RF tutorials - Diode balanced signal mixers
ฝัง
- เผยแพร่เมื่อ 19 ม.ค. 2025
- #107 In this video I look at signal mixers implemented with diodes. First I look at what a mixer is supposed to do, and then I look at various constructive options, where complexity increases but also performance. Finally I try building the double balanced ring diode mixer and then perform some measurements to see just how it works.
References and further reading:
www.qsl.net/va...
www.robkalmeij...
Special Thanks to all my supporters on Patreon! Especially @afiskon and Ralf B.!
If you liked this video be sure to check out my other videos and you can also subscribe to be up to date with all the new ones!
If you want to support the creation of more and better videos please consider checking out: / feszelectronics
Great presentation. Very well explained with clear video and audio.
Thank you, very glad to see another RF-related video.
Speaking of filtering you should be mindful of a little problem with diode ring mixers. Filters have high input impedance outside of the passband. This means that unwanted signal will be reflected back to the mixer. For some mixers this is not a problem. (E.g. for mixers based on dual gate MOSFETs, if I'm not mistaken it's also possible to build a mixer with two FETs; interestingly an AM modulator you've shown before can be considered as an BJT mixer with a little high LO leakage.) However diode ring mixer is a symmetrical circuit. It will accept the reflected signal on IF port, mix it with LO and pass to RF port. Where it can (depending on the schematic and the frequency) be reflected back, once again be mixed with LO. For this reason you have to A) use diplexers instead of filters B) use a buffer between the mixer and the filter. Either will look like a nice ~50 Ohm load for the mixer and will not reflect anything back.
Now speaking of improvements of the circuit. I've never seen anyone using additional capacitors in the diode mixer. However you often see a 50 Ohm potentiometer as shown in this article eax.me/superhet-receiver/ It allows to get better balance in the circuit and better suppression of LO signal. To get an even better suppression of LO all diodes should be matched by forward bias voltage. The turns on the transformers should be wound with a trifilar winding of an enamel wire ~0.5 mm in diameter. The wires should be twisted tightly to form transmission lines with ~50 Ohm impedance. 7-10 turns on FT37-43 core will make a good mixer for HF.
Interestingly you can use a diode ring mixer as a DSB modulator. Just use IF as an low frequency input and RF as an output. With some attenuators and RF combiners you can add a carrier to DSB and get a very clean AM.
800 mVpp (into a 50 Ohm load) is about 2 dBm. This is a little low for LO. It should be about 7 dBm or 1.4 Vpp. In practice values between 4 dBm and 10 dBm will work OK. RF should be at least 8 dB lower than LO, which for 7 dBm LO is -1 dBm maximum or 560 mVpp. Otherwise RF starts to drive the mixer instead of LO which is not something you regularly want.
Hello Alex. By adding capacitors I meant something like fig 14 ( users.tpg.com.au/ldbutler/MixerTheory.htm ) or C3 and C1 in ( www.pan-tex.net/usr/r/receivers/ra02019.htm ); I saw similar approaches with single balanced mixers also - the idea would be to create a resonant filter between this capacitor and the double inductor. I guess, capacitors could also be added on the output side, again to form resonant filters.
The answer to reflected mixing products or harmonics of the LO is to use a diplexer on the IF output of the mixer. The diplexer will properly terminate the other frequencies. In the alternative, a 50 ohm resistive attenuator can force the termination of everything. Of course, you'll get some insertion loss, but you're going to amplify the IF signal anyway. If you use metal film resistors for the attenuator, you can minimize adding thermal (Johnson) noise.
This is a great description of how mixers work. I too use mixers in many projects, I must admit, that for higher frequencies (greater than 400mhz), I will be lazy and use the `mini circuits` mixers, Ade-1, sra-3500 etc. But for the lower frequencies below 50mhz, I much prefer to make my own where I can get better performance by very careful manual selection of each diodes parameters. At the moment I am building a 40 meter ssb exciter which uses two double balanced mixers. Great work! You have another subscriber. Chris, UK.
Another great video. It's a great format you are using, keep up the good work Fesz.
Great explanation as usual, keep up the excellent work!
Keep making videos. Your content is GOLD!!
Nice talk about mixers. I use mixers a lot for my projects.
Nice! What sort of projects do you work on?
@@FesZElectronics I'm making a hf exicter taking soundcard output and shifting signal up to hf bands
.
This is an excellent video! Showing the practical as well as the theory really helps this layman.
Good job , keep doing videos like that
well done - great content and to the point
Exceptional video mixing theory and practice as you used to do. I will spend hours to replicate and understand as far as possible! A question about the level of power needed: Some radio ham explains the input signal could be very weak if the local oscillator is at +7dB. Will try to experiment starting from your schematic. Again a big thanks for sharing.
Great video !
i need a mixer for a 27mhz carrier and audio signals, how would you design the coils at my spec ? Would the same diodes work ? Also ... how would the coils change (their values)?
Great video! Is this LTSpice you're using?
Hi. Great content. I have a small question. Could we use a double balanced ring mixer with a square wave signal and a sine wave for the carrier? I know that we have odd harmonics for the square wave but how would a mixer like this react to these two inputs?
Found you! Great channel
With the double blanced mixer, why are the ports of the IF and LO sometimes swaped?
Is there a rule on where to best apply what signal?
The left and right ports are obviously symmetrical, however i would assume that the bottom port is best connected to an LO that provides a squarewave?
In the circuit at time 12’, if you swap the feed points of the 10 KHz audio and the 100KHz carrier, you would have stronger outputs (less conversion loss) while not needing the larger transformer with 100 mH inductances. Same thing applies to a demodulation mixer: do not couple the low frequency input/output to the mixer via magnetic coupling.
Really good, interesting how this mixer (just like the ade-1ask) works so well and is used in many HF radio receivers with antenna signal at uVolts and LO at 200mv. Seems amazing that with 1000x amplitude difference it still works! Do you know if their is anything much better than the ade-1ask?
Just great... Keep going sir..
jesus man, you rock !
I've been looking for info on the ADE-1ASK+ as the ADE-1ASK is not readily available, but can't find anything that indicates what the + version change is. If anyone knows, please let me know!
The + versions are RoHS (Reduction of Hazardous Substances) compliant. Basically, no lead (and certain other substances) in the solder.
So nice
Can anybody explain what differences I will need to consider when designing a mixing circuit for LF vs VHF vs microwave? Thanks.
This cleared up a lot of the theory for me. Thank you. What is the name of the cable connectors on the veroboard(used to connect the signal generator and the output to the circuit)?
Those are "BNC" type connectors.
thank you. I am currently building a frequency domain reflectometer and a mixer is essentially one of the components as you probably know. What do you think might be the best way to connect sma or bnc directional couplers to such a mixer and other components involved? What would the interconnection be? Simple jumper wires? A coax? What do you think?
I'm thinking trifilar winding on the torroids would give more consistent results.
I'm trying to understand the mystery of mixers.
It feels like I am getting closer with every video I watch.
I haven't achieved the Ah-ha moment where all the pieces fall into place, but I know I am close.
I think I need a few more sleep cycles for my subconscious to sort it all out.
Merci for this video.
Very good
Is it possible to balance them with potentiometer ?
not really. mismatches of inductance, capacitance cannot be corrected with resistance tweaks.
Excellent!
👍👍👍
Unfortunately there are no subtitles. If there were English subtitles, I could understand it by translating.
you dont shot from what point are graph
Make RF transformers from variocouplers and get perfect balance.
Respect from Pakistan.
Thank you for your videos, but please add translation to your videos. Automatic translation at least