F4 power amplifier

[GRollins]

Originally posted by Vix


Someone on diyaudio once said "Stop pushing my Woo-woo button!"

Push mine, instead...
I have a nearly infinite capacity for 'woo.'

This sort of circuit would work well for headphones, but you could use something like IRF610/IRF9610s for the outputs and lower the bias. Smaller heatsinks, smaller power supply, as well.
Bypassing the input buffer stage will work, but it's hardly going to be a groundbreaking circuit, as it will be pretty nearly the Zen follower circuit. At least this way Nelson is able to publish specs for the circuit. If it did not have the buffer, the performance would be highly variable, depending on the Zout, current capabilities, etc. of the source. Being DIY, we don't really have to worry about such things. We have more control over those variables than the average consumer.
Patrick,
If I remember correctly, the 2SK170/2SJ74 do not have the same voltage ratings, which always struck me as odd, given that they're marketed as complements. At any rate, whichever is lower will be the weak link in the chain. Everything else will be tethered to that one part. I'd love to find 100V JFETs...or even 75V, for that matter...
Of course, I'd like to find a large sack of unmarked $100 bills, too, but that's not going to happen.

Grey

[Jozef]

Voltage is almost same as this linestage.
On one PCB maybe ?

[mpmarino]

the 2SK170/2SJ74 do not have the same voltage ratings

I noticed that and wondered how they could be complements?



It didn't give me a warm and fuzzy feeling

[GRollins]

While the rails are similar, I'm not a big fan of running tubes at reduced voltages.

Grey

[jupiterjune]

If it did not have the buffer, the performance would be highly variable, depending on the Zout, current capabilities, etc. of the source

And what might that look like -- perhaps a phase lag in one channel?

I like the unity gain amp concept--at low listening levels, the output from my preamp has to be attenuated to ridiculously low levels before being fed to my amp.....
This amp lets you bypasses that whole process.

Good ol' P5 in the BZLS might actually work as a good volume control.

JJ

[GRollins]

The Gate capacitance of the output MOSFETs is the fly in the ointment. It's big, it's unruly, and it's nonlinear. Quite an ugly beast. By isolating that capacitance from the outside world, the performance of the amp becomes much more predictable. Yes, the input JFETs have Gate capacitance, but an order of magnitude (actually probably more like two orders of magnitude) less. Quite tame by comparison.
If you leave off the JFET buffer, be sure to go for as low a source impedance as you can get and provide as much current as possible.
Basically at that point you're building a whole 'nother circuit, as the output stage is an entirely conventional compementary follower--pretty much an industry standard sort of thing. It will become whatever you use for a front end; an A-75, for instance.
Yes, I'm overlooking the bias dingus. On purpose. You could build this amp with a conventional Vgs multiplier, or a Vbe multiplier, or use diodes, or a CCS and variable resistor, or...
Really, it's the JFET follower front end that's different.

Grey

[flg]

Yea You mean that bootstrapped TL refrence bissiness??? And he gets almost all that supply voltage to the output too
I would go simulate it but, I don't want you to shame me for doing something so disrespectfull to N.P.s work

[GRollins]

Simulate?!?
Gasp!
Lightnin's gonna strike you, boy!

Grey

[flg]

There is something a little confusing here though Grey The old Class A peak output current rule... That you can get double the queicent current out. Nelson gets more than that???
Dosen't this thing idle at 1.5A or so??? Gets 20V peak to the load??? O.K. thats good for 25Watts out...
But when you drive a 4 ohm load... If you can drive 20Volts out, thats 50 watts like he says but thats 5A pk??? Well beyond the double the idle current addage of old. Does this thing do Class AB for that power level???
Did I miss something???

[Babowana]

Like a mad dog, RRRRRRRed to Reno Hi-Fi, but no F4 there.
RRRRrrrrRRR . . . torn off the drawer and squeezzed into it.
RRrrrrRRRrrr . . . found all from TL431 to the blue LED piece by piece.
RRrrRRRRRR . . . but, missing matched 240/9240s and jfets.
RRrrrrrrrrrrrr . . .

Do you know who has them (matched mosfets and jfets)?
Will give him one spicy bite.

Thanks for info.