OPA2675: How to drive a 100 Ohm Twisted pair with a Max gain and 100MHz bandwidth?

what frequency range is required?

on the minimum low F side?

Hi Michael,

Thank you for your reply. The bandwidth of around 1MHz to 100MHz is required.

Incidentally on your TINA file the V+ op amp inputs do not have a DC path for bias current, you need to add R's to ground for that at some ok value - also, normally you are driving the input balun with some source impedence?

Hi Michael,

Thank you for your notes.

I have revised the design to include a DC path for the V+ inputs. Additionally, I have set the recommended passive values according to the datasheet. However, I have noticed that the bandwidth is significantly lower than anticipated. Could you please advise if there might be any other factors I am overlooking?

Thank you

5850.LineDriver.TSC

Well 1st of all Mojtabe, we need to get a real balun model in here, the ADT4-6T looks promising, here is its output with a matched secondary 1:2 step up

Now you don't need to match across a balun and increasing those 200ohm terminations will reduce the insertion loss and shifts the response up in frequency

here is 499ohm on each side - pretty promising, flat at 1Mhz then peaking a bit at 100Mhz which will equalize the eventual amplfier rolloff - can push its gain up until it is rolled off .5dB at 100Mhz and this will equalize flat to the amp outputs

So the OPA2675 is a recent addition to a  long line of dual CFA for line drivers - this one has active off line control so aimed and PLC apps for in house wiring comm networks.

I did open the TINA reference design for the OPA2675, that modelled AC response is way low vs datasheet - that model is apparently way off, can't use it, 

Try the original DSL line driver with the 2011 model update, the OPA2677 - ok, this is looking pretty good, 22dB gain where I replaced your output balun with just a diff to single sensor, that whole output line model still needs to be added, but the flatness is only 0.7dB down by 100MHz. Work on your load model. There is a whole nuther world of output twisted pair active output impedance things out there too that was used everywhere in DSL, notBalunInputOPA2677.TSC there yet for this question 

here is that file in TINA V9 so you can open it, 

1145.BalunInputOPA2677.TSC

Hello Mojtaba,

Please let us know when you have additional questions.  I am going to review the OPA2675 model and let our team know about potential performance discrepancies. 

The OPA2677 model, as Michael proposed, should give you an understanding of what to expect in the interim.

Best,

Alec

Hi Michael,

Thank you for the provided design. It meets all the requirements.

Two questions: how did you extract the inductances of the balun (ADT4-6T). I am asking since I could not find them in the datasheet.

OPA2677 model has an extra pin which is not in its datasheet! is this an enable/power on pin since you connected it to Vcc?

Thank you

Thank you, Michael and Alec,

Your information has been incredibly helpful in resolving the issues and optimizing the circuit. I have one final question: Sometimes, two capacitors are needed in series with the transformer to block the DC, but TINA gives an irregular circuit error. Is there any way to fix this?

Thank you.