If I understand this right these new Semisouth Jfets can handle some pretty high voltage. What do you guys think about driving a 300B, 2a3 or possible a 847 tube with one? Is there something I am missing with this idea?
Interesting read, something I may try at a later date but not quite what I am talking about. What I am talking about would be an amp with a jfet directly driving a tube.
Im not knocking your idea, but isn't the virtue of tubes the linear voltage amplification and that of solid state devices the ability to deliver current? What is your motivation for doing it the other way round, apart from "because you can"😀
Correct me if I'm wrong, but I thought that both valves and FETs are essentially voltage controlled current sources.
They don't look very linear to me, although it's hard to tell, because the curves on their spec sheets are for much higher currents that you'd ever use in a driver for an output tube. Also, the input capacitance is about 10 times or more than would be typical for a driver tube, so your source would have to be capable of driving it.
Of course, you can get around those issues, with cascoding, feedback (degeneration), etc., but I'm not sure what the benefit would be.
Sheldon
This was more of a "could it be done" question. I have not seen an amp that uses only a single high power output tube and no other tubes. I wonder if it would be one of those things where the amp sounds better then what it's measurements would show.
A tube driving a power jfet would seem more obvious but similar hybrid amps already exist.
A tube driving a power jfet would seem more obvious but similar hybrid amps already exist.
Sure, it can be done. But as mentioned previously in the thread, the high voltage FET's that are available are pretty high current devices. And in a tube amp, it doesn't make much sense to have your driver stage eating up much current, certainly not more that the output tube. That doesn't mean that you can't use these devices at lower current, but I think they will be much less linear than available tube choices for the same application. And, they have much higher input capacitances. You can deal with those issues, but what have you gained, as a practical matter?
I'm by no means an expert on any of this, take the following with that in mind. The FET's designed for audio are more linear, but also for the lower voltages typical of SS amps. And there seem to be small signal devices (linear, low voltage) and output devices (high current), with little in between. I would guess that audio FET's could be designed to substitute pretty well for a driver tube but there's no commercial reason to do so. So unless it accidentally fits well with some other application, it won't happen.
Sheldon
I'm by no means an expert on any of this, take the following with that in mind. The FET's designed for audio are more linear, but also for the lower voltages typical of SS amps. And there seem to be small signal devices (linear, low voltage) and output devices (high current), with little in between. I would guess that audio FET's could be designed to substitute pretty well for a driver tube but there's no commercial reason to do so. So unless it accidentally fits well with some other application, it won't happen.
Sheldon
I assume that the JFET is driven as a follower, with the tube for voltage gain? The OP is asking about the JFET as and input/driver.
Sheldon
I just know enough to realize I don't know enough. What happens if you drive an output tube with to much current and how much is to much current? I know tubes are best at voltage gain. Does that mean that it is best to drive a tube with a low current high Voltage signal? What problems or benefits could be gained from driving a tube with a moderate current, high voltage signal?
My thinking is that these new jfets are very low distortion, and very linear. Would a 2a3/300b etc be better driven by them? I'm also curious if they may do a better job of splitting phase when used in a PP amp. From what I understand up until recently we didn't have transistors that could handle the high voltage of a tube amp, now that we do it seems like a cool idea to play with.
Why is a high input capacitance so bad? Don't you have the same issue with SS amps that use jfets?
If we don't try new, untested ideas in life we never grow. I don't have the level of technical knowhow to try this on my own......yet. I am focusing on building my first F5 amp right now but I have been on this site for years reading and learning.
6P15P thats the same as an SV83 right. My Decware zen select uses that tube and I love that amp. Do you have a posting about this amp?
The tube requires some current to drive the capacitance at the grid, or to supply some extra current if you want A2 operation. But the tube takes only the current it needs. It doesn't care how much extra you run through the driver. The problem is, that if the FET requires lots of bias current to be linear, then your power supply has to supply it. You can do that, but you may be putting more power into the driver than the output tube. Not very efficient.
They might be OK at reasonably low currents. I don't know what they look like in the range of 5-25mA. There are plenty of FET's that can handle tube amp voltages.
High input capacitance requires high drive current, or your high frequencies will roll off too soon. Many sources are not designed to do that. If the source is designed for it, then it's not a problem. You don't have the same problem with normal SS amps that use jFETs on the input, as those input devices are small signal devices, with much lower input capacitance than the power FET's we're talking about here.
From an engineering standpoint there are some disadvantages to using these FET's as a direct replacement for an input/driver for a tube amp. What are the compensating advantages? Answer that question to your satisfaction, then go for it.
Sheldon
They are really lousy if used as amplifying devices, high distortion and bad linearity. As these are current devices they should be local feedbacked(Zen-style) and hard loaded.
If you use them as source-followers they are probably great.
No, But I have a thread about similar (simplified) amp, called Tower-III.
Yes, 6P15P is SV83, and it is very nice tube in triode mode.
And I am answering, that vice-verse is more effective. 😉
And yes, he can use JFET to drive 300B. Like, Michael Koster posted here schematic of his amp where he used MOSFET to drive DHT tube.
Attachments
Last edited:
To drive a 300B you'll need about 80~85 Vp (160~170 Vpp) of signal swing on the 300B grid. Of course, the exact value will depend a lot on the quiescent point chosen and a little bit on the tube-to-tube variation. But figure 85 Vp as a worst case.
Most power amps will be driven by a preamp that sources a voltage on the order of 1 Vrms (= 1.4 Vp). Hence, the 300B driver will need to provide a gain of roughly 60 V/V. If the source or preamp is capable of delivering a higher output voltage, this requirement can be relaxed some.
Can the JFET be operated as a grounded source amp and provide this kind of gain at a reasonable performance? If not, then you're probably looking at something like the PowerDrive circuit where the FET is used as a source follower. In this case, you'll need 85 Vp swing on the gate of the fet. Swinging that large voltage across a high input capacitance will require high drive current. Hence, a FET type with low input capacitance is preferred. AFAIR, the 2SK2700 that George uses in the PowerDrive circuit has an input cap of some 500 pF. There are some other types that'll work as well. 2SK3564 for example.
I don't know the numbers for input capacitance of the brand of JFETs you mention. It varies from device to device anyway. But a JFET is basically a reverse biased PN junction, so the capacitance should be manageable. However, the capacitance will also depend quite dramatically on the voltage across the depletion region, so it wouldn't surprise me if the performance limit in terms of distortion at high gate-source (or gate-drain) voltage swings would be set by the voltage coefficient of the FET input cap. But that's just gut feeling, you'd have to simulate it or measure it to find out. Maybe even do the math ... 😱
~Tom
I've done what MK has done too (for a low impedance output on a phono amp). I know that he asked about driving the tube, but the OP made it clear that he is asking about substituting the FET for an input gain stage. By "driving directly" I think he means substituting for a single tube gain/driver stage. Let's not confuse the issue.
Sheldon
Last edited:
I don't see why the issue would be confused.
The OP asks for a "power JFET" driving a 300B output stage.
He does not make clear if the Fet driver should amplify voltage, or mere current.
Both options however can be called "drivers" IMHO.
He was referred to the power drive circuit, which uses an FET to load the input tube, and and FET as a source follower to drive the output tube. Neither FET is used as a voltage amplifier. He made it clear that this was not his intent. But let's ask him. I don't want to get involved in a pointless semantic go round.
DJNUBZ?