integrated power amplifier 700kHz

[raff5184]

So if you are looking at 700kHz, gain=100x, Vin=+-1V, then we are talking about 70M GBW at +-100V output, that's quite a lot.
Keep in mind that even at 500 Ohm load, the load is only +-0.2A, which makes it easier.
Apex PA90 might be a good choice, but the $137 price tag makews me think twice. For such an amplifier (700kHz, 20W peak), I won't spend anything more than $15.
I would recommend a +-17.5V powered AD825 OPAMP to provide 15x gain (10.5M GBW), then use a bootstrapped AD825 to obtain another 6.7x gain (4.7M GBW), finally buffer the output with a complementary class B BJT buffer.

Please can you give me more details bout the second part?
Bootstrapped opAmp is to reduce the impedance so that I can give a higher voltage in input without distortion?
Why a BJT buffer?

Someone suggested me a multistage amplifier with the AD825 but I can't simply give the output of the first AD825 in input to the 2nd AD825, assuming I configure both with the same say inverting feedback configuration

[raff5184]

Do not cascade an open loop opamp to another one, providing only global feedback. It may well turn into an oscillator!

Yes, I meant to configure both in inverting configuration, with two similar -but separate- feedbacks.
Thanks for the paper I'll read it and see how to create the cascade.

[raff5184]

I would recommend a +-17.5V powered AD825 OPAMP to provide 15x gain (10.5M GBW), then use a bootstrapped AD825 to obtain another 6.7x gain (4.7M GBW), finally buffer the output with a complementary class B BJT buffer.

so, I'm re-arising this old post. About the powering of the 2 cascaded op amps, do I have to use 2 independent supplies like -+17.5 V (by the way, why exactly 17.5?) for the first stage and another +-Vsupply for the bootstrapped op amp?

Also, I was able to have the second opAmp with a gain of about 6-7 (without bootstrapping network) but when I connect the opAmp to the bootstrapping circuit, it doesn't work any longer or its output is the same of the input. I am using a bootstrapping circuit very similar to the one in Figure 6 on this paper, but with and AD825: http://www.proaudiodesignforum.com/images/pdf/Bootstrapping_Your_Op_Amp_Yields_Wide_Voltage_Swings_King_Watkins_EDN_May_13_1999.pdf

[raff5184]

As for the 17.5V, it's chosen so it is lower than +-18V max operating voltage of AD825, while it is still high enough to give you output swing range. If you use my gain configuration (15x and 6.7x), your first stage has to be able to output 15V peak for 2V pp input (assuming ground centered).

Yes I got very close

AD825 has 2V max output dropout, so to drive +-15V, you need to supply it with at least +-17V.

Do you get this info from Figure 7 of the datasheet, right? Because I don't see it elsewhere on the datasheet.

Assuming you used R2:R1=R4:R3=9:1 (+-120V VCC, +-12V opamp supply), then Vo=100V, Vcm=50V, Vco=112V, Veo=88V, Vcm<Veo, therefore it won't work.

No I used R2:R1=R4:R3=3:1. What's the criterion to choose of R1,R2,R3,R4?
Also  I noticed that it is better not to insert R_B