High frequency, high voltage amplifier

Thread Starter

JosipMiller

Joined Nov 8, 2010
4
Hello,

I searched net and and got a lots advices from colleagues but I am having trouble to build relevant amplifier that meets requirements I need. I would not go to details of amplifiers purpose because it would take large amount of text because it is intended to be part of very complicated scientific instrument (old amp is fried beyond repair and new amps are very expensive).

The amplifier needs to meet following requirements: to amplify sine wave signals to 75V (peak to peak) on the output, in range from 300 kHz to 4 MHz. Amp needs to drive capacitive load of 10 pF.

I have build some prototype, but I have problems with linearity in this frequency range. Does anybody have some idea or even project that could be recommended ?

Thank you,

Josip
 

t_n_k

Joined Mar 6, 2009
5,455
You've only specified output voltage into a 10pF load - is there any power requirement for the load? Or is the load just a 10pF capacitance across the full frequency range?
 

Thread Starter

JosipMiller

Joined Nov 8, 2010
4
You've only specified output voltage into a 10pF load - is there any power requirement for the load? Or is the load just a 10pF capacitance across the full frequency range?
Thank you for the response. :)

Answer is yes. The load is mass spectrometer measuring cell; amp is connected to two plates in high vacuum (basically, a capacitor of 10 pF). So, there is same capacitance over full frequency range.

I have made three prototypes so far and they drive load very well, problem was not the load but amplifier itself. It shows significant signal fall when signal is higher frequency than 2 MHz regardless of the load (even if the load is disconnected). I tried many solutions but so far, the best was solution based on vertical output of cathode oscilloscope.

..

Josip
 
Last edited:

Thread Starter

JosipMiller

Joined Nov 8, 2010
4
Something like this might be useful.

http://www.national.com/ds/LM/LM2451.pdf
Thanks. :) I didn't try this one but I tried with some other video amps from ST. Bandwidth was OK, but output voltage was the problem. However, I will try this chip, I guess it is available in local electronic stores.

During past months I also tried with some chips from Apex but problem was the bandwidth. PA85, for instance can work on 2 MHz in best possible scenario (with additional preamp, so gain can be reduced and thus bandwidth improved). I guess that only professional solutions are MSK products:

http://www.mskennedy.com/

Few hours ago I just simulated another discrete prototype based on oscilloscope vertical output so this is the schematic (in this simulation it performs excellent - schematic is just finished, it is not yet arranged properly):
 

Attachments

Thread Starter

JosipMiller

Joined Nov 8, 2010
4
Older CRT driver ICs didn't have to be high bandwidth as resolution wasn't that critical, the newer ones do.
Yes, I've notice. :) First I used one from old Nokia CRT monitor, these new ones are certainly better - only problem was the output voltage (at least I tried those directly available from local stores like TDA 9533 and TDA 9535). Anyway, video chips are of reasonable price so if this new discrete prototype does not work properly I will turn to one of these solutions.

Thanks.

..

Josip
 
Top