2540 is not available. Is it possible to design it with JFETS where the supply voltage is 350V-360V. I will be obliged.

Regards

[andyjevans]

2a3 is good. You can also try 31, 71A and 46. All on ebay. And 31 is cheap from US dealers.

Andy

[Miles Prower]

How about a source follower to drive the 300B? Most of the complaints about inadequate 300B performance seem to be related to inadequate grid drive. Since a MOSFET has a much lower Rd(on) and much higher g(m) than any VT, it also presents a very Lo-Z load to the grid, with more than enough current sourcing capability to take care of any grid current (power triodes tend to pull grid current even before Vgk actullay goes positive) and to charge up Ci + Cmiller + Cstray for good slew rate performance at the upper end.

[minhaj]

Dear Andy,

You do not need to be humble, 2A3 is not that good as 300B driver. I know cause I heard VT62 for a short time as 300B driver and that was way better. We can not buy online and import into the country (too many government formalities) else I could try 10Y, 71, 31 (as you suggested) before writing in the forum. Obliged for your help, pls tell me more.

Dear Miles,

I know you from the beginning, I feel good that you are trying to help me. I have some education and experience in tube circuit, zero for SS. Your point about using MOSFET source follower is well taken, but what about high capacitance of MOSFET? Can you give me some schamatic/drawing/direction in the same line?

Regards

[CLS]

I found this is a good reading:

http://www.tubelab.com/powerdrive.htm

[minhaj]

Hi CLS, thanks for the great link I read it earlier also I read it now
I understand some aspect I do not understand many, my limited knowledge/education!

Somehow I accept SS CCS/mu follower type thing at the top of the tube plate. In-between driver and output tube....it gives me uneasy feeling, I have big respect for Tubelab's work though.

Somehow I can manage few DN2540 and cascode them in mu follower style, is it good for Aa DHT at around 2 mA? Need your educated input to proceed. Will the mu follower act optimally at 2 mA?

Regards

[Miles Prower]

Originally posted by minhaj
Dear Miles,

I know you from the beginning, I feel good that you are trying to help me. I have some education and experience in tube circuit, zero for SS. Your point about using MOSFET source follower is well taken, but what about high capacitance of MOSFET? Can you give me some schamatic/drawing/direction in the same line?

Regards

With MOSFETs, the largest capacitance is the input: Cgs. The source follower solves that problem since the voltage at the source is following the voltage at the gate. Therefore, the current through that Cgs is very small. This a type of "bootstrapping" that reduces the effective value of the capacitance to a very small value. The main contributor to source follower input capacitance is the reverse transfer capacitance, which isn't all that high. That's what you look for: a small reverse transfer capacitance. It's not all that formidable to drive it properly.

See here for how I intend to use it as a power triode driver: Wolverine Main Schemo (Preliminary). The thing to keep in mind when using power MOSFETs is that these are very high gain devices, moreso than any VT, and they like to make RF if given half-a-chance. That means keeping all leads as short as possible and including a gate stopper. In keeping with the usual RF practice, make that stopper with a carbon composition resistor, and ferrite beads. This should keep it from taking off and oscillating, but be sure to o'scope to make certain that it's stable. I'd get a half-dozen of the MOSFETs since RF instability usually results in poofage, and you just might poof a few before getting it right. It's always nice to have some in the junk box even if you don't poof any.

As for how to design, that's even easier than VT design. Once you know what the static bias voltage is going to be, you can figure that the static gate voltage will be 4 -- 5Vdc more positive to forward bias. Figure on a drain current of about 20 -- 30mAdc, and calculate the source load resistor accordingly. Include enough Vdd so that the Vds stays high enough to prevent the device capacitances from going nonlinear. As Vds approaches the rail, MOSFET capacitances go pretty wonkey (see MOSFET data sheets) usually, 20 -- 30Vdc is enough at signal peaks.

This is really the best way to go, and the 845 is even nastier to drive than the 300B. Back in the good ol' days, they simply said t'hellwiddit and advised that the 845 be biased at -155Vdc, with the signal limited to 150Vp to just keep the Vgk negative at all times to avoid that grid current problem. These days, it's not necessary since any power MOSFET can slap that control grid silly.

Regardless of what silicon allergy some bottle heads have, it really is the best solution.

[andyjevans]

Out of interest, how do you import into Bangladesh?

- do you mark parcels "gift"?
- do you disguise small componants?
- anything else?

Seems surprising - we think of Bangladesh as a modern society. I'd like to hear more details!

andy