LM324: Indirectly driving capacitve load with DC accuracy

Hi Praful,

there are learning videos discussing stability issues of OPAmps from TI:

training.ti.com/ti-precision-labs-op-amps

Kai

Hello Paul ,

I have designed my circuit ( given in previous post )with following values:
1. CF= 100pF, 50V, 0603
2. RF = 22K, 1%, 0603
3. RISO = 47E, 1%, 0603
4. CL = 0.1uF, 50V, 0603
Also, i have added cap of 100pF directly between input pins of op-amp. Non inverting pin is connected to 1.65 VDC. (1.65 VDC needs to use as reference for ADC)

For the application freq range of 80 MHz to 1 GHz, i am observing gain peaking for 110 to 120 MHz only which is causing offset addition on ADC pins. Else gain response curve is 20dB / decade.

Let me know, how i can further reduce this gain peaking.

Thanks in advance.

Prafulla

Hi Prafulla,

is the LM324 used in this circuit? And is your frequency range really 80MHz...1GHz?

Can you post a schematic of your application?

Kai

Prafulla,

A schematic is needed. I would also highly recommend removing the cap between the input pins of op-amp, that is usually not required and most of the time only causes issues for stability.

-Paul

Hello Paul,

I am using this circuit to generate DC offset reference voltage of 1.65 V for my ADC measurement ( Current of 11mA to 6A). Currently i am using MCP6L01 op-amp which i needs to be replaced with TI device (If gain peaking remains unsolved).  While performing RS test (IEC61000-4-3) on my board, where radiations of  freq range 80 MHz to 1 GHz are used, my reference voltage shows peaking in freq range of 110 MHz to 150 MHz. Which is causing shift in my DC offset resulting in wrong ADC measurement. If i remove cap of 100pF between input pins of opamp, same peaking is observed for 110 MHz to 300 MHz. For rest of freq range , ref voltage remains stable. 

Kindly suggest me the design Aid.

Below is my Circuit diagram:

Hi Praful,

the best cure against radiated interference is to add a shielding. Just take a piece of metal, connect it to a suited signal ground and cover the susceptible circuitry.

In any case, your circuit is unstable and needs to be modified. Also, as Paul already mentioned, the 100p cap between the inputs of OPAmp heavily erodes the phase margin and can cause instability and oscillation.

I would give the following circuit a try:

The phase stability analysis assuming an open loop output impedance of OPAmp of up to 2k shows, that this circuit can work in combination with an OPAmp with a unity gain frequency of under 1MHz. But if you take a faster OPAmp, C3 should be omitted!

Kai

Hello Kai,

Thank you for your support. I tried out the RS test with your given circuit and found the gain peaking for frequency range of 210 MHz to 220 MHz.
Then i changed C2 to 1nF instead of 10nF resulting into no gain peaking. (R3 * C2 = R2 * C1)

Can you please share your considerations and calculations for above given circuit.

Regards,
Prafulla

If it works, take 1n.

Kai

Hello Kai,

Thank you for your support. Can you please share your considerations and references for design calculations for above given circuit.

Regards,

Prafulla