Wideband amp circuit

DC frequencies.

That's +/-20V at +/-400mA = 8W peak, right? And it requires a 250V/us slew rate. The APEX PA09 amplifier can do that if compensated for high-frequency G = 100 (with a loop-gain reducing R + C across the + and - input pins).

One of these with an output current booster will do the trick, plus it's much cheaper than anything from Apex (although Apex products are great, too):

The Buur-Brown opa552 is a nice part, able to be powered from +/-30V supplies. But sadly it's got a slow slew rate of only 24V/us. If Luke wants 40Vp-p at 2MHz, then he'll need S = 2pi f A = 251V/us slew rate, as I said, which is 10x faster than the opa552. Sorry.

One part that might do the trick is Analog Devices' ad815ay. This can only be powered from a maximum of +/-15V, but it's a dual amplifier and when used in bridge configuration it can deliver +/-25V at +/-500mA, with a 900V/us slew rate (I was able to pick up a supply of their 15-pin SIP heat- sink tab-style power package before they discontinued it).

That's just a simple CE with a simple simulated inductor step-down load...piece of cake. Take your question to the Basics ng.

Nowadays, with high-performance DC wideband amplifiers available, it's tempting to use one of them instead of a traditional class-A amplifier. But Luke could consider a class-A amp amplifier running with a Vce of roughly 25V and a current of roughly 500mA, which would have a quiescent power dissipation of about 12.5 watts. Then a 100-to-900mA collector-current excursion would deliver +/-20V into a 50 ohm load. The collector-supply inductor would have to be greater than 80mH (that's pretty large!) for a 100Hz low-freq rolloff. The output-node capacitance would have to be under 1600pF for a 2MHz high-frequency rolloff. No doubt the large 80mH choke would deliver some of that, so a second smaller series inductor would be wise. There's more to consider, emitter degeneration, input gain, class-A biasing, etc., but it's evening and duty calls.