Ping SY

[JoshK]

Stuart,

Did you ever work out your problem with your scanner? I'd be interested to see your schema.


FWIW, I grabbed a few pairs of 6LW6s off ebay reasonably by being patient.

 

[jon_010101]

Now that you have them all, have you seen what they fetch on Ebay. They are GE's by the way. I plan on doing some more screen drive experiments in the future, but I have been infatuated by cathode followers lately. The 6FW5's are for the screen drive experiments.

I bought a sleeve of GE 6FW5s for ~$25 on eBay not too long ago. I tried them instead of 6AV5s in my Bogen DB230A, and the extra transconductance is helpful for feedback -- biased almost the same, too. I had been hoping to find more for projects and such, but it seems JoshK bought them all

 

[SY]

Josh, I'm putting together a website, sloooooowly. I'm still trying to figure out the mechanics... In any event, here's a version of it. The one in my "big" amp used the same mediocre CCS (I'd redo that!) though set at 12mA, and split rails at +/-350V to get the swing higher. 150k on the plates of the 6SL7. There's a follower (MOSFET, though my original used 6SN7) between the driver stage and the output tubes.

 

[JoshK]

Thanks Stuart. I see what you were talking about.

If you wanted to use 5842 (for the big spenders) that would help alleviate the Cgp problem wouldn't it (for the input tubes). Or would the gain not be high enough?

How much gNFB did you have to use to get a decent damping factor? I guess a related question would be how does one figure out the output Z of a screen driven output stage?

 

[SY]

Because of the configuration of the input stage Cgp isn't very important, so you can use a high mu tube. If cost were no object, some matched D3a, triode connected, would be pretty cool.

I'm redoing the stage now for publication at the new website (that's why I had found that old drawing). Besides showing the bipolar supplies and using a better CCS, there's an improvement when the cathode resistors in the first stage are changed to LEDs.

To do a screen drive calculation, you can use the curves given at Vgk = 0, with screen voltage as the variable; several sweep tubes' datasheets show this. Just do the same sort of construction you'd do with pentodes in AB push-pull. Damping factor will never be anything to write home about; there's not a lot of open-loop gain to play with, so 10dB feedback is pretty much what I was stuck with. Fortunately, the MOSFET drive for the screens means that there's no blocking at overload.

 

[TonyTecson]

some matched D3a, triode connected, would be pretty cool.

hey SY, how about the 6BN11, dual pentode compactron tube, care to look at it please?

 

[JoshK]

To do a screen drive calculation, you can use the curves given at Vgk = 0, with screen voltage as the variable; several sweep tubes' datasheets show this. Just do the same sort of construction you'd do with pentodes in AB push-pull.

Like the second graph (dotted lines) in this data sheet?






6DQ6A

 

[SY]

The second graph, yes. Use the solid lines for the calculation of power, load, etc. The dotten lines will tell you how stout your driver needs to be- as a practical matter, an IRF820 source follower or similar will laugh at the requirements.

 

[JoshK]

Thanks Stuart.

I have the 6FW5 data sheet and the curves look identical to the 6DQ6A ones. At least in the GE data sheet.

I was thinking the dotted lines looked funny. I mixed which were associated with which axis.

I was thinking of attempting plate to plate feedback around the output stage and driver, using the 6J52P's as drivers (source follower driving screens too obviously). The 6J52P looks like a D3A and I'd operate it like a pentode. I haven't worked out the op points and gain yet to see if it will work out or not.

I don't know if it'd be better to stick with two triode stages and gNFB or one pentode stage and p2pFB. I like the idea of a shorter feedback loop.

I am probably in a bit over my head here, but this is suppose to make me learn faster.

 

[JoshK]

P.S. what is a good source for explaining Class AB pentode PP load lines? Morgan Jones discusses Class A triode strapped push pull only.

I've read bit here and there but nothing very clear. Is the Radiotron handbook the right place to look?

 

[SY]

Actually, I learned about it from the RCA Tube Manual, though RDH will certainly have a good explanation. As a rule of thumb, start with the B+ voltage, then draw a line up to the knee of where you think your driver will swing. I tend to load low, idle low, and bang the screen hard. My 6LF6 amp had 800V of B+, 3mA of idle current per tube, and 1250 ohms plate to plate. That gave me an easy 150-180W.

 

[SpreadSpectrum]

SY,

Apologies if you have answered this before, but why do screen driven amps avoid crossover distortion?

 

[SY]

The low current linearity of this connection is surprisingly good. I think I've posted some curves that Bascom King published in Audio showing the 6JN6 at very low currents.

Now one does have to be somewhat careful about the output transformer to avoid notch distortion, but a decently stacked and wound primary will usually avoid that issue.