Musical Fidelity P170 Bias

[13oots2]

Was experiencing some problems with soundstage so checked the bias on each channel, the readings were widely differing. The problem I now have is I have no idea what the bias settings should be, the amp is supposed to be heavily biased into Class A. I have read quiescent current could be anywhere from 75mA up to 200mA, depending on various forum threads.

I have settled on a value of 150mA which seems to give a very good sound, 75-100mA sounded a little bright and insipid. Obviously I don't want to blow the output transistors and the heatsink at 150mA is very warm to the touch, I can hold my fingers on it for about 4 seconds after about an hour at low volume levels. Anyone have any idea as to what the quiescent voltage should be for this amp?

[jez]

Anywhere from 75mA to 200mA is correct with 100mA about right. There is no specific correct value with a mosfet amp such as this and, as you have found, heatsinking is the main issue!

[Edmond Stuart]

The CPU of my PC is water cooled, so why not do the same with power MOSFETs?

Cheers,
E.

[AndrewT]

Try plumbing it into the return circuit of your central heating.

[13oots2]

Originally Posted by AndrewT

Try plumbing it into the return circuit of your central heating.

Have dismantled the fridge and have icicles forming

Finally settled on about 124mA, nice and toasty but not silly. Once amp is warmed up am within 0.1mV across the 0.22Ohm output resistors, not too bad at all considering they are one turn pots on the bias.

DC offset is 0V on one channel and 0.1mV on the other, seems all my servicing has turned out well. The upgraded OPA 627 are an absolute joy and well worth upgrading from the LM 318's.

[13oots2]

Originally Posted by jez

Anywhere from 75mA to 200mA is correct with 100mA about right. There is no specific correct value with a mosfet amp such as this and, as you have found, heatsinking is the main issue!

Have finally settled on just short of 180mV and it sounds smooth, detailed and powerful. A few of the words said in jest set me thinking and now have 120mm PC fan blowing on heatsink, it doesn't even get warm to the touch now. I now have to build up an enclosure for the fan, got to keep the furry ones paws away from the blades.

The amp is now a different sounding beast and have never heard it sound so musical before. I would much preferred the bias to sound better at lower values but the higher I go the better it sounds

[astankov]

Originally Posted by 13oots2

Have finally settled on just short of 180mV and it sounds smooth, detailed and powerful. A few of the words said in jest set me thinking and now have 120mm PC fan blowing on heatsink, it doesn't even get warm to the touch now. I now have to build up an enclosure for the fan, got to keep the furry ones paws away from the blades.

The amp is now a different sounding beast and have never heard it sound so musical before. I would much preferred the bias to sound better at lower values but the higher I go the better it sounds

And if you get bias to 1A will sound even better believe me!
You will rediscover class A and will have a heater in the cold winter days (or nights maybe)

[13oots2]

Originally Posted by astankov

And if you get bias to 1A will sound even better believe me!
You will rediscover class A and will have a heater in the cold winter days (or nights maybe)

It is my understanding that the P170 was full class A for the first 18W. I never use that living in an apartment with people next to and below me.

[AndrewT]

1A of (total) output bias current gives 2Apk of maximum ClassA output current from a Push-Pull amplifier.
2Apk into 8r0 is 2*2*8/2 = 16W of ClassA output power.
Above that, the 1A bias amplifier will transition automatically over to ClassAB.
Whereas 450mA of bias, gives 0.9^2*8/2 = 3W of ClassA before transitioning over.

[13oots2]

Originally Posted by AndrewT

1A of (total) output bias current gives 2Apk of maximum ClassA output current from a Push-Pull amplifier.
2Apk into 8r0 is 2*2*8/2 = 16W of ClassA output power.
Above that, the 1A bias amplifier will transition automatically over to ClassAB.
Whereas 450mA of bias, gives 0.9^2*8/2 = 3W of ClassA before transitioning over.

Ok! I confess this is all new to me and I haven't touched on these subjects since taking my physics O level 30 years ago.

If I am doing all the math correctly I have a current of 0.8mA (0.18mVx0.22Ohm) per transistor per channel. So that gives me 0.8x2x2x8/2=12.8W, more than enough for my needs. Please correct me if I have this wrong.