Tinypico Radio Part 10 - Double Balanced Mixers

[@rfdave3980]

Excellent lesson. I wish I had all that nice test equipment. Thank you.

[@na5y]

Heh - thank you. I have been accumulating for a while now ;)

[@NeedleBender785]

What kind of LO drive level do you use when using IF as an input.? I am trying it with 1.41V P-P and getting nothing but distortion, also what is the audio signal level presented at the IF port? I am able to get a decent signal but only if my LO is 400mV P-P and my IF in is 300mV P-P Thanks so much for great videos! KE8SPI

[@na5y]

Its been a while since I did this but the LO in both the ADEX and SBL were audio signals at about 0.7 V p-p. The RF input in both cases came from an si5351 and so was around 3.3V p-p. I blew a few of both mixers with too high an LO voltage though. Overall I got better results out of the ADEX-10. I hope this answers your question. To be honest I prefer the FST3253 based mixer. Easier to get a hold of and more robust. IMO of course. Thank you for the comment

[@kg6hum]

Is there a reason you didn't try using the FST3253 for the transmit mixer also? It would be interesting to see how it compares to the others.

[@na5y]

Excellent question Rob - I actually plan to get around to that! There's a little more involved with that style of mixer - you have to have an opamp network that extracts the 0, 90, 180, 270 components from the I and Q signals. Definitely coming up!

[@amirb715]

excellent video. very interesting stuff

[@na5y]

Thank you Amir - glad you enjoyed it!

[@MegaCadr]

On the ne602, is there an issue with not impedance matching on the input? Also, what is the turn ratios on t1/t2? Thanks!

[@na5y]

You might be right there. I do have a 47 ohm resistor in series on both inputs and the single ended input impedance is 1500ohms. That might be one of the reasons I have to drive the Audio so hard. I'll have to experiment. Other circuits such as this one www.arrl.org/files/file/QST/Homebrew%20Challenge/HBC%201%20Winner-KD1JV.pdf though don't bother with an impedance matching to pin 1 for the audio signal. Also turn ratio was 25:3 - If my math is correct that is approx 60:1 impedance down from 3000 ohms balanced output to 50. I tried a few different winding values from 25:5 - 25:3 and it didn't seem to make much difference Thank you for the comment!.

[@MegaCadr]

NA5Y yeah, I honestly don’t know. I have been playing with one to do a dsb transmitter, but I don’t have an SA.

[@na5y]

@@MegaCadr I wanted to include an SA602 version is this because I have a pile of them at the moment. They are pretty robust. I did blow up a pair of SBL-1s in the course of experimenting by driving the audio too hard into the IF pin.

[@amirb715]

w.r.t. the NE602 mixers a couple of things to consider for improving their performance: (1) drive the RF port differentially not single ended. This reduces those spurious tones. Also the IF should be taken differentially which you have already done so. Note that the RF level should not be high or you will get large spurious tones. it must be 10dB or so less than the LO drive. (2) I found that impedance matching at RF port is not that critical but at IF port it is more critical. The self reactance of your transformer at IF port should be large enough and at the same time the transformer and load should provide the right loading impedance for the mixer. (3) as mentioned by another viewer, play with the LO drive level to find the best conversion gain.

[@na5y]

Thank you for the great feedback Amir. I will definitely try differentially driving the RF port. The performance of the NE602 was pretty disappointing

[@na5y]

I was thinking more about your reply Amir. When you say the self reactance at the IF port are you talking about the reactance of the 25 turns on that T50-2 w.r.t the output impedance of pins 4 and 5 of the NE602? I have read something similar elsewhere (reactance should be 4x impedance) but was unable to find a reference. If thats the case then I'd have to use a different core. Thoughts?

[@amirb715]

@@na5y yes, that's what I meant. Basically only having the correct turns ratio for impedance transformation is not enough. The transformer self reactance should not heavily load the mixer. I dont have a reference for that 4x-5x value but it's a rule of thumb that is usually used when winding these type of transformers. It may not be possible to do it sometimes if the frequency is not high enough and the source impedance is too large (these probably apply to your case) then i think you should go for a self impedance that is at least not too low compared to the output impedance of the mixer.

[@m0rph171]

Thank you for sharing this experiment and your series on the phasing topology. The ne602 and friends are also very critical to LO drive level, something like 500 mV peak-to-peak should make it happy or if you can, try to find the sweet spot. I've had some headaches with this because I don't have a proper signal generator and fiddling with attenuation pads is slow, but when you get the LO amplitude correct there is a huge difference. Who knows might not help at all :) It would be amazing if you could test the VK3YE phasing transceiver mixer as presented here: kzbin.info/www/bejne/rambkIWopLimrac . That mixer would be very cheap to build with matched diodes. 73s OH6UAV :)

[@na5y]

Thank you for your comment m0rph. I'll have to check out your attached video - always interested in a challenge. I did attempt to build up my own discrete diode rings in the past but without much success. One of the things I do plan on trying is a Tayloe style transmit mixer using the FST3253 at the heart (like I had on the receive side). That should be an interesting comparison. 73s!