Using mains transformer as output transformer

[Fuling]

MTP2P50E

Thank you for mentioning these mosfets, I didn´t know 500V P-channel devices existed!
I immediately bought 10pcs for 2 dollars each at ebay.

[revintage]

Erik,
Thanks for the link, will look deeper into this interesting solution, have lots of 6080 lying around.

Rod,
Great advice, ordered 10 samples through OnSemi immediately. As you say BJT is not the obvious choice.

Also got the spicemodel from OnSemi to make some initial sims before testing IRL.

Fuling,
Do you have any direct plans for the PMOS?

[Fuling]

reVintage:

Constant current plateloads! So far I´ve used MJE350 in cascoded bipolar CCSs for input and driver stages up to 300V B+, I also have a bunch of DN2540 that can take 400V but I haven´t used them as they seem very intolerant against ESD (people destroys them all the time).
A 500V 2A 75W P-channel device is something I have wished for for years!
With these devices I´m looking froward to squeeze at least 450Vp-p out of an 12B4A stage with 500V B+ and fixed bias, excellent for my cathode follower projects!

[revintage]

To make things more complicated a N-ch MOSFET could be added on the active side.

M1 makes a highZ CCS and M2 makes a LoZ Voltage source.

Could not find any highvoltage NMOS at OnSemi, any suggestions Rod?

R3 sets Iq, R2 sets balance in the current-mirror and R11 sets anode-voltage.

Must admit that "sand" is not my territory so I leave to someone else to adjust the circuit.

[Rod Coleman]

hello Lars,

I like the 900V FQPF3N90 from Fairchild for all of this kind of work. The extra high voltage helps reliability! It's an isolated case TO-220, so building is easy.

Even so, some precautions using FETs in this kind of circuit:

- use 100nF or more on mains trafo secondary, else turn-OFF transients can wreck your FETs.

- use 330..1KOhm gate resistor as a stopper. Low inductance chip resistor is best (see pic). Power FET can give unwanted oscillation at 1..30MHz or more!

- With current source/mirror and cascodes in the same node as a triode, sometimes stoppers are not enough to stop oscillation. I have had success using Chip Ferrites (Murata BLM21A102S is very effective in this kind of position)

- 6V..10V zener (gate to source) to protect the oxide is vital! See the zener in the photo.

[Rod Coleman]

hello Fuling, I have had some good results using current sources in anode circuits. For current source applications using the P-channel 500V FETs, try using a chip resistor in the source circuit, and the PNP Zetex ZTX560 "cross-coupled" to form the current source.

When I built this circuit and compared to circuits with LEDs, zeners, opamps etc - it sounds MUCH better!
Put 1K in the P-ch gate, and about 6,2K in the PNP base circuit to compensate. Run about 1mA in the PNP, and increase the 6,2K if there is any oscillation.

for low power circuits, another choice for P-channel at 450V durability is the superb Zetex ZVP0545 -

http://www.zetex.com/3.0/3-3-2b.asp?rid=35

the capacitance is even lower than the ON part, so HF power supply rejection is improved still further, and makes a superb source.

[revintage]

Hello Rod,
Do you have a source that can deliver the FQPF3N90 in small quantities?

[Fuling]

Rod: Thanks for your advices! I guess cascoding the mofset thing would improve things overall, right?

[smoking-amp]

"kenpeter: I simmed your (as usual weird circuit )and mating the 6V6 with the 2A3. Think you got something here. Have you actually tested it?"

As Ken mentioned in posts #92 and #94, one can get twice the power of SE using a P-P xfmr and still get SE sound using dissimilar Gm devices. Some previous threads on the subject:

http://www.diyaudio.com/forums/showt...07#post1446107

http://www.diyaudio.com/forums/showt...81#post1597781

http://www.diyaudio.com/forums/showt...005#post638005

No reason to be using air gapped SE sfmrs anymore.

also regarding HV Mosfets:

There are some other HV P channel Mosfets besides the MTP2P50E which is somewhat dated, like FQP1P50:
500V, 63 Watt, Output capacitance only 40 pF. (the MTP2P50E has 100 pF output cap.) also the FQP3P50 85W 70 pF

For N channel there are FQP1N50,FQU1N80, FQP2N90 and STP2NK60, STP2NK90, 2SK3563, 2SK3566


Don

[Michael Koster]

Hi,

I'm looking for a 900V (N-ch) MOSFET in a TO-3P or TO-247.
It should have relatively low charge, ideally the 2N90 or
3N90 but in a TO-3P. The deal is I'd like to actually use it
in a 700V circuit. What I could find is a 7BN90, which will
probably work OK but has a lot of capacitance. I worry
also that it's got a lot of small junctions in parallel and
could current-hog.

I'll try the ones I have but would like to have an alternative
in my back pocket in case these are less than ideal.

Thanks,

Michael