How can I amplify a 0.5Vpp 45Mhz sinusoidal signal to get a gain around +40

Selim Kara

Other Parts Discussed in Thread: THS3061, OPA847, THS3091, LMH6702, TINA-TI

I'm trying to develop a QCM (quartz crystal microbalance) measurement system to drive a special 5MHz quartz crystal on it's overtones to track it's resonance frequencies. I'm using a DDS signal generator by AD9851 which has an output signal gain 1Vpp at 5MHz to 0.5Vpp at 45MHz. I've succeeded to design a circuit to drive that crystal on 1st (5MHz), 2nd (15MHz), 3rd (25MHz) at around 14Vpp. I'm using OPA847 (Vs:+-5V) and THS3061 (Vs:+-12V). But I couln't amplify the 35 and 45 MHz signals to achieve those levels of gain. It seems on the datasheets of these op-amps that it's possible to get that gain level. But,... What can I do to amplify my 35 and 45 MHz signal to achieve 20Vpp voltage gain? Any idea?

Dr. Selim Kara

over 10 years ago

0 Hooman Hashemi over 10 years ago

TI__Mastermind 22825 points

Hello Salim,

Your application requires a high slew rate (>3,000 V/us, most likely current feedback type) amplifier which can operate from a large supply voltage (+/-15V) and that it can maintain close to 100MHz of bandwidth when operated at a large gain (20Vpp / 0.5Vpp = 40V/V).

The closest device I can find is the THS3091 which can do 135MHz with 4Vpp output (G=5V/V):

But you need more gain and more swing! Operating at a higher gain / larger swing will impact large signal bandwidth so that you could jeopardize your 45MHz need.

So, you may have to get to your 40V/V gain in a couple of stages, say two equal gains of about 6.3V/V, with THS3091 as your 2nd stage active element.

For the 1st stage, you could use a lower supply voltage (+/-5V for example), higher speed device due to much lower swing. The LMH6702 might be a good choice for the 1st stage. Here is the LMH6702 response with 2Vpp output swing showing in excess of 100MHz large signal bandwidth with G=10V/V:

Here is what TINA-TI predicts for large signal (20Vpp) swing at the output of the 1st and 2nd stages (1.5ns rise time). Simulation shows no problems with 20Vpp output swing at 100MHz repetition rate (but you may have to build it to make sure this matches bench behavior):