Building a balanced mosfet amp with the IRF9140/140 and at a stage of "tweaking" the sonics to my preferences and the balance is now a little bit soft the first 10 minutes, very good for an hour or so but after say 3-4 hours they start sounding a bit thin and maybe a tad grainy on some material.
I have only listened to and worked on BJT amps for the last few years and they have typically gotten smoother and more refined the longer they stayed on.
Did some THD measurements today as I got frustrated not being able to sustain my preferred sonics over a longer time and it showed a THD increase of between 5-10dB when warmed up. I realize it's running more in class A until it's settled heatwise which may explain the increased distortion.
Either way I wanted to check if anyone have similar experiences with mosfet amps?
I have only listened to and worked on BJT amps for the last few years and they have typically gotten smoother and more refined the longer they stayed on.
Did some THD measurements today as I got frustrated not being able to sustain my preferred sonics over a longer time and it showed a THD increase of between 5-10dB when warmed up. I realize it's running more in class A until it's settled heatwise which may explain the increased distortion.
Either way I wanted to check if anyone have similar experiences with mosfet amps?
The only amp which changes sound a lot is M2, due to large interstage coupling cap charging. According to NP it goes from classB to classAB and finally to classA over a course of minutes. The change is drastic and ideally one would have timer on input or standby mode of some sort. But knowing NP, why? One can wait.
Many other mosfet amps i have work almost immediately and change in sound is negligable.
It seems to me you do not like low distortion. To everyone his own preference. Hard to understand without actual circuit. Balanced circuit tend to cancel our some harmonics. It may be a good thing, may be not. Used to run Son of Zen balanced classA mosfet amp with Lowthers long time ago.
Ok, thx. Not sure why you would assume I don't like low distortion?
If this is not a common trait of mosfet amps I will focus on other areas to see if I can get it fixed.
If this is not a common trait of mosfet amps I will focus on other areas to see if I can get it fixed.
If after warming up.
Might just mean thermal tracking is over compensating.
Been known to happen with BJT transistor used in VBE.
Getting warmer it lowers bias too much.
numerous ways to change it with different thermal track circuit
or even reduce thermal transfer to the BJT
Might just mean thermal tracking is over compensating.
Been known to happen with BJT transistor used in VBE.
Getting warmer it lowers bias too much.
numerous ways to change it with different thermal track circuit
or even reduce thermal transfer to the BJT
It's after warming up for 3-4 hours but the bias is stable at 200'ish mA per bank after 15-20 minutes and 6 hours later.
Thing is I remember having had similar issues in the past which I thought maybe was my first mosfet amp but I just realized it was my 2009 powerful Onkyo HT BJT receiver that run very warm and after 5-7 years it behaved in a similar manner.
At the time I assumed it was the caps getting old because of the heat. The ones in this amp are new but maybe not intended to heat up and stay on for 16 hours a day (Nichicon KG(M) and KW(M)).
200mA is the recommended bias for the amp but maybe not necessary, I will lower it to 150mA, do some THD tests and if ok let the amp run a bit cooler.
Thing is I remember having had similar issues in the past which I thought maybe was my first mosfet amp but I just realized it was my 2009 powerful Onkyo HT BJT receiver that run very warm and after 5-7 years it behaved in a similar manner.
At the time I assumed it was the caps getting old because of the heat. The ones in this amp are new but maybe not intended to heat up and stay on for 16 hours a day (Nichicon KG(M) and KW(M)).
200mA is the recommended bias for the amp but maybe not necessary, I will lower it to 150mA, do some THD tests and if ok let the amp run a bit cooler.
It's supposedly a well designed, fully balanced, global feedback amp.
Just to get the distortion issue into perspective, after the thd increase it was still under -100dB with 3.2 ohm speakers connected which is not really a concern even though it annoys me.
It got me on the track thinking as the mosfets RDS increases with temperature maybe that meant that a mosfet AMP's generally would lose some performance as they heat up.
Unfortunately, you are giving very little useful information. A schematics is needed. First, try to measure if the idle current does not increase with time. Rs or Re of output stage, measure voltage drop.
If the idle current is really stable, then you should check for oscillations with a scope (not with a sound card).
The idle is really stable and I have looked for oscillations but can only measure up to 500 kHz with a simple scope which was clean.
For now I am good knowing this issue is not inherent to mosfet power amps.
For now I am good knowing this issue is not inherent to mosfet power amps.
No, I have already done the "in-amp" hw tweaking as far as the design allows me with different capacitors, placements, ground routing, power supply alternatives/filtering, cables and what not.
I am well aware and as I already shared this is a proven and well respected design incl the pcb. It will not help getting into the design details of the amp for this issue.