Filtering out the carrier with a double balanced mixer

[G7LAS]

Hi all (my first post.. what an excellent forum!),

I've built a small oscillator which operates at around 17MHz and I plug this into my HF radio and I get a good strong carrier. Great start.

Then I've made a double balanced mixer like this : http://www.radio-electronics.com/inf...nced-mixer.gif

I've made the trifilar transformers myself.

I apply the oscillator output to the Lo port, an antenna to the RF port and the HF receiver to the IF port. This shifts the stations frequencies received by the receiver as to be expected.

I'm *trying* to get DSB out of this, but the mixer doesn't seem to suppress the carrier. I don't understand why. I apply the oscillator to the Lo port, an AF source to the RF port and the receiver to the IF port. I get great AM out of that (it plays my music source lovely!) - but the carrier is still present.

I understood a "double balanced mixer" should suppress the carrier as it should filter out the two source frequencies.

Any ideas appreciated.

Many thanks!

Rob

[SM5JAB]

With any unbalance in the circuit you will have some carrier. Normally one needs to trim for lowest carrier in some way. I did that in my HW-8 mod: http://www.da.isy.liu.se/~mj/HAM/QRP/index.html which links to KL7R:s site. Or google for "HW8 Double Sideband Mod" and you will probably find it. The carrier went down some 20 dB re the DSB peaks for me. With more care even more should be possible. There are also IC:s for this (was it MC1496?).

/Micke

[G7LAS]

thanks for the reply Micke...

I'm using schottky diodes for this. any idea how well made the transformers need to be? mine are certainly similar, but definitely not identical. could this be contributing to my problem?

what kind of tolerances are we talking about here?

thanks again

[SM5JAB]

In my case I think that the trifilar winding did the trick, that retains some symmetry. But even with the trifilar winding it was necessary to adjust the balance with the trimpot for lowest carrier. I imagine there are all sorts of capacitive and inductive couplings going on which can be tricky to foresee. The trimpot did the trick though. I watched the result on an HP spectrum analyzer. The vestiges of the carrier is still clearly audible on another receiver in SSB mode. With no modulation perhaps adjusting for lowest S-meter reading would do?

/Micke