TSSA - The Simplest Symmetrical Amplifier

L.C., i would like to have you sentiment about this idea (May-be it was yet in your head ?) to have two different output stages technology (BJT, FETs) with the same previous stages ?
To make the things more clear, i believe between all SSA combinations, this had not be tested yet:
I presume it would need some little changes of the temp compensation in the previous stage.
 

Attachments

  • ultrasymfet.jpg
    ultrasymfet.jpg
    27.9 KB · Views: 592
Last edited:
L.C., i would like to have you sentiment about this idea (May-be it was yet in your head ?) to have two different output stages technology (BJT, FETs) with the same previous stages ?

I thought about it too but never tried all mosfet triple like BIGBT. This configuration would be nice to settle with exactly 50% of current splitting between the output mosfets. First Rgs could be trimmers and after finding the value to exactly split the current, can be replace by fixed resistors. The only problem would be loosing quite a portion of output amplitude in regarding to double Vgs value from the rails. Clipping can be around 5 V below the rail potential, lower efficiency in compare to regular emitter follower.

Also as you already mentioned BJT and mosfet outputs could be easily swapped to check sound signature of certain configuration or just because experimenting to find best sounding solution. :up:
 
The only problem would be loosing quite a portion of output amplitude in regarding to double Vgs value from the rails. Clipping can be around 5 V below the rail potential, lower efficiency in compare to regular emitter follower.
Right, but no problem to provide separate power rails, isn't it ?
Anyway, for experimenting, it will not be a great difference in listening level, with rails >50V. Just power numbers.

An opportunity for Nico Ras to share with us with minimal work more experiences about the sound signature of power devices, BJT, Hex FETs, MOSFETs and try different combinations of them. ;-)

I will try to find some times and courage in next months to explore that too, it really tease-me... I use to prefer the sound of power Hexfets, but this is far to be "scientific" as so many things differed in my comparison's amps.

Can be simulated too, to see if any obvious difference in virtual measurements and save time to eliminate some combinations...
Lazy Cat said:
Mosfet driver would practically operate at constant Id, its Rds will change just to sustain constant voltage drop on Rgs resistors.
Agree. Not to forget we can play too with gate stoppers and, as you mention, Gate/Source resistances values of the driver.
 
Dear ersa191

In a glorious spirit of enthusiasm you have made a great leap forward but maybe just a little too fast, so now I only can please you to wait for Idefixes to test the prototype for functionality and than you will both cross the finish line simultaneously. :cheers:

P.S. ... to ersa's post which vanished ehhehehe:Popworm:
 
Last edited:
With the driving topology, I think the BIGBT with HEXFET+BJT is a logical contender for competition with all-BJT output. But for HEXFET+LATFET, I prefer BJT driving LATFET because in the original circuit, the HEXFET is not really enjoying much current.

For very high power low distortion SSA LATFET, my plan has been to drive the LATFET with emitter driver (2N5401 or MPSA92), with cascoded VAS or non-cascoded VAS. That is what I think is more feasible to create the best sound at high power.
 
At which frequency?
6/ 10Mhz depending of comps or gate stoppers. Gates stoppers or not. But just a quick run, adding it in a existing schematic. It has not worked out of the virtual box. Murphy was there.
I need to look at this as stand alone to explore it more deeply.
No problems in reality, 100 kHz square waves sharp with no ringing on 4 ohm resistor. ;)
Did-you tested-it in real (I'm exited :)? How do you feel -it?
HEXFET-> LATFETs ?
Is-it faster than without driver ?
 
Last edited:
I thing it was definitively a bad idea.
MOSFET driving complementary BJT can have a x10 FT than the BJT, so it do not add a pole.
We are not in this situation with MOSFET driving MOSFETs.
This added pole divide by 2 the Open loop bandwidth where we wanted the reduced Drain/gate capacitance to increase-it :-(
The added non linear GMs add distortion.
We do not need any current gain at low frequencies, where power MOSFETs are near infinite.
The low transconductance introduce a peak to peak of -10v on each rail with my parts, as L.C said..
I feel stupid, as a simple reasoning was able to bring-me to those conclusions. I was upset by the aesthetic symmetry of the schematic and by the disparity of the N and P gates capacitances of the power units.

Nice looking schematic, poor results.
 
Any ideas from the amp gurus on this thread on how to increase the number of output devices on the TSSA circuit on page 16, post 159 as my load will be 4 ohms or less?
Also I am not sure what typology to use on the output after reading Christophe’s informative posts.
Perhaps I should be using the SSA Low Impedance circuit from the Simple Symetrical Amplifier thread??
http://www.diyaudio.com/forums/solid-state/193923-simple-symetrical-amplifier-118.html
Comments are welcomed as I am not an expert.