High voltage amplifier and high frequency

[reeddreed]

Hey, I did so here is layout now ... I added a high-speed operational amplifier ...
the input signal is 5 MHz sine wave 2 Vpp
problem is the slow growth of the signal up to 100 Vp
underneath the chart of simulation

 

[erikl]

Hey, I did so here is layout now ... I added a high-speed operational amplifier ...
the input signal is 5 MHz sine wave 2 Vpp

Congratulations!

problem is the slow growth of the signal up to 100 Vp

Seems this has something to do with your C1*R3 time constant = 80µs. You could check this easily by changing R3 .

 

[FvM]

You should also remember, that non-overlapping gate voltages would be preferred for safe operation will real MOSFETs. The Supertex driver has two separate channels rather than one to allow individual gate signal generation. At least, you should slowly increase the supply voltage, checking for excessive transistor losses when implementing a real circuit.

 

[reeddreed]

thank you very much for your advice ...

Now I have this problem .. I put the feedback in this way as seen in the picture ...
Unfortunately, the signal is still distorted, I do not know how I can reduce voltage, so that the signal does not try to driven as to the max ...
I try to combine the feedback, but I do not know what else to do

The second circuit is left only the capacitive load, and as seen in the top half have a nice sine wave, but I do not know how to do it on the bottom

 

[erikl]

Your overall gain setting is too high: it is not -(1+(R10/R11)) - as you might expect - but -(1+(R10/( R11||R8 )) , which is -111 for the bottom circuit - neglecting your frequency generator's output impedance. Hence your circuit still is totally overdriven.

 

[reeddreed]

. I just tried to throw R8 in order to reduce the enhancement .. but did not improve the simulation

What else can I try to eliminate this overdrive?

 

[FvM]

I'm surprized, that you intend to make a linear amplifier. You never mentioned that before.

A linear amplifier would use e.g. a class AB output stage, involving at least a small bias current for the output transistors. I don't think the presently implemented method to drive the output transistors will work for a linear amplifier.

 

[erikl]

I just tried to throw R8 in order to reduce the enhancement .. but did not improve the simulation

Not sure what you did by "throw R8", but if you removed it, your gain probably is now -(1+(R10/50Ω)) = -201 (if the output impedance of your frequency generator is 50Ω), i.e. larger than before!

What else can I try to eliminate this overdrive?

Unlink R11 from GND and use it instead of R8, to achieve an overall gain of ≈10 !

It is not safe, however, if this configuration is still stable.

Another possibility is to reduce the signal output from the frequency generator by about an order of magnitude.

 

[reeddreed]

I did like you saidwith the R8 ... but unfortunately, the amplifier is not stabilized ...

Trying to control circuit as in Push Pull Power Amplifier

**broken link removed**

but something it does not go for a frequency of 5 MHz

 

[erikl]

I did like you saidwith the R8 ... but unfortunately, the amplifier is not stabilized

That's what I - most likely - anticipated.

Did you try my other suggestion

Another possibility is to reduce the signal output from the frequency generator by about an order of magnitude.

with your original (gain ≈ 100) circuit configuration?

 

[reeddreed]

yes .. reduced the output signal from the generator, but it did not work properly .. too low signal level

that in the original signal G≈100

 

[FvM]

In my opinion, you have to move one step back and specify the amplifier requirements.

If you need a feedback amplifier to achieve a certain gain accuracy or output signal quality, then the output stage circuit and loop compensation has to take this into account. I think, a resonant (= small bandwidth) piezo actuator won't derive much benefit from a low distortion driver amplifier, because harmonics are suppresed anyway.

 

[reeddreed]

My professor wants the signal was that distorted as little as possible ...
It was nice to that sine signal was no distortion, I know that such a distortion of the transmitter was not prevent to the action ...
For me, it could have already been, because I do not have any idea how to control it perfectly

 

[FvM]

As i already mentioned, the output stage should be redesigned for class AB operation. And a suitable compensation scheme apllied.

 

[reeddreed]

I came to conclusion, together with my professor ... that the program probably ignores feedback in circuit

I am now trying to implement circuit to PSpice ..
I have a problem with the implementation of components such as transistors and operational amplifier ...
I have a library but in simulation is an error that the elements are undefined
know how to define the elements of spice?

 

[FvM]

I came to conclusion, together with my professor ... that the program probably ignores feedback in circuit

Do you mean Multisim? It basically knows about feedback.

I have a problem with the implementation of components such as transistors and operational amplifier ...
I have a library but in simulation is an error that the elements are undefined

The respective circuit libraries (*.lib) have to be imported in addition to the symbols (*.slb).