diyAudio

diyAudio (http://www.diyaudio.com/forums/)
-   Tubes / Valves (http://www.diyaudio.com/forums/tubes-valves/)
-   -   John Swenson's New BDT preamp (http://www.diyaudio.com/forums/tubes-valves/123402-john-swensons-new-bdt-preamp.html)

jkeny 20th May 2008 10:19 AM

John Swenson's New BDT preamp
 
2 Attachment(s)
Have people seen John Swenson's new take on his BDT preamplifier http://www.audioasylum.com/cgi/vt.mp...bediy&m=141747

Apart from the unique use of Beam Deflection Tubes, there's an interesting twist on Gary Pimm's self bias CCS to make it into a Voltage regulator (par excellance, according to JS). This involves the addition of an error amplifier - full explanation is given on the thread referenced.

I'm wondering how this might be scaled up to higher currents for use in a tube output stage?

Is it a matter of using the Pentode configuration as per Gary Pimms explanation of higher current CCS & if so what other changes might be involved?

jkeny 20th May 2008 10:22 AM

2 Attachment(s)
Unfortunately, the image, when compressed, is not very clear - here's the voltage regulator part uncompressed

jkeny 20th May 2008 10:24 AM

2 Attachment(s)
And the BDT preamp part uncompressed

Edit: Oops edited the first image - it just needed the pixel count reduced to be accepted for posting!

Sheldon 20th May 2008 06:43 PM

Re: John Swenson's New BDT preamp
 
Quote:

Originally posted by jkeny
I'm wondering how this might be scaled up to higher currents for use in a tube output stage?
Those output FET's are rated for 30W. Assuming you are regulating at a generous 50V below your supply input, Q2 would drop 90% of that - let's make it 45V. At 100mA output, you'd be dissipating 4.5W. With good heatsinking, you should be able to at least double that. You could probably parallel a couple of Q2, if you want a big safety margin.

Sheldon

jkeny 20th May 2008 07:43 PM

Thanks Sheldon,
It seems it would be fine then in the current config - with heatsink on output Mosfets.?

I've been reading up on GP's Pentode version for higher currents in which he uses the pentode as an intelligent resistor to dissipate heat - plays no other role apparently! Can leave off the heatsink on Mosfet in that case.

If I was looking for a low voltage high current version say 40V/6A - I wonder would this circuit do the job?

Sheldon 20th May 2008 08:47 PM

Both LND150 and the IRF820 are 500V parts. If your input supply is never over about 450 with no load, you can regulate to 0V and not exceed part ratings. Input can be higher, but should not exceed output by more that 450V. A bigger margin will be better, say 400V in either case. If your input is higher than about 400V, go through the thought experiment to make sure that the voltage across any of the FET's will not exceed that during start up. Charging time of C2 should not be an issue if I've done the analysis correctly, but double check that.

Yes, I see no reason why you couldn't use this circuit for lower voltage higher current applications. In that case you'd need bigger FET's for the output. Input capacitance for the output FET will be an issue, especially since you'd probably want minimal voltage drop (look at the CRSS curves, vs ds voltage). So in that case you might substitute DN2540 or 10M45's for the LND150 and run about 10 or 20 mA through that string. Factor your resistors and capacitors accordingly. It looks like this circuit runs about 0.4mA, so for 16mA you'd need to decrease R1,2,3 by 40X and increase caps C1,2 by the same.

If you are thinking of doing it for the Zeus, it's overkill. PSRR is very good for that design. But hey, overkill is just enough here.

Sheldon

Edit: BTW, if you're planning to build the big Zeus, I've got a set of matched FET's, I'm probably not going to get to.



jkeny 20th May 2008 11:30 PM

Ah, Sheldon, you remember me! The pack rat who got those lovely Lovoltech LU10140D from you but haven't done anything with them yet.

One of the issues holding me back was a preamp to drive the Zeus i.e. balanced & good drive.

But now this new BDT preamp comes at the right time & I'm looking to build it as the front end to a Zeus & a Baby Huey. I'll be building the "BIG" Zeus with FQA34N20 (these are replacement parts for the STW34NB20) Mosfets from which I matched two good pairs.

I believe this BDT will replace the Zeus input trafo & provide Vol control, gain & phase splitting making for a nice integrated amp.

I wasn't thinking about the low voltage regulator for the Zeus but for a Panasonic SAXR57 digital amp that uses Power Supply Voltage Control & needs a very good PS

I'll have to think about all you said as this doesn't come naturally to me!

Thanks again for your help - I'm sure I'll be leaning on you again if that's OK?

jkeny 21st May 2008 12:18 AM

Oh meant to ask - how do you like the Zeus? Which version do you like the best?

Sheldon 21st May 2008 12:54 AM

It's a nice little amp. I've only tried the JFET version. I played with a tube differential input stage, but it didn't offer any real advantage, other than higher input impedance compared to the transformer input. My source has an output impedance of 75R, so it's not really an issue, and I'm using it with the transformer now.

Sheldon

Gary P 21st May 2008 02:46 AM

I have not had time to try John's regulator version of the CCS circuit yet. It looks like a very nice circuit. It should scale up nicely though.

Back when experimenting with filament CCS ideas one self bias CCS got loaded with FQPF50NO6L (60V 36A) MOSFETs with the cascode bias voltage dropped to just a few volts. Just to see what the circuit would take I kept increasing the current to see what would break first. Had to stop at 8 amps when the runs on the board started smoking! No part failures...

I gave up on the low voltage high current endeavors because they almost required running the MOSFETs at low voltage where the capacitance is high.

From experience gained from using the CCS circuit as an amplifier, the voltage across the lower MOSFET needs to be kept at 20 volts or higher for stability and performance issues. 20 volts is fine in circuits running up to the low hundreds of milliamps but not easy to do when talking about amps of current.


All times are GMT. The time now is 12:43 PM.


vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2014 DragonByte Technologies Ltd.
Copyright 1999-2014 diyAudio


Content Relevant URLs by vBSEO 3.3.2