Why a Dc Return path is required in mixer?

Hi all,

Can someone kindly answer my query? I want to know why we require dc return path in a mixer? (With reference please)

The dc return path can be via an inductor or directly grounding one end of the diode. Which one would be better and why?

Thank you all.

All mixer diodes, to operate correctly, must be installed in a circuit allowing a DC current to flow.
So, check your mixer schematic if this condition is met. If not, you should add a circuit element not affecting the RF performance but allowing the DC current to flow.

For instance, if one diode end is grounded, check if the other end is DC- grounded through an inductor. If it is not, install a choke to pass the DC current to ground.
You can wind a small coil of a thin wire to make such choke. If your signals are UHF or higher, use a piece of wire approximately quarter-wavelength long, or longer.
For lower bands you can wind your choke on a small ferrite core, or try using a commercial product.

Many designs of mixers, mainly at microwaves, only offer component selection for high frequency. Designers should know that DC circuit is as important as other parts.

Thank you for your reply. My circuit has a receiving antenna matching network for Rf (Stub) , diode, matching network for IF (stub) and then transmitting antenna. The diode is either in shunt or series. In series It has got and inductor parallel to it for dc return path. In shunt cathode end is grounded. Both designs work work. IF I want to dc bias my diode what should I do now, I tried using high inductance and a high impedance transmission line, but the dc bias had no effects at the output.

Thanks for reply again.

It is good that you can find the DC circuit for your mixer diodes closed.
Now, to introduce a DC bias, you have to "lift" one diode end, preferably above ground (it is a good idea not to touch or modify the "live" diode point where RF voltage is high).
Separating the grounded diode end is easily done by putting a small-size capacitor in the interrupted joint; then, I would suggest to use a resistor if you wish to use a small-current bias to the diode DC circuit. The resistor, to make sure it does not affect the RF operation of the diode, should have a substantially higher resistance than e.g. 50 Ohms which usually is the RF impedance. Say 500 Ohms or larger value, and the new capacitor 10 nF or more.
If you need to use more than 1-2 mA DC bias, replace the resistor with a choke. A choke allows for a larger current while it blocks the RF current.
It is a good idea to keep all modifications and added components small in size, to make sure they do not affect mixer function.

Thank you. I have another question, hope you will be kind enough to answer it. I am using a Schottky diode CD7621-000 of Skyworks for my mixer. Spice model for the diode is available in datasheet. When I connect the diode directly with the RF Signal of 10 dBm, my output is -5.092dBm, conversion loss of 14.57 which is too much. Can you guide in modelling a diode in Microwave Office. I guess there is really something wrong with the model.

Thanks

This time I am sorry. I prefer making real things, not modeling.
Usually, one diode is not the best option for a good mixer. It is difficult to match it in a wideband application; a matching circuit can be good only in a narrow band,

The best option is to use a diode pair and wideband baluns or RF transformers.

Good luck with your model- I prefer to make things on a breadboard and test them.

14 dB conversion loss for a single ended mixer sounds about right because these mixers tend to be somewhat inefficient. Try a balanced approach (single or double) to improve conversion efficiency. Although this requires a balun, the benefits are significant in terms of insertion loss, bandwidth, isolation and cancellation of intermod products.

Also, make sure you select the correct diode for your specific RF power requirement. Not sure what diode you have, but a +10 dBm input is considered moderately high, so you need at least a medium to high level diode. It is possible that if you are using a low level diode that your 14 dB conversion loss is partially caused by the fact that you are running the mixer into compression. You can check this by reducing the RF power to 0 dBm and measuring conversion loss, again. You may find the conversion loss to improve several dB. Remember: the diode determines the compression point!