Just curious, but has anyone every tried building Class A amps with logic level mosfets? I am assuming that a lower Vgs threshold means potentially lower rail voltage and hence less waste heat for same AC music output...
I may be spouting drivel....?
I may be spouting drivel....?
use bigger , more efficient and more sensitive speakers , so build less potent amp
in Watts , not Cojones
if in sensible range (not chipping much from rails , which is also dependable of amp concept) value of Ugs is much less important than at which voltage mosfet capacitances are becoming so-so linear
in Watts , not Cojones
if in sensible range (not chipping much from rails , which is also dependable of amp concept) value of Ugs is much less important than at which voltage mosfet capacitances are becoming so-so linear
I've wondered if a "dynamic bias" could be used, where if the amp knows what's coming, it can burn more power when it has to (to accommodate musical passage) then go back to the "first watt" level. Probably via some kind of delay on the music...
Unless of course it takes 100W quiescent dissipation to get the 1st watt correct, then my idea is off -
Unless of course it takes 100W quiescent dissipation to get the 1st watt correct, then my idea is off -
@ZM,
Playing Spice on MoFo, so if I remember my `Pass readings' capacitances not so important in common drain application..(?)
Playing Spice on MoFo, so if I remember my `Pass readings' capacitances not so important in common drain application..(?)
No.
Power output depends on rail voltage, period.
Making it easier to drive does not change that.
Pa done it all , ages ago
not so, you're right
but , again - Spice is great and useful, while amp must be also great in vivo ; there is a practical lowest limit for rail and for Iq , when speaking of A Class amp
if going bellow , D class is sounding better while costing peanuts in energy
but for me , that's the same as spinning vinyl with top lid closed
so....
Would anybody like to feel generous and share some reasonable models for IRL640 and other common logic level mosfets?
Cheers
@JM
I agree heat dissipation depends upon bias current and rail voltage, but power of AC sine wave amplified without clipping in choke loaded follower depends upon other things like bias voltage versus rails.
Would anybody like to feel generous and share some reasonable models for IRL640 and other common logic level mosfets?
Cheers
@JM
I agree heat dissipation depends upon bias current and rail voltage, but power of AC sine wave amplified without clipping in choke loaded follower depends upon other things like bias voltage versus rails.
Last edited:
OK
I'm missing something
if you can achieve the current bias you want at a lower voltage, then headroom to a 24v rail is more than starting at a higher number...
No?
I'm missing something
if you can achieve the current bias you want at a lower voltage, then headroom to a 24v rail is more than starting at a higher number...
No?
you're pretty much obsessed with irrelevant thing
use part which you have
choose Iq and rail which suit your needs , but don't go beneath 15V or so , if using IRFP family of parts
if you observe rail and possible output voltage envelope , Rdson is most important factor there ( losses in part itself ) while gate (swing) can go freely above rail , if you're using some external source for modulation , being capable of greater swing
if not , if modulation source (FE, if you like) is internal part of your amp , then feel free to establish whatever set of prerogatives you want ....
that's why you need to study , for instance, F4 schematic - and why there are protection diodes .......
use part which you have
choose Iq and rail which suit your needs , but don't go beneath 15V or so , if using IRFP family of parts
if you observe rail and possible output voltage envelope , Rdson is most important factor there ( losses in part itself ) while gate (swing) can go freely above rail , if you're using some external source for modulation , being capable of greater swing
if not , if modulation source (FE, if you like) is internal part of your amp , then feel free to establish whatever set of prerogatives you want ....
that's why you need to study , for instance, F4 schematic - and why there are protection diodes .......
Last edited:
F4: I believe that the diodes on the gates of the input Jfets are there to prevent excursion beyond the voltage rails that would occur with excessive voltage swing from source and would otherwise drive the F4 into clipping..?
And yes likely irrelevant, but one might argue that many posts on here exhibit obsession to some degree.
So, and apologies for flogging a dead horse..., in the case of a MoFo where o/p is voltage rail limited. My premise is that max AC output swing prior to clipping is achieved when output waveform centered on midpoint between rails. Hence, using a device that operates at the desired current bias when the source is at the mid point of the rails is ideal (assuming all else equal). Is this accurate?
And yes likely irrelevant, but one might argue that many posts on here exhibit obsession to some degree.
So, and apologies for flogging a dead horse..., in the case of a MoFo where o/p is voltage rail limited. My premise is that max AC output swing prior to clipping is achieved when output waveform centered on midpoint between rails. Hence, using a device that operates at the desired current bias when the source is at the mid point of the rails is ideal (assuming all else equal). Is this accurate?
make proper analysis of MoFo, both AC and DC
I've been there , done that zillion of times ....... beating not dead one , but wrong horse
I've been there , done that zillion of times ....... beating not dead one , but wrong horse
Can you/anyone be a little less cryptic?
I haven't been **there** and I'd appreciate a bigger clue...
I haven't been **there** and I'd appreciate a bigger clue...
point is , you're insisting on one tiny detail , without observing bigger picture
in short , your hypothesis is mainly wrong
for better understanding what's place of your detail in bigger picture, use LTSpice on MoFo ( and you said you're using it), make simulation of MoFo , and you'll see overall behavior of circuit, same as ratio of input/output voltages , their DC displacement and ratio of rail vs. output voltage envelope
I believe most of these things are already covered in MoFo thread with textual contributions
........LTSpice - crank input voltage and see how much you can squeeze in and get out before serious clipping ........ due to inductive loading , output swing is much higher than rail ; if that's not in lieu of energy preservation , I don't know what is
in short , your hypothesis is mainly wrong
for better understanding what's place of your detail in bigger picture, use LTSpice on MoFo ( and you said you're using it), make simulation of MoFo , and you'll see overall behavior of circuit, same as ratio of input/output voltages , their DC displacement and ratio of rail vs. output voltage envelope
I believe most of these things are already covered in MoFo thread with textual contributions
........LTSpice - crank input voltage and see how much you can squeeze in and get out before serious clipping ........ due to inductive loading , output swing is much higher than rail ; if that's not in lieu of energy preservation , I don't know what is
edited file , to show what's important and what not
I did use different mosfet , one which I had in my LTSpice
read Note in sim file
I did use different mosfet , one which I had in my LTSpice
read Note in sim file