drive output transistors with an IC

Is this what the figure at the botom of page 8 is doing?
that design increases the output voltage, but not the current..
Are there any designs that take an IC and drive bigger output transistors to provide a high current amplifier??

Would such a design be very inferior to a standard layout? It seems to me like there must be a way to use a $2 IC to take the place of a lot of components in an amp and still allow a decent power rating (~250 watts)

Anyone have any idea's suggestions, explanations why this idea wont work etc. ???

(BTW tried to search, but Ic is too short for search engine !!)
It would seem that this set-up also provides current drive. The print at the bottom states that it brings the output swing up to +/- 70V at +/- 10A. I'd say that this is pretty high current.

Seems to be two camps, 1 that scoffs IC's (OPAmps) and one that embraces them. I fall somewhere in between. While I can't see an IC being 'ideal' I'm of the opinion that they are usable and can generate acceptable results, even the lowly TL072. There is a lot of info on using various ICs for building amps. One way is to use IC's for the voltage stage and discrete components for the OPS (current amplification). A common TL072, a few resistors, a drive transistor, and a couple of mosfets are enough for an approx 15-25W poweramp. Not too bad for 15$ in parts (not including power supply). There are also ICs that are essentially 'amps on a chip' that can be used to build an amp. These can be parrelled or bridged to give higher outputs. I think that the PDF file that you pointed to is a good start for building higher powered amps in this manner. I admit that I don't understand all the topologies that the datasheet suggest and I tend to approach it from the OPamp as VAS point of view, choosing to use an OPS of my liking. If you could get an LM12 to swing 70 Volts without clipping as the datasheet suggest, you could theoreticly make a 300W amp.

the lm12 swings +- 35 volts at 10 amps, that design swings it +-70 volts at 10 amps. how many watts is this ? 612 ? or is it more than just ohms law? This is not enough curent to drive a 4 ohm speaker tho, so could I just double/triple up the big transistors to provide extra curent capability?
Not Cheap!

If you are considering using the LM12, it is not cheap, ~$30.00 as a single unit. I have seen it used in a couple of "high-end" car audio amplifiers that are manufactured by MMATS. You can use this device as an audio amplifier by itself and a few supproting components. I believe its good for 80W. However, I think its original design was for industrial control applications to drive motors, etc. If I remember correctly, you can parallel several together for greater output capability.

Don't forget the heatsinks!


I would think that an IC with a Buffer IC inside the feedback loop would be acceptable to drive a power stage. Say a OPA627 and a BUF634 combo as a combination gain/driver stage. I've got a PCB layed out for this, if you would like I can send it to you. The board is about 2"x2.5" per channel.

The equasion is a little more than ohm's law. For Wattage output it is:

(V*V)/Z/2 - z=speaker impedance. There is a thread discussing this if you want more info. My estimation of making a 300W amp was from the presumption that in the +/- 70V Schematic you could get a 70 swing without clipping. There is a possibility that you could get more, but as you pointed out the 10 Amps would not be enough for a 4 ohm load. You would need 17.5 amps for this.

The Datasheet for the LM12 shows a simple amp for audio that claims .01% distrotion, i'd say that ain't too shabby, but specs tell you nothing of the sound. I was thinking about the topology that you origianally pointed two and I came to the conclusion today that this topology is probably running the OPAmp in some form of Class B, swinging 70 volts into the chip in either direction as the wave form swung that direction. Again this is just guessing. Using this method might bring the distortion levels back up.


I would love to see the schematic for the board that you have, the buffer in the NFB loop sounds interesting.

Serious opamp indeed

Here's a pretty solid opamp that would drive just about anything so it seems. Price is $9.88 US in quantity, so it's barely bearable.


SUPPLY VOLTAGE, +VS to –VS 350V :att'n:
OUTPUT CURRENT, cont. within SOA 60mA :att'n:
INPUT VOLTAGE, differential ±16V
INPUT VOLTAGE, common-mode ±VS

Here You go Dozuki.


  • multi loop buffer.gif
    multi loop buffer.gif
    2.4 KB · Views: 721