F5 power amplifier

Jump to Latest
#9,257
This may have been answered in some page, but here I go; I have seen that rising up the bias to 2A per device is a quite common practice for this design if you want more power but...
What if I down it from 1.3A to, let's say, 1A? Is quality affected or does it decompensate the circuit in any way?

My heatsinks are running a bit hot and I don't use to listen to more than 1W power, so there's a lot of "unused bias"

Thank you guys, regards
Regi
 
#9,258
jackinnj said:
I can probably fit this on a DIP8:

An externally hosted image should be here but it was not working when we last tested it.


Of course, when you parallel JFETs you run the risk of oscillation, and each SOT23 device has to be matched up. With 4 parallel J270's I get about 20mS.
Click to expand...

jackinnj: I'm certainly interested, although I won't be building anything anytime soon; I think it looks a really cool idea! I did some BF862 buffers (and juma's BF862 preamp) on DIP8s and DIP16s, and thought it was great fun, and certainly successful. (Listening to one of them now.) I wondered a year or so ago about trying N- and P- complementary circuits (like Hiraga's Le Monstre) using SMD, but I couldn't find complementary pairs, and I don't have the understanding or experience to figure out something like you've posted. So do you mind a bunch of daft questions? (Some of them a little OT, maybe... sorry...)

1. "Matched up" here means for Idss? And I presume you mean a matched quad. right?

2. Why four? Again, is Idss the issue, or capacitance or something?

3. Apart from the challenge, is there any reason to squeeze five fets on a DIP8? Why not just use a DIP16?

4. How "complementary" would you expect such an arrangement to be, compared to a matched 2SK170/2SJ74 pair? And would this be closest to BL-grade? How would you go about trying to "fabricate" a substitute for Y- or GR-grade?

Well, I'll be fascinated to hear more!

Cheers

Nigel
 
#9,259
The transconductance of a single J175 or J270 doesn't come close to matching the 2SJ74 -- lower distortion goes hand in hand with higher gm. In the case of the F5, the gain of the first section is 20dB, so it's a bit of a "dog food" problem getting the right gm, and bias for the IRFP240;s.

I've done it on a breadboard and there were no issues with oscillation. The J175 is actually MMBFJ175 -- SOT23.
 
#9,260
regiregi22 said:
This may have been answered in some page, but here I go; I have seen that rising up the bias to 2A per device is a quite common practice for this design if you want more power but...
What if I down it from 1.3A to, let's say, 1A? Is quality affected or does it decompensate the circuit in any way?

My heatsinks are running a bit hot and I don't use to listen to more than 1W power, so there's a lot of "unused bias"

Thank you guys, regards
Regi
Click to expand...

Why don't you just try it and let your ears be the judge.
 
You must log in or register to reply here.
Top Bottom