0.8W Class "A" Guitar Amp (MOSFET Output Stage)

Started by DDD, March 22, 2017, 07:45:54 AM

Previous topic - Next topic

DDD

Hi men,
Please take a look at one more guitar PA for bedroom use.
It has about 30 Ohm output impedance like many tubed amps, and very nice overload curves with input signals up to 12Volts peak-to-peak.
The amp is thermally compensated.
The MOSFETs should be mounted on heatsink (about 5.5 Watts for each MOSFET).
More useful info and pictures are at:
http://forum.gtlab.net/cgi-bin/yabb2/YaBB.pl?num=1488919485/60#76


Too old to rock'n'roll, too young to die

bool

Perhaps inserting "another R11" between M1 D and M2 S would add a little more drive to the amp; making it more a push-pull class-A mode. (and perhaps readjusting R10 to cca. 910 Ohm).


DDD

The "existing R11" is not necessary at all.
It's intended for quiescent current measurement (95mVDC at M2 source), and may be omitted for simplicity.
Moreover, since 2SK1058 has source terminal on the metal flange, it's possible to mount M2 just on the amp's metal case (chassis). Hence one of two heatsinks may be omitted.
================
The "another R11" between M1 and M2 changes the distortion curves drastically, but I like the present curves very much.
Too old to rock'n'roll, too young to die

DDD

Below is the PA output voltage at 0.5 - 1 - 2 - 3 - 4 - 5 Volts, 200 Hz input signal.
It's clearly visible that magnitude of the distorted signal increases with the input signal (This is the particular feature of the present amp I've written above).

Too old to rock'n'roll, too young to die

bool

I assume that you tweak the upper (positive) excursion curve with the C5 value?

I had a small srpp-ed - made with BS170s - "amplike" preamp for DI bass recording and also tweaked the driven curves a bit. But ultimately the concept failed with battery power, because when battery started to loose the power, bias points would mess the curves and change the tone ... but it was quite convincing for a single totem / 2 fets circuit.

The only other negative was that it flipped the signal polarity. But hey ...