OPMOSFET Amp

Tought I'd share the schematics for an amp I designed about a year ago. Nothing too fancy, but quite easy to build. This is a quiet design and sounds good to my ears. I use it in my studio for my reference amp and sounds as good as i've heard in other ppls studio. Anyway, it doesn't use any Global Negative Feedback, but it does use Op amps (Makes for easy building) I chose the output Mosfets because they were designed for audio and I got into this Hobby to build a DJ sound system, so Mosfets seemed like a good choice. I used sockets for the OPamps so I could try out different ones. It was developed specifically for the NTE cross of the TL072. I've also tried: BB-OPA37 (didn't work in this configuation) BB-OPA2604 and the AD-OPA270. But not the actual TL072. My impressions of these different OPamps are thus:

NTE: More forward in the very High end (10K Hz +) making for a distinct sizzle, and a good full and round bottom end (250 and lower) Mids less forward but not stressed out or fatiguing. Highest output before clipping and the cleanest sounding distortion when it does. I've used these Op's in this amp for a year now, and they will remain there.

OPA2604: My least favorite. Over all the sound was more veiled and muddy with no part of the spectrum standing out. Clipped at lower levels of input and sounded AWFULL when it did.

OPA270: Highend less forward but more open. Mids very open. Both the Highs and the Mids were more crisp and had a very good sound. The bottom end was a bit less rounded and deep. Again, clipped at lower levels and while the distortion wasn't quite as harsh as the OPA2604 it was still harher than the NTE. I liked this chip and am thinking of designing A similar amp using a class A output stage to compliment the sound of the mids and the highs. The local feedback around the 2 OP blocks will need to be tweaked to make the best use of this device.

The Power rating is right at about 25W calculated with a 1K sine from a CD player. I have no idea what the THD is as I don't have the equiptment to test this. I've also Designed a higher power version that is about 60-100W and can post schematics for it if there is any interest.


Anyway here's the schematic:

The Higher power version actually has fewer parts. it just uses an OPamp that can handle higher voltages and thus swing a higher output. I'll Post the schematics tomorrow or so, as I only have a hand drwn version at the moment (and no scanner).

-D.

Okay, so it didn't really take me that long to get this new schematic into the computer... and I made a few mistakes in the above schematic (getting used to the sch cad program). so...


here is the corrected above schematic for the lower power version:

And the Higher Power Version:

45 Volt rails would be a bit too high as the 452 is rated for a Max of +- 40. Tho I feel that this would be pushing too much. I'd prefer to keep a bit of a saftey margin.

Adding more output transistors would be a good idea for 4 ohm operation with the higher power version. The lower power version is fine with a 4 ohm load.

-D.

The type of PNP isn't really all that important. I use whatever predriver transistors i have on hand, they just need to be able to mount to the heatsink. I wouldn't use this circuit as it is with a 2 ohm load, i would hesitate to use it with a 4 ohm load without at least one more pair of output transistors. For bridging I suppose you could use a bridging adapter. For this type of heavyduty use again, i would add more transistors in the output using source resistors, and add protection circuitry.

As for its simplicity, the OPamp provides all the voltage gain and the Output transistors provide the current gain. This is not really a new concept, but i had not seen one done with a higher voltage OPamp like the ones from TI.

-Dozuki