0
\$\begingroup\$

As a hobby project I'm trying to drive a small RL load which should draw about 1A pp using a sinusoidal voltage coming out of a series of op amp stages. The signal should be applied directly to the load. How can I do that?

With BJT it is pretty easy to do that when using a DC voltage drive and output such as below:

schematic

simulate this circuit – Schematic created using CircuitLab

The schematic is just for reference, to show you the simple driver I've been using for dc loads. In the case of this question you might as well substiture the load with a constant current source at 1A.

However, suppose my signal is a sinusoidal 1V pp wave coming from an op amp. The op amp cannot supply 1A and the transistor will be under stress with 0.3Vce and 1A pp current. This "issue" however can be solved by choosing a transistor that can handle 400mW and a 1.5 A max collector current. The bigger problem is that this circuit is going to reproduce only the positive half of the driving sinusoidal wave since the load is grounded.

A possible solution is to change the ground voltage to -1V but that solution has two main disadvantages:

  1. During the upper part of the sine wave, the voltage across the load becomes 2V drawing which is not acceptable.
  2. The load draws more current making the transistor dissipating even more power and probably exceeding the maximum ratings.

What other simple options do I have? Maybe using n channel mosfets?

EDIT: The sine wave signal frequency is in the range 10 to 50Hz.

\$\endgroup\$
  • \$\begingroup\$ You haven't mentioned what frequencies you are using. \$\endgroup\$ – Andy aka Jan 6 '17 at 14:08
  • \$\begingroup\$ @Andyaka very low frequencies, ranging from 10 to about 50Hz \$\endgroup\$ – mickkk Jan 6 '17 at 14:09
  • \$\begingroup\$ You have POSITIVE feedback in your circuit. You will be driving with a square wave, not a sine \$\endgroup\$ – Scott Seidman Jan 6 '17 at 14:11
  • \$\begingroup\$ In better wording: this circuit can only drive the load with a voltage > 0 (zero) Volt. You can either superimpose the sinusoidal wave on top of a DC offset. For example 1 V DC + 0.9 Vpeak sinewave, then the signal will be between 0.1 V and 1.9 V. If that is not OK then you should look at how audio power amplifiers solve this because a loudspeaker is very similar to your load. \$\endgroup\$ – Bimpelrekkie Jan 6 '17 at 14:11
  • \$\begingroup\$ @ScottSeidman maybe I was not that clear in my wording, I get that this circuit is not the best for driving with a sinusoidal wave (without some weird work around), that's why I am asking here for a circuit (or some guidance at least) that can. The circuit is just for reference and for explaining my problem. So far I've been using that just for DC loads. Please let me know if my wording is a bit obscure, english is not my first language :) \$\endgroup\$ – mickkk Jan 6 '17 at 14:20
4
\$\begingroup\$

You can use an op-amp and power transistors in a push-pull, emitter-follower configuration to get what you want: -

enter image description here

There is no need to bias the transistors (to avoid cross-over distortion) because, at such low operating frequencies, the overall negative feedback will rapidly push each transistor through that region.

And, if you know that the maximum |peak| voltage on the load is (say) 2 volts, the +/-12V rails to the transistors can be dropped to (say) +/- 3.3 volts or maybe even a shade lower. This will significantly reduce power dissipation in the output transistors.

\$\endgroup\$
  • \$\begingroup\$ Would it be desirable to use a matched pair of NPN/PNP transistors for that circuit, even though the OP did not state any preference for the fidelity of the output w.r.t. the input? \$\endgroup\$ – Andrew Morton Jan 6 '17 at 20:56
  • 1
    \$\begingroup\$ @AndrewMorton given the amount of negative feedback and the really low frequency response required, I would say not. \$\endgroup\$ – Andy aka Jan 6 '17 at 21:19
  • \$\begingroup\$ @mickkk if you are happy with an answer please consider formally accepting it. It's a small price to pay for continued support on new questions. \$\endgroup\$ – Andy aka Jan 31 '17 at 21:36
2
\$\begingroup\$

Realize that 1 amp is no longer an impossible or even uncommon amount of current for OP-AMPs designed specifically for that purpose. In a quick search I found a couple of op amps good to 2Amps output, one being a Linear Technology LT210 and another by TI, an OPA569. I suspect all such options will be more pricey than an add on buffer made from BJTs such as already suggested, but if is a one shot deal you should be able to sample a couple. And of course even at a higher price, a single chip solution might be best if board space is cramped.

\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.