Hello there,

I'm trying to build a "doomsday-machine" and I'm kinda stuck at this problem.

I have a 200mV peak-to-peak signal riding on a +6V offset, now I would like to remove the +6V offset so that the signal oscillates between 0-200mv, however the problem is that i'm constrained by:

1. low low frequency 0.001Hz or less, so I can't use a simple cap
2. +12V single supply environment, so I can't use a diff op-amp

I would appeciate any suggestions, I'm really stuck.

Thanks in advance

 

You just need to create a virtual earth at 6V for your differential op amp.
Careful, I think Austin Powers uses this forum sometimes.

 

Dear Markd77

Thank you for your suggestion,

however I tried to run a simulation in Proteus based on your suggestion, and it seems that it is not possible to reach 0V output from a op-amp with single supply (virtual ground) due to the op-amp's internal voltage drop of about 1.5V i.e. minimum obtainable DC offset is 1.5V, or maybe I did something wrong?

Damn that Austin Powers and his mojo, evilness will soon prevail (I just need to get this circuit to work).

 

You need to try a different op amp. There are many that will allow outputs to ground when using a single supply. there are also many that won't.

 

Dear StayatHomeElectronics,

Thank you for your reply, since I don't have that much experince with different op-amps, maybe you could kindly suggest one that will allow output to ground?

thank you and best regards

 

Such as the LM324, industry jellybean, that will work down to ground.

 

Sorry about that... I got called away...

The LM324 is a good recommendation. One thing to know is that these op amps will get you close to ground (typically within 5 mV, maximum within 20 mV) but do not get you all the way to ground. That is typical performance of these kinds of amplifiers. For small signals, that can be an issue as well. At least something to keep in mind.

 

Dear tom66 & StayatHomeElectronics,

Thank you for your suggestion, I tried run the simulation with the LM324, it is definetly an improvement as it goes down to 500mV, I guess it would go even lower in reality, but it is a small signal and according to StayatHomeElectronics best case for me would be 2.5-10% error, which unfortunately is not acceptable. So I guess I will have to generate a negative supply for the op-amp, I was hoping I could avoid this additional circuitry, but it seems to be the way to go.

thank you and best regards.

 

What you want is a "rail to rail op amp" which can drive its output all the way down to gnd. Except it won't do it perfectly--to get to real zero volts you'd need a negative power supply. You could do that, if it's absolutely necessary, with a charge pump. But would a system with an offset (pseudo-ground) work for you? Say that you decide that instead of 0-200mV you use 1.0-1.200V--could your circuit work that way? It would solve some problems.

 

Dear John P,

Thank you for your suggestion about pseudo-ground, I don't think it would work for my application since currents are up to 10A so I would need expensive heavy-duty components. As you suggest yourself, the way to go must be to generate a negative supply via a charge pump.

Thank you.