SS Followers to drive power tubes

[gingertube]

Guys,
I am using ZVN0545A MOSFET Source followers (direct coupled) to drive output tube grid 1 (with the output tube bias applied to the mosfet gate). I use a "Ring of Two" BJT current source for the source load.
I recently saw an article (which I can no longer find) where it was suggested that a bipolar emitter follower would do a better job than the MOSFET source follower in this duty.

Can anyone confirm, refute, offer a WAG as to whether this is indeed correct?

Thanks,
Ian

[Miles Prower]

Originally Posted by gingertube

Guys,
I am using ZVN0545A MOSFET Source followers (direct coupled) to drive output tube grid 1 (with the output tube bias applied to the mosfet gate). I use a "Ring of Two" BJT current source for the source load.
I recently saw an article (which I can no longer find) where it was suggested that a bipolar emitter follower would do a better job than the MOSFET source follower in this duty.

Can anyone confirm, refute, offer a WAG as to whether this is indeed correct?

Thanks,
Ian

If the driven grids stay in Class *1, it really doesn't make much difference. If the grids are driven positive, then it does make a big difference. For emitter followers, the input impedance is dependent on the load impedance. That's why you often see cascaded emitter followers used when you want a Hi-Z input. What you don't want happening is that varying load on the stage that connects to the emitter followers. This is why FETs are preferred for grid drivers: the input impedance doesn't vary with a steeply varying load, which is what Class *2 grids are.

[Yvesm]

Hi Ian and Miles,

In other words, you must provide a base current equal to the expected grid current divided by the beta of the BJT.
And yes, it will vary !

Yves.

[Jaap]

If I remember well I read somewhere that Allen Wright thinks that tubes sound better than SS around the position of a cathode follower. Would it be a good idea to have the Wright/Rasmussen all tube SLCF before outputtubes in a pp amp ? Or is this restricted to pre-amp territory ? Like to try that some day.

[Allen Wright]

I believe you'd need a very large amount of output tubes to go to the bother of adding SLCFs to drive them.

Our dpa300b amps get by with a cascoded 6H30 running at 15mA very nicely - and even our new 150 watt amp with paralleled push-pull KT88s also uses cascoded 6H30's, although now at 20mA.

KISS is my key principle in signal path design, and I'd never add an extra stage unless essential.

Regarding MOSFETs, my opinion is that they have NO PLACE in an audio signal path, either as source followers or gain elements - NONE!

Regards, Allen

[Zen Mod]

Originally Posted by Allen Wright

.......

Regarding MOSFETs, my opinion is that they have NO PLACE in an audio signal path, either as source followers or gain elements - NONE!

Regards, Allen

you are old enough and toobie enough to have all my respect and love ..... - but say that to Papa ........

I had impression that topology is always more important in your designs than parts ...... so I'm kinda surprised



in any case - your writings regarding Tectronix-SLCF topology urged me (many moons ago) to try any (known to me ) active part in that manner , and for my ears what you wrote is true - dang above that cascode , dang bellow that CCS - and active part is invisible

[Allen Wright]

Zen Mod,
>>I had impression that topology is always more important in your designs than parts ...... so I'm kinda surprised<<

So was I. I thoght theMOSFET'd SLCF was just fine, until I made up the all tube version - and then I heard the difference!

>>in any case - your writings regarding Tectronix-SLCF topology urged me (many moons ago) to try any (known to me ) active part in that manner , and for my ears what you wrote is true - dang above that cascode , dang bellow that CCS - and active part is invisible <<

Sorry, don't quite follow you here - you mean you like the MOSFET in thr SLCF or you don't?

Regards, Allen

[SY]

Curiously, I went the other way- I used to use tubes for driving AB2 output stages, but got much better results when I switched to MOSFETs.

Of course for screen drive, most CFs will curl up and die. It takes a MOSFET to deliver the current punch necessary. I've seen comparative results with tubes appropriate for that service (like 6S45), and the MOSFET still wins.

[Zen Mod]

Originally Posted by Allen Wright

........

Sorry, don't quite follow you here - you mean you like the MOSFET in thr SLCF or you don't?

Regards, Allen

I liked even "all mosfet" SLSourceFollower

[leadbelly]

At the risk of being off topic, what about bandwidth limiting? If you had cheap iron, an SS follower might make things worse with a wider bandpass, no? In this case, wouldn't you be forced to add rolloff at the stage before the follower?