VHF Mixer

[dmowziz]

Hi

Trying to make a mixer for 88-108 MHz... IF is 10 MHz
Transistor is BFR92p. Conversion gain about 12dB

Built it but not like simulation. IF is not 10 MHz (maybe due to input loading?)

The output toroid has to be exactly 1.15 uH to get 12 dB gain.   1.2 uH gives loss.
Maybe that's also a problem

Please what's wrong with circuit? / If there's another circuit
U310 simulation is not decent


Thanks for any input

[Joel_Dunsmore]

to start, every capacitor needs a series inductor included (probably not a big deal at these frequencies, but I would use something like 1 nH), but more importantly, ever inductor needs a parasitic capacitance and that will depend entirely on it's construction.  You almost need to measure it, or use some SMT library. The part might have a Self Resonant Frequency (SRF) defined from which you can compute the parasitic (parallel) capacitance.  Below SRF, the effect is to cause the apparent inductance to be much larger than expected, typically by 50%.

[TimFox]

What do you mean by "IF is not 10 MHz"?
With an input RF = 100 MHz and an oscillator LO = 90 MHz, the output frequencies of any operating mixer will be IF = 10 MHz and 190 MHz, with LO and RF leakage at 90 and 100 MHz.
The output network needs to select 10 MHz and reject the other three frequencies.
Do you mean that the output tuning network is peaking at an unwanted frequency?

[dmowziz]

Thanksss

I measured the inductors with a VNA but didn't bother to model (simple RLC) it (atleast take the capacitance into account)



Yes, the output tuning network is peaking around 12 MHz.

It just occurred to me that I can atleast test the output network with my FG (2 outputs that run up to 25 MHz).
I'll try soon


Thanks for your input

[TimFox]

Luckily, your output network is peaking at a higher frequency than you want.
As Joel_Dunsmore pointed out, practical inductors have a substantial parasitic shunt capacitance and practical capacitors may have substantial parasitic series capacitance.
By "luckily", I mean that it should be straightforward to add a trimmer capacitor to lower the peak frequency of this network;  it is more difficult (but not impossible) to raise the peak frequency.

[fourfathom]

You can also increase the inductance of a toroidal inductor by squeezing the turns together. But probably not enough to shift the resonant frequency from 12 to 10 Mhz. For that matter this big a difference isn't likely due to mere parasitics. Check your calculations, core material, and winding count.

[szoftveres]

The resonance of C5 and L1 is indeed at 10MHz, but then isn't C3 in parallel with this LC tank?

[fourfathom]

The resonance of C5 and L1 is indeed at 10MHz, but then isn't C3 in parallel with this LC tank?

Yeah, the way I look at this is the L1, C5||C3  combo is resonant at 9.33 MHz, and the L2,C6 combo is series resonant at 1.96 MHz.  The inductive reactance of the L2, C6 combo does happen to pull the parallel tank circuit resonance up to near 10 MHz (probably by accident), but the output level is extremely low.  Not a useful circuit for a 10 MHz filter!