The Designer's Guide Community Forum

The Designer's Guide Community Forum
https://designers-guide.org/forum/YaBB.pl
Design >> RF Design >> passive mixer design
https://designers-guide.org/forum/YaBB.pl?num=1284799984

Message started by T3N on Sep 18^th, 2010, 1:53am

Title: passive mixer design
Post by T3N on Sep 18^th, 2010, 1:53am

I am designing a simple double-balanced passive CMOS mixer. When I ran the transient simulation and measured the IF voltage, I noticed that the IF voltage is very small. It is in the range of nV. The RF and LO voltages are in mV range. It means that my circuit has a hug insertion loss.
Do I need input and out-matching to simulate a mixer? Currently I have no input and output matching.

Thank all
:)

Title: Re: passive mixer design
Post by vp1953 on Sep 18^th, 2010, 8:47am

Hi T3N,

passive mixers do not have as much gain as active mixers, so output in general will be smaller, though it looks really small in your case. Why LO is so small, it generally is at least several 100mV. What is the input impedance looking into the mixer and the output impedance of the LNA. If possible, please post a schematic of your mixer.

Your mixer has any filters at the output, if yes the downconverted signal is within the bandwidth of the filter?

Title: Re: passive mixer design
Post by T3N on Sep 19^th, 2010, 6:22pm

I don't have any filter at the output. I also don't have any input or output matching. Right now, I am simulating a stand alone mixer.

Title: Re: passive mixer design
Post by vp1953 on Sep 20^th, 2010, 6:47pm

Hi T3n,

Can you try increasing the LO to maybe 1V in amplitude? I think in your current setup, none of the fets are turning on really.

Title: Re: passive mixer design
Post by T3N on Sep 22^nd, 2010, 2:19am

Hi vp1953,

Thank a lot for your help. Now I bias the gate of all the transistors at 1V and the source at 0.5V. The result is much more reasonable. I notice that if I connect the IF to a 50ohms port, and measure the IF at the output of the 50ohms port, the IF voltage is very low. But if I measure the the differential IF nets ( as seen in the schematic), the output is exactly what I expect.
Is it because of the 50ohms load of the port? Do I need to add a buffer?

Title: Re: passive mixer design
Post by vp1953 on Sep 22^nd, 2010, 5:03pm

Hi T3N,

50 ohms is like a short load for your mixer. Try using a larger port resistance like 1k or 10k or higher.

Title: Re: passive mixer design
Post by T3N on Sep 22^nd, 2010, 6:38pm

But for measurement, aren't we supposed to design a mixer for 50ohms load?

Title: Re: passive mixer design
Post by vp1953 on Sep 22^nd, 2010, 10:17pm

Hi T3n,

Well not necessarily, it does not always have to be 50ohms. But if you have 50 ohm load, then obviously you need to increase the drive on those fets - use larger sizes, larger biases and larger amplitude for LO. I think there is not enough drive current that is why the voltage goes low for 50ohms.