This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

THS3491RGTEVM: Adding offset voltage to the amplified signal

Part Number: THS3491RGTEVM
Other Parts Discussed in Thread: THS3217, THS3491

Hell there,

I have a THS3491RGTEVM and want to use it to amplify 20-40M square wave to 10Vpp (with input up to 1Vpp). I can get the waveform amplified to 10Vpp with a gain of 20V/V with input of -250mV to 250mV (the opamp is set to default non-inverting configuration, Rf=499 and Rg=27ohms); however my system needs the 10Vpp reference to 0V, ie, between 0V and 10V; I've tried to amplify a 500mVp square wave using the amplifier but the output is significantly distorted, I have to increase +/-VSS to 20V to get proper squarewave (but still distorted and I don't think you would recommend it). Could you please provide me with any suggestions (please note that I cannot use bias-Tee to do the job). My plan was to configure the EVM to inverting configuration and connecting J2 to a voltage source to provide the offset voltage. In this case I hope I can use 500mVpp to achieve the 10Vpp required voltage. Do you think if this is a viable solution? How should I select Rf and Rg (Do you have a recommendation table for the inverting configuration)? I've tried a few combinations but the output is not great (but I can offset the waveform by adjusting the voltage on J2).

Many thanks 

Kelly

  • Hi Kelly, 

    Using an inverting configuration with a bias provided to the non-inverting pin should work for what you are trying to do. You just need to make sure you have enough headroom on the output swing to the supply. As for the resistor values, you can just choose them using the non-inverting table but add 1 to your gain value. For example, if you want a non-inverting gain of 9V/V chose the table value for 10 V/V (or whatever is closest). 

    Regards, 

  • Hi Jacob,

    Thanks for the answers.

    I did find my amplified signals are much better with 20V dual supply setup, with +/- 15V setting I got much worse overshoot and undershoot (I’ve tried a few Rf and Rg combinations). If I stick with+/- 20V setup, will I kill my opamp eventually? The current consumption does not change much between 15V and 20V. Also I understand that section 8.4.3 of the opamp datasheet gives suggestions on configure the opamp to single supply mode, but I’m still a bit confused. Can I simply modify my EVM circuit to covert it the single supply mode? If I can which components to move/add?

    Many thanks for your help.

    Kelly

  • Well Kelly, if you need a 0 to 10V output, +/-15V should be fine, it seemed you might have been trying to use it as a comparator. (with input up to 1V??) if so, do the comparator function first and have its output drive the op amp.

    1. Running non-inverting, your gain is 1+499/27 = 19.48 and your V+ input needs to be 51mV. If you put in +/-500mV you will clip. 

    2. You can run offset supplies, say 20V and -10V. 

    3. That is a low resistor to drive if you go inverting. - you can also drive +/-51mV in the V+ node and provide an output level shift with a -0.27V bias on the Rg - that has to be low impedance to get the pulse response. 

  • Hi Michael,

    Thank you for your reply.

    The input is limited by my DAC which has output up to 1Vpp (-500mV to 500mV), and I want to amplify this DAC to 10Vpp (0 to 10Vp). Using my configuration (gain of 19.48 as you calculated) I can achieve 10Vpp with small overshoot and undershoot with input of -250 to250mV, in +/-15V dual supply mode; but the output is between -5V to +5V and I need to add a 5V offset to shift the voltage up to 0 to 10Vp. I've tried to configure my DAC to output 0 to 500mV but the output is then distorted and I have to increase the supply to 20V and get proper output (but still distorted).

    I was wandering if I configure the EVM to inverting mode, and use J2 to add voltage offset, in this case I might just need to amplify -250mV to 250mV signal to -5V to +5V, and then input a 25mV DC to shift the signal up by 5V, then I can get 0 to 10V output. But I haven't succeeded. Could you let me know if my modifications to the circuit are correct: change R1 to 0, remove R2 and R7, Rf =499 and Rg=25;

    I've also added some test results to this post. 

    Kelly

    opamp tests.pptx

  • Thanks for the detail, your 0 to .5V input should work on +/-15V supplies. 

    Looks like it is getting slew limited, must have a really fast input edge? What is your load. 

    The other part you should be aware of that is very flexible for DAC interfaces is the THS3217. Lots of places to do level shift there - that is what the onboard buffer was for, to drive the reference line with a DC offset. 

  • The input rise time is around 0.7ns and the load is 50ohm.

    I do have a THS3217 but it cannot provide 10Vpp easily (also I think even it can the bandwidth is small and the amplified square wave will not be great). I can give it a try. If you are familiar with THS3217 could you let me know how to modify the EVM to achieve my requirement (ie 0-10Vp with DAC input up to 1Vpp)? 

    Also back to one of my questions, are my modifications for inverting mode correct?

    Kelly 

  • Well now - that is a demanding output, 10V in 0.7nsec is 14000V/usec - exceeding even what the THS3491 can do, you need to RC filter that input edge to >1.25nsec to stay under 8000V/usec. Also, that 10V into 50ohms is right at the limit of 500mA peak, this may call for paralleling 2 THS3491 stages. 

    Your DAC must be complementary output to have that kind of speed. Yes, the THS3217 cannot do this directly, but you could still use the D2S stage to get the level shift and some gain, disable the OPS and drive the D2S output into the THS3491.