F5 power amplifier

[juma]

Originally posted by jacco vermeulen
Ya mean the Yugo Ferrari ?
http://video.google.com/videoplay?do...83761175801872

Yes, that's Stojadin !

[Skorpio]

Here is a headphone amplifier looking very much like a bipolar F5...

http://users.otenet.gr/~ATHSAM/headphone_2.htm

[tms0425]

Originally posted by Skorpio
Here is a headphone amplifier looking very much like a bipolar F5...

http://users.otenet.gr/~ATHSAM/headphone_2.htm

Except it's full of capacitors

[jackinnj]

I saw this on the simulation, and had to see in real life. The response continues to rise the output was 2.83V into 8 ohms. I let the amp cook at 1.3 Amps, 24V for an hour and it stabilizes quite nicely. The heat sink is cooled with forced air at around 100LFM.

Simulation:



Results at 1W into 8 ohms:
(the AP only goes to 200kHz)

[GRollins]

Easily enough fixed if it bothers you. Just put caps in the feedback loops...or the drive line...or the input resistor network...or...
I wouldn't lose any sleep over it.

Grey

[jackinnj]

Originally posted by GRollins
Easily enough fixed if it bothers you. Just put caps in the feedback loops...or the drive line...or the input resistor network...or...
I wouldn't lose any sleep over it.

Grey

What bothers me is that the results are dissimilar from those illustrated in Chart 12 of the AX article.

The distortion I see is mostly 2nd order as well.

[TimS]

Originally posted by jackinnj


What bothers me is that the results are dissimilar from those illustrated in Chart 12 of the AX article.

The distortion I see is mostly 2nd order as well.

Have you tried trimming R1 or R2?
By trimming the gain between the top and bottom halves of the circuit It is possible to significantly reduce the second harmonic.

Cheers

Tim

[Nelson Pass]

What is the gain of the amp?

What semiconductors did you use?

[jackinnj]

15.4dB

I used 2SJ74 and 2SK170 matched for Idss ~ 7mA. IRFP240 and IRFP9240 that I obtained from Avnet.

FWIW, the power supply will be running in the Belmont Stakes.

[Nelson Pass]

I built one channel with an IRfp240 and the Fairchild P channel
part named in the article, so I'm going to assume that the primary
issue is the IRFP9240.

Somehow I remembered commenting in the article about the
IR version of the IRF9240 with regard to the performance, but going
back, it appears that I probably remember it from this thread.

The IRFP9240 from IR suffers from the high frequency shelving of
the transconductance figure which makes it difficult to use in
Common Source applications. You need to replace it with a 9240
made by Harris or Fairchild (or anybody else). I used the Fairchild
FQA12P20 with good results, the ones you saw in the curve.

I recommend that you get a pair of FQA12P20's from Mouser as
a start.

The IRFP9240 from IR accounts for the second harmonic, which you
will probably note favors the + half of the output with increased
amplitude (2nd harmonic), and you should also see that this effect
occurs from the mid audio band up and is not seen at low frequencies.
This stems from the decreased gain of the P channel device.

The response peak on the top is likely also an artifact, and if it's
still there after replacing the P devices, there's a number of ways
to tweak it out. Alternately if you don't mind the distortion figure
of the IR P channel parts, we can tweak those for flatter response.

Last question, what is your supply voltage?