Discrete OPAMP Using discrete components (BJT,FET,TUBES) only help

I'm trying to build and piece an multistage amplifier with the specifications below:


1. Two inputs one plus , one negative

2. Capable of driving a 50 K ohm load to + 10 V

3. Frequency response DC - 10MHz + 2 db.

4. The input impedance must be at least 1 Megohm

5. The output impedance must be less than 50 ohms

6. Only discrete devices (BJT, FET, TUBES) may be used.

7. This amplifier will be packaged in such a way that it does not exceed more than 3 inches on any side.

8. Temperature range: 25 -- 100oC

9. The gain must be between 1000 - 1500

So far *attached is what I got (PSPICE) but I have trouble getting the gain to be atleast 1000. Also the frequency does not exactly match at ~8Mhz. I also tried building the circuit on board but is giving erroneous results. Can someone guide me on how to proceed to satisfy the requirements?

The main problem with that circuit is that Q16 and Q10 are connected as emitter followers, so the output voltage is the same as the voltage on the collector of Q5. This severely limits the output voltage swing.

It would be better to connect the collector of Q5 to a common emmitter stage. For example, you could do this:
a) Remove R6 and Q10
b) Connect the emmitter o Q16 directly to +15V
c) Connect the collector of Q16 to the output, instead o to ground

To get a -good- op amp you would have to be pretty
selective about components and especially matching.
The old timey module op amps did a lot of binning and
trimming.

You can find many op amp schematics in old National
databooks; LM301, LM324, LM741 are about as simple
as it gets. And nowhere near "best".

A single stage op amp design may be hard pressed to
deliver Av=60dB without some really good transistors
operated at their ideal bias point. And board level
parasitics will kill your frequency response and/or add
feedback coupling that destabilizes the amplifier.