Fast Class A Amp

[Nico Ras]

I HAVE FOUND THAT A FAST AMP PERFORMS BETTER THAN A SLOW AMP.

THIS LITTLE 30 WATT CLASS A THAT I AM DESIGNING FOR SOMEONE HAS A SLEW RATE OF 580V/uS AND THERE IS NO RINGING OR OVERSHOOT, JUST A NICE SMOOTH-OVER.

THE FREQUENCY SELECTED WAS 250kHZ. THE AMPLIFIER BAND WIDTH IS CURRENTLY LIMITED 0.1 HZ TO 7.4 MHZ.(-1dB)

SOUND IS CRISP AND VERY CLEAR WITH INCREDIBLE DETAIL. THD AT FULL POWER AND 100Khz IS BELOW THE PROVERBIAL 0.0008%. THE THD AT 1 MHZ 0.003%. AND AT 10 HZ 0.0002%. aT 1 KHZ IT IS 0.000012%

TWO TONES 19KHZ AND 20 KHZ GIVES INTERMODULATION PEAKING AT -134 AND -138 DBC, JUST BELOW THE NOISE.

I AM USING FOUR PAIRS OF 2SK1058 AND 2SJ162

I WAS TASKED TO MAKE THE ULTIMATE AMPLIFIER, IT Makes my toenails curl when listening to Diana Kral.

ISTHERE MORE ISSUES THAT I SHOULD CONSIDER. BESIDES GROUP DELAY, WHICH IS JUST ABOUT 0.138 DEGREES OVER THE 0 - 200 KHZ RANGE.
COMMENT WELCOME.

[AKSA]

Nico,

With those measurements, and particularly with that sound as you describe, I doubt more needs to be done.

Are those measurements made empirically, or using PSpice?

Except perhaps that you should have it made in China and sold worldwide for $1k!!

Hugh

PS: Your graph measurements indicate 520V/uS, in fact, and you should specify a particular frequency if you quote group delay in degrees......

[darkfenriz]

Nico

You are either a genius or a clown.

[MJL21193]

Originally posted by Nico Ras
IT Makes my toenails curl when listening to Diana Kral.

Excellent stuff Nico.

[darian]

Good job... It sound quite out of this world performances, even too good, especially with mosfets. You must rely on a heavy negative feedback to get that out of them, I guess. Linearity is not their best usually. I am curious about a schematic. I didn't think it was possible to get such numbers. Seeing is believing...

Regards

[Steve Dunlap]

If those specs are from a real unit no wonder you are SHOUTING about it. What distortion analyzer were you using? Mine doesn't go that high in frequency or that low in distortion measurement.

[CBS240]

Hi


Of course a fast amp is better than a slow amp, but the ability to drive a reactive load with stability is quite important. Mosfets are plenty fast enough, if you use them correctly. Do you use local feedback or error correction for the outputs? Vertical or lateral fets? I would guess verticals..... A schematic would be interesting, but I bet your layout has a lot to do with the performance at such high frequencies. I would be more interested in the layout. The class A operation would contribute a lot to the performance because there is no switching off of the transistors. Approximately what price do you pay (in inefficiency) for the class A operation?

just curious

[lineup]

HAS A SLEW RATE OF 580V/uS
AND THERE IS NO RINGING OR OVERSHOOT

How did you do this?

You have full information about your other amplifiers here:
http://www.digisec.co.za/ras/

But if that amplifier is that good, I would consider selling Diy KITS.
If I were you

Regards, Lineup

[darian]

Yes, I am quite curious, at 1MHz it would be better than a diamond buffer actually, which has no amplification by the way. I just can't even imagine such performances, and even less a way to measure them accurately. Let us know about the design more precisely. Even with a full current mirror topology and diamond buffer output I couldn't obtain that, would it be with feedback. PLEASE... some schematic!!!

Regards

[Jacobsen]

Originally posted by lineup


How did you do this?

You have full information about your other amplifiers here:
http://www.digisec.co.za/ras/

But if that amplifier is that good, I would consider selling Diy KITS.
If I were you

Regards, Lineup

Hello - Prove it and we will make a lot of money - per@navi-star.com