Hello all,
I have to read a low amplitude dc signal (1~2mV) where I also have a low amplitude ac signal (10~20mV 100kHz). How can I isolate these signals eachother. I tried to use a passive low pass filter by grounding the ac signal over 1000uF capacitor. But I still see some parts of the ac signal.

 

I would suggest amplifing the signal first with an op-amp circuit to increase your DC voltage to a more useful level; then deal with the signal processing to seperate the two signals.

You mentioned a low-pass filter? Would not a high pass have been the first choice-to block the DC voltage (desired signal) and allow only the 100 KHz signal to pass?

 

I would suggest amplifing the signal first with an op-amp circuit to increase your DC voltage to a more useful level; then deal with the signal processing to seperate the two signals.

You mentioned a low-pass filter? Would not a high pass have been the first choice-to block the DC voltage (desired signal) and allow only the 100 KHz signal to pass?

I have the same idea with you, first I should increase the dc but sometimes I have dc peeks so I should also protect my op-amp, is zener diode enough what you think?

About the filter, my idea is by using a simple rc circuit, which has the capacitor grounded, pass the ac to ground and block the dc.

 

Think either active filters or LC types, it will do a cleaner job. How high are your DC peaks?

 

Isolating the AC is relatively easy, although level of accuracy is also relative, but just use a series cap to block the DC. Another possibility would be to drive a signal transformer, this will also block the DC voltage. Something like a T1-1T type of xfmr maybe a good starting point, I think they're good down to 100KHz or so; you could also use a T1-3T, or so, if you wanted some gain (these are designed for 50 ohm RF circuits).

To measure the DC voltage, you may need to use some really good (very low bias) opamps to phase cancel the AC signal.

 

Think either active filters or LC types, it will do a cleaner job. How high are your DC peaks?

15~18VDC peaks I did measured, my idea is to use a RC Circuit, which has 100Hz cutoff frequency and get rid of AC, then I can use an opamp and get a reasionable DC voltage for measurement.