Centering output with a cap,then to comparator ..???

[wannabegeek]

hi all....

I 'm stuck with something that seems pretty basic...

I have a triangle wave coming out of an active filter that's all above zero, Vpp = 0 - 2 V.
I want to connect that output to a decoupling cap to center the signal again, so that
Vpp = -1 - 1 V.

Looks great on the scope...Next, I want a comparator to get make a square pulse train that swings both pos and negative.

This is where I fail, as the input impedance on the op amp comparator
is huge and the cap just charges up.

How is this done...?

TIA,
wbg

 

[(*steve*)]

You need to provide a DC path from the input of the op-amp to ground.

 

[wannabegeek]

You need to provide a DC path from the input of the op-amp to ground.

Yes, that sounds like a basic ...thanks Steve...that makes total sense...
Did I mention that my electronics background comes from one crash course
in a physics program...?

If I get my project to do something interesting I'll start a thread in the projects section to
"show it off"...

Cheers !
wbg

 

[wannabegeek]

Trying out the "fix"...

I tried putting the DC path at the input of the OP AMP...no luck.

But I did stick a 1M Ohm resistor to ground and that worked. I'm
using a 0.1 micro F cap.

In fact, this solution makes more sense to me...blocking caps in general never seem to work for me as initially expected.

 

[wannabegeek]

It appears that I'm stuck with this again...

I've attached a schematic...

The input is actually a photo-diode with op-amp for gain. I'm sending a PWM
20kHz carrier via 650 nm laser diode. The photo-diode receives this pulse, which has a lot of chatter in it, and gets sent to the filters.

To keep the schematic simple, I just put an AC voltage source in.

I'd like to not use a negative voltage source. I'm using an OP07 op amp, and wondering if I should be using uni-polar op-amps....

I've got the circuit working when using a negative supple and no voltage divider.
I don't see why I need a negative supply though.

Thanks again,
wbg

 

[(*steve*)]

But I did stick a 1M Ohm resistor to ground and that worked.

Yep, that's what I meant. The input of the op-amp requires a DC path to ground (or to a virtual ground).

 

[wannabegeek]

this http://www.analog.com/static/imported-files/application_notes/AN-581.pdf
article is helpful...I love these older application notes...

wbg

 

[wannabegeek]

I totally forgot about setting all the f_3dB points...I was taught how to bias a transistor in school...the application note I posted covers it.

I got the circuit to work with a single positive supply....and it wasn't worth it...
too many more parts and a total PITA...two batteries is not such a big deal....

However, I did try pushing more current through my laser module, 30mA, and it now
has a small instability...or the Rx has an issue...

Not sure but I broke my circuit...while making it better...

Now I've got something to do all weekend...

cheers,
wbg