how to interface a differential op-amp to a single ended output

[mageta]

I would like to have a regular op-amp (single ended) input driven via differential outputs from a differential op-amp. I assume this will be better than going to a single ended configuration right after the diff op-amp (typical configuration).

The question is how to best interface the ad8137 in differential mode to the lmh6619, and is there any need to impedance match or load the ad8137 for best signal integrity?

The input to the diff op-amp is a parallel resonant circuit, this configuration has a high Q. Is there a way to reduce the Q and therefore make it more stable?

 

[Skogsgurra]

The standard way of doing it is to use an instrumentation amplifier. But that is overkill in this application where both signals have low impedance sorces.

So, just use a differential amplifier and connect directly. You only need four resistors for that and you can google for hundreds of diagrams.

Gunnar Englund

http://www.gke.org

--------------------------------------
Half full - Half empty? I don't mind. It's what in it that counts.

 

[djs]

Connect the single ended op-amp as a standard inverting amplifier. Connect the input resistor to one of the differential outputs and connect the non-inverting input of the single-ended op-amp to the the other output.

 

[mageta]

thanks for the replies, trying different things.

now i am finding the LC networks is being unbalanced by the diff op-amp. do the diff op-amps have a high differential capacitance?

 

[Skogsgurra]

What frequency range are you working in? The standard opamps are not HF devices.

Gunnar Englund

http://www.gke.org

--------------------------------------
Half full - Half empty? I don't mind. It's what in it that counts.

 

[mageta]

I am working with 400 KHz, and 2 MHz. I am assuming that the LC network does not require impedance matching since it has such low impedance.

 

[Skogsgurra]

You need quite fast opamps for 400 kHz and damn fast for 2 MHz.

Your LC network is a parallel one. They have high impedance at resonance.

Gunnar Englund

http://www.gke.org

--------------------------------------
Half full - Half empty? I don't mind. It's what in it that counts.

 

[mageta]

i am using the very excellent ad8137 diff op-amp. it should be fine at 2 MHz with a gain of 10.

the LC network should be impedance matched than? how do i calculate the source impedance of the LC network?

 

[Skogsgurra]

Usual thing. It involves the Q factor, lower Q, lower impedance. I am not going to do that work - that is something you have to do on your own.

Gunnar Englund

http://www.gke.org

--------------------------------------
Half full - Half empty? I don't mind. It's what in it that counts.

 

[mageta]

thanks so far and i don't expect you to, but could you give me a good-to-the-point reference that i can learn it from. i have found much material but most of it is not very practical.