Protecting cv inputs

Can anyone help? I was following a thread a few weeks ago called ‘cv inputs: how to protect them from negative or too high voltages’. Pichenettes gave two possible solutions - quad rail-to-rail op amp and Schottky diodes. I was sure that there was a circuit diagram for the op amp solution, but it seems to have disappeared. (I think it involved feeding the V out back into the - in). Is this a viable solution? If so, is it possible to power it directly from the +5v on the control board (I would use a twin op amp rather than the quad).

I haven’t tested it yet, but fcd was helping me out with that. Need a couple 4148’s (one to the CV IN & one against ground) and a Zener 4.8 between the CV IN and ground (against ground). I’m on vaca atm, but I was planning on breadboarding on Monday.

EDIT: oh and a resistor on the CV IN after the 4.8v.

I don’t recommend the op-amp solution and I have edited my thread to avoid confusions - because it will only work with op-amps with built-in input protection and rail to rail I/O (which are not always easy to find!).

My favorite solution, which would also be the one with the highest part count, would be to put one (summing) inverter powered by 5V / 5V so that you could sum multiple signals to the same input of the Shruthi ; then a second inverter powered by 0V and 5V (made with a rail to rail op-amp such as a cheap MCP6002) which would do the clamping. The op-amps could also be used to provide active filtering of the CV to prevent aliasing.

Thanks Pichenettes - I was wondering where it went! I’ve already ordered a TLV272IP (Texas Instruments) - would that be safe for the simple op amp solution - it seems much simpler than your revised suggestion. Also, is the Schottky diode solution OK?

The diode solution is OK.

but not the TLV272 ?

Sorry to keep bothering you Pichenettes - I’m looking at forward voltage drop values for Schottky diodes - the values vary according to the current - what kind of current could I expect across the diodes? 0.25 V @ 0.1 mA and 0.45 v @ 10mA. Is that too high?

That should be 0.25 V at 0.1 mA and 0.45 v at 10mA.

Depends on the output impedance of the gizmo you connect to the inputs…

OK - sounds a bit risky - too easy to exceed the maximum voltage and fry the chip!