A new ESP 3A amplifier!

[Duo]

Well everybody, I finally built a p3a... Just one channel so far and I'm adding the second later.

It works very very well and seems to accept all kinds of different transistors quite well... I used A1263/C3180 outputs, TIP41/42 drivers, TIP32 VAS, and 2N2222 for the LTP, bias and LTP current sink.

So, anyway, it's obvious that I didn't use a single transistor specified in the compliment, however, it works well any way.

I have a few questions from other people who've built this amp..

1. What's a good bias current for a smooth class A'ish sound but not too hard on the transistors? I see that 100mA is the accepted value but I feel like I want to run it at 300 or so, does that sound okay?

2. It seems to amplify a powerful radio carrier in town very well and this causes lots of heat dissipation.. Are there any good ways to get the amp so it won't amplify past perhaps 35khz or so? I have a zobel network on it, good miller caps and everything so I doubt there's anything wrong with my construction.

Anyway, the bias is no big deal, I don't care if it's not the exact value it's supposed to be as long as both channels are the same and the sound is smooth. As for the radio amplification, that drives me nuts so I really want to get that out of the way..


Thanks.

[paulb]

I think you can run the bias as high as your heatsinks and power supply can handle.
As for your RFI problem: how do you know your heat dissipation is caused by the pickup of radio stations? Are you seeing something on the output? You'll need to find out where the RF is getting in. Capacitor C2 should get rid of most of the stuff on the input.
I've had some oscillation problems with the P3A, but it only seemed to happen when I was running from my lab supply; adding more power supply capacitance made it go away. I could see a steady 250 kHz signal on the output. Running off its regular supply (which has 61000 uF per rail nearby), I've never had the problem.

[Duo]

Okay, yeah, the way I can tell it's amplifying radio signals is because I can hear high frequency (about 18-20Khz) on the speaker that changes with the input impedance and also, when this occurs, the transistors get hot because of overlapping current since they don't switch off fast enough to handle the signal...

[mlloyd1]

As I remember, this design has no thermal feedback between the bias regulator and either the output devices or outputs drivers. So, I'd be leary of cranking up the bias unless I was anxious to see some molten silicon.

Send Rod some email with your questions. He probably gets this question a lot.

Originally posted by paulb
I think you can run the bias as high as your heatsinks and power supply can handle.

[paulb]

Originally posted by mlloyd1
Send Rod some email with your questions.

Can't argue with that.
Let us know what he says.

[Duo]

Will do, I have emailed him before and he seems to reply pretty fast even though he's so busy.

[carlmart]

Originally posted by mlloyd1
As I remember, this design has no thermal feedback between the bias regulator and either the output devices or outputs drivers. So, I'd be leary of cranking up the bias unless I was anxious to see some molten silicon.

If you're going to increase bias, you can put the bias transistor over one output device to track heat.


Carlos

[Duo]

I've been thinking of doing that. Perhaps I'll simply put it right on the power heatsink like usual and make sure that the cooling is good so that my bias doesn't move around too much...