Bias adjustment for power amp? Help please

[cdl]

Hi,

My power amp (Schematic, SMT module) has a trimpot near the driver transistor, just left of the centre of the schematic.

I have been thinking of it for some time, and (sorry to be a moron) I can't really figure out if it is for adjustment of the bias current, or for DC offset (or?).

I would be very grateful if anyone could help...please?

Thanks in advance,

cdl

 

[subwo1]

I'd say it is the voltage adjustment for the amplified diode. That would make it the bias adjustment. Do you know at what current the output mosfets idle at?

 

[cdl]

Hi subwo1,

All I have is the amp and the schematics, so I have no idea what current the mosfets are supposed to idle at

Can it be measured on the left channel (which has not been fiddled with) somehow?

/cdl

 

[subwo1]

Hi cdl,

Lets try the easy way first. Under link1 there is a 1R resistor. If that has a power rating of several watts, you can pull out the fuse, link1, and mesure the voltage across the resistor. Since it is 1R, 1 volt will indicate 1 amp of current. Measure the value for the good channel, and carefully adjust the other to match it. Move it very slightly at a time or the 1R resistor could blow. Be sure there is no audio going to the amp at the time.

 

[cdl]

Hi Subwo1,

Sounds like very sound advice!

The 1r resistor isn't more than ½ watt or so, but should it blow ... well 1r resitors are expendable ... so I'll try it!

Thanks a lot,

cdl

 

[subwo1]

Thank you too, cdl,

for mentioning it. With no audio, the current drawn through the resistor should not be more than the .5 amps it would take to blow the resistor even if it is a 1/4 watt type.

Take care.

 

[cdl]

more questions...

I haven't gotten around to actually adjusting yet - disassembly and reassembly of the amp is an hours work in itself (!) so I guess I'll opt for the weekend.

I guess the adjustment should be done without a load (speaker) connected, although it shouldn't really make a difference (no DC)?

I also considered that actually a bias current as high as possible (without blowing anything) would be desirable? So as to push class A to class B transition into as high a volume as possible?

The downside would of course be a decrease in total power, but who cares - I live in a flat with neighbors etc. and can't play disco levels anyway.

Cooling & power supply should be OK - the heatsink makes up the entire top of the case. The transformer is rather large, with a 5 amp mains fuse, and 6.5 amps fuses for on the 60V supplies.

So what remains to consider is the mosfet's power capabilities, and how much the paralleled 1R resistors in the output can handle, but hey, there's 16 resitors for each halfwave... and I believe the mosfets handle about 20 to 40 amps each.

Am I right?

Ta - cdl

 

[subwo1]

That is a good idea to me. I think conventionally designed amps with mosfet outputs should be idled high because the gate charge keeps the mosfets on longer after crossover at high frequencies, so as I see it, may as well make the amp more consistent and try to keep them on longer all the time. Your amp schematic put it in a conventional output stage category.

It is a good idea to leave the speakers off during the procedure. That way if something does happen to go amiss, one less thing damaged.

If you are not concerned about heat, that is a plus. Mosfets have heftier structures and can tolerate more heat than bipolar transistors. It is just a good idea to try to keep the heat from being trapped in the case so as not to affect more delicate components.

 

[iso_510M]

This may help.

20N10L Datasheet

Looks like heat should be a concern. P-tot is 105W but derate .7W/*C


That is a familiar looking design. What amp is it?

 

[BobM]

Why not call the manufacturer, get an engineer on the phone and ask what the bias should be for that model? Going "hotter" is not always better. Sounds like a better idea than blowing resistors to see how far you can push it.

Enjoy,
Bob

 

[cdl]

Hi,

This may help.

20N10L Datasheet

Looks like heat should be a concern. P-tot is 105W but derate .7W/*C

Hmmm, you might be right. Actually, have been forced to exchange most of the 20N10L's with a newer HexFet type (20N10L is no longer available anywhere) - and I think the newer type might be better in that regard. But there's still the other channel...

That is a familiar looking design. What amp is it?

It is the Aura PA-200.

Why not call the manufacturer, get an engineer on the phone and ask what the bias should be for that model? Going "hotter" is not always better. Sounds like a better idea than blowing resistors to see how far you can push it.

Enjoy,
Bob

That's the fuss. Aura was produced by B&W, and they discontinued it completely some abount 5 years ago... and now will know nothing of it At least that is the answer I was given at a previous occasion when I asked for some detail about Aura.

But it's a good idea, so I have submitted a question to B&W ... now let's see if they reply.

Thanks a lot - cdl

 

[cdl]

Another question...

By the ways, while you are looking at the schem...

Isn't the INA103 in the input doing absolutely nothing for me, as I am driving the amp from a single-ended preamp anyway?

As far as I can see, it just amplifies the signal with a gain between 5 or 10 - which I might just as well do in the gain stage of my DIY preamp (active preamp is needed for this one anyway, input impedance of only 10K). And feed the single ended input directly to the SMT amp module ... that would remove three "useless" opamps from the signal path?

Any opinions?

Cheers, cdl