Differential Probe (diy), i designed myself a differential probe

Differential probes are so expensive.. So i thought. Why not design one myself using opamps.

so.. needs for a probe that is useful :

Very high input impedance

selectable 1x/10x/100x. because the voltage at opamps should not go to rail. and also if looking at ripple one might like the least attenuation of the signal.

DC block and AC block


so. the input is a differential T attenuator with imedance always over 1M. the 1x/10x/100x control is the single resistor in the middle.

because of simulating with Ideal opamps i got some specs from lm1260 that have an average input impedance of 17M and I placed those into the circuit. It revealed that i had to compensate for input impedance, since the input stage already has an impedance of 1-2 Meg.


Now. the opamp configuration is a simple instrumentation amplifier.

The coax cable is simulated as simple as possible, a 200pf capacitance to common.
and that gives a cutoff frequency at 8 MHZ.

now for my first rf circuit. i find that quite good. also, it is more than enough to look at the basic circuits we use and construct at school.

so, here comes the big question, is my circuit crap?
I'm still only in my first year of electronic engineering school, so I might have missed many things.



Original file

Qucs schematic

thanks ,
- Magnus

C3-C6 and D5-D8 = do nothing circuit

OP5 = ?

R20 and R21 = ?

C3-C6 and D5-D8 should have been represented as switchable instead of direct coupling into the opamps. sorry
the network with C3-C6 and D5-D8 is a DC blocking scheme with reverse voltage protection for the caps.


wow. I cannot believe that I did not see the OP5.

I am ashamed of myself!

Thanks for looking at it

Simple enough, but getting HF performance, and maximizing CMRR, are challenges of design, understand, and a lot of tweaking. Very high precision resistors are required; 1% will not do, because 1% error in a 100:1 differential attenuator makes 100:1 common mode gain as well.

Tim

I thought of using a wheatstone bridge to match resistors, lots of work, but better than buying 10$ resistors

balancing the resistors is easy, keeping them together isn't.

at a minimum you need lowtemp co resistors, then match them yourself.