MOSFET practice amp-help needed...

[sajy_ho]

Hi guys, I'm new here, I was trying to build a FET-based practice amp and I found this:http://www.hawestv.com/amp_projects/amp_solid_tube/art_guitar_amp/2_IRF510adj_Dif_In.GIF
I modeled it with LTSpice with two stages of ROG Fetzer valve and I've got some interesting news:
Input signal of 100 mVhttp://i587.photobucket.com/albums/ss312/sajy_ho/FETInferno100mv-1.jpg?t=1388776143
Input signal of 300 mVhttp://i587.photobucket.com/albums/ss312/sajy_ho/FETInferno300mv-1.jpg?t=1388776150
The distortion characteristics is alot like tubes, but I'm wondering does the output sinal need to be in the same phase as the input signal or not? And what do you say about the circuit? Does have a worth of trying?
Thank you guys and sorry for terrible English :duh

[J M Fahey]

It's a weird amp to say it kindly and doubt very much it's a practical solution.
Build a TDA2030 or 2050 chip amp and add the ROG preamp you like to drive it, or any other.

[sajy_ho]

Thank you man, I've built a TDA chip amp before, but this time I wanted to avoid bipolars and build a complete FET amp. How about the signal's phase? Do I need the output phase to be the same as input's? Is it necessary for a guitar amp?

[J M Fahey]

No, ear can perceive phase difference between 2 signals which reach it at the same time, so if you wire 2 speakers out of phase it will *definitely* hear it (and it sounds horrible) but does not perceive *absolute* phase, so if you wire a speaker to a switch to invert phase and flip it, it will sound the same.
So, it's irrelevant whether the guitar signal changes phase in the middle of its path.

As of that amp, although it's actually built with Fets and Mosfets, it won't behave any closer to a tube amp at all, it still has an SS amp "structure" if that means something to you .

If you want to try that way, search here for the **AMAZING** work done by our friend KMG who generously contributes his designs here.
You will be thunderstruck, no kidding.
Write KMG on the top right search window and enjoy.
Specially listen to the unbelievable MP3s .

[Roly]

Hi @sajy_ho, and Welcome!

Well ... that's an interesting take.  Looks a bit ... primitive to me.

Do I need the output phase to be the same as input's? Is it necessary for a guitar amp?

It doesn't matter.  In solid state amps generally the output is out of phase with the input, at least at the power amp input.  Now, does that cause the speaker cone to move forwards, or backwards?   

The velocity of sound in air is ~1000ft/sec (333m/s), therefore one cycle of 1kHz (=1mS) in the air is about a foot long, Middle-C is about 250Hz so its wavelength is about four feet, and that's a full 360º phase rotation.

Then some acoustic guitar control boxes have an "Invert" button that flips the output phase.

If you sit close to the amp and the guitar gets a lot of wash there may well be "sweet" and "sour" spots, but more distant, say sitting on a stool four to six feet away, and facing away from the amp behind you, I don't think the phasing would make any real difference; it depends on a few things, YMMV.

If you really want to explore this you could swop over the coupling caps from the diff pair collectors.


Both CRO traces show considerable crossover distortion, and that won't sound nice; you need more standing current down through your output FET's.  I don't see any thermal compensation, so they had better be self-compensating as reputed. 

I'm also a bit surprised that there isn't any Negative Feedback at all, DC or AC.  By convention a volt in should give full output, so you seem to have enough spare forward gain to apply NFB to the grounded Gate of the second of the input LTP.

[sajy_ho]

Thanks alot, That's lots of helpfull information.