I have a project where I need 30Vp-p, 20mA at 4KHz Sine wave.
I am currently using AD9833(A DDS chip interfaced with PIC18 Micro) to generate the signal with 0.6Vp-p with 0.5V offset at 5mA . I have a boost regulator which is capable of 36V and 1A DC output. I wont have a negative power supply.

I have tried standard inverting and non inverting opamp ciruits from google without much success. I was able to amplify the wave upto 5V only, even though I had VCC as 15V.

Can anyone help me come up with a circuit using opamps to obtain the required output?

Thanks

 

The total supply voltage for the opamp will have to be about 3V higher than the 30Vpp you are trying to get out of it. The opamp will have to be rated for operation on supply voltages greater than 36V. You will have to do some offseting of the input signal to half that supply voltage.

 

The total supply voltage for the opamp will have to be about 3V higher than the 30Vpp you are trying to get out of it.

For example, you could use a split supply of +-16.5V to the opamp. The opamp will have to be rated for operation on supply voltages greater than 35V.

Can you deal with a split supply? If not, then you will need a ~33V supply, and do some offseting of the input signal to half that supply voltage.

I am limited to one voltage source only which is good for 36V. Also the input AC signal has an offset of 0.5V

 

Here is a hack at it, assuming that the DDS output is like a voltage source:

Note the required matching of R1-4. Note the effect of adjusting the trim pot (Green is 300K and violet is 400K. Also note, that while the LT1001 has the voltage rating, its slew-rate is too slow, making the output wave more like a trapezoid. You will have to select an opamp with a faster slew-rate...

 

If you can capacitively-couple the DDS signal to the input of the opamp, here is another approach:

 

I recommend the AC-coupled version. Since you have a tightly controlled signal bandwidth, you can use two resistors and a capacitor to create a virtual ground at 50% of Vcc and eliminate offset voltage errors in the opamp and the need for super-tight resistor matching. Also, is it just me or does the output in #5 look slew-rate limited? Not curvey enough?

ak

 

It is definitely slew rate limited. Here is a third version, this time using a faster opamp.

 

Here is a hack at it, assuming that the DDS output is like a voltage source:

Note the required matching of R1-4. Note the effect of adjusting the trim pot (Green is 300K and violet is 400K. Also note, that while the LT1001 has the voltage rating, its slew-rate is too slow, making the output wave more like a trapezoid. You will have to select an opamp with a faster slew-rate...

View attachment 82218

Thank you good sir! I am going to give it a try right now.

 

Thank you good sir! I am going to give it a try right now.

I used 30V instead of 36V and used 200K instead of 250K. But I got the amplification.
I thank you for the instant help and support.

 

But my digital trimpot only varies from 1K to 40K. How will that affect other resistors in my circuit.

 

Are you using the first circuit I posted in #4? The one in post #7 doesn't need the trimpot...

 

Are you using the first circuit I posted in #4? The one in post #7 doesn't need the trimpot...

I am using the circuit in #7 and I placed a 100k Pot in place of 249k. Sorry for the delay.

 

I am confused. You changed the supply voltage to 30V?

You want to be able to vary the gain of the amplifier using a 1K to 40K digital pot?

Or you want to do it with a manual pot that goes to 100K?

What is the required frequency range of inputs from the DDS?

 

I am confused. You changed the supply voltage to 30V?

You want to be able to vary the gain of the amplifier using a 1K to 40K digital pot?

Or you want to do it with a manual pot that goes to 100K?

What is the required frequency range of inputs from the DDS?

I didnt have a 36V in hand. And I found out that my digital pot only works in 5v circuits, so i am having to do a two stage amplification from 0.5v pp to 5v pp to 30v pp.

 

Assuming that the input to the second stage is now 5Vpp centered on 2.5V, here are the modifications to the second stage:
Output is shown as a function of different values of R5.

 

Assuming that the input to the second stage is now 5Vpp centered on 2.5V, here are the modifications to the second stage:
Output is shown as a function of different values of R5.
View attachment 82871

Thanks I will give it a try... Also I have been looking for "charge pump circuit using PWM" on this forum but havent found much help. I wonder if you could help me or maybe direct me to a thread?
So the requirements are, 40V peak positive and negative alternating pulses of 100 microsecond width. Using 5V powersupply and a micro controller
Thanks again MikeML

 

Thanks ak. I had searched for a fast op and found LM7171, which unity gain bandwidth is 200MHz, slew rate is 4100v/us at the conditions when Vin 13v and Av is +2. Per the characteristics graphs, gain is around 30dB~45dB at 1M~10MHz, GBW at Vs=+/-15v is 200MHz. These look promising. Do you think this op can fit to my purpose?