HTML-код
- Опубликовано: 29 июн 2024
- This video describes how a classic double-balanced diode-ring mixer operates. Very basic mixer theory is quickly reviewed, which describes how the sum and difference of the LO (local oscillator) and RF (Radio Frequency) inputs are generated at the IF (intermediate frequency) output. It is also noted that the sum and differences of the harmonics of the LO and RF are also present at the IF output. Math waveforms on the oscilloscope are used to illustrate the operation of the mixer, and the actual waveforms from the mixer are compared to the math waveforms and the differences are discussed. A detailed description of the operation of the mixer is presented, including the switching action of the diodes. Finally, the frequency components are are expected from the mixer are shown on the spectrum analyzer. For more information about how diodes are used as switches, please see the video at this link:
• #82: How to use a Diod...
Also, for another tutorial on mixers and their use in frequency conversion and translation, check out this video:
• #83: Basics of RF Mixe...
And, here's a video that shows how I wound the trifilar transformers used in this circuit:
• #166: How to wind a tr...
And, here's a video that shows how I wound the trifilar transformers used in this circuit:
• #166: How to wind a tr...
A copy of the notes in this video can be found here:
www.qsl.net/w/w2aew//youtube/D... 
Наука
Man, can I just say you are one hell of a good explainer of RF stuff
W2AEW, aka Alan, you're the KING of RF electronics videos. You hit the stuff I never totally understood, and in the best way. Rock on man
Fantastic video, thank you! Literally almost all the information on the internet is calling a 4 quadrant multiplier a ring modulator (It's not!!). This is the only video I've seen that gets it right.
Thank you for this awesome presentation. As I was following the "Scope Overview" section where you list the 2 frequencies (7 MHz & 10 MHz) being combined, in my head I'm already calculating the resultant frequencies as 3 MHz & 17 MHz. You really threw me for loop at 2:08. I did a "What?!?" double-take & went back to hear that again when you stated that the upper frequency would be "13MHz". I started doubting my own RF & LO knowledge. Thankfully, the rest of the time you referred to the upper one as 17MHz. [You may wish to consider posting a notice of this misstatement.] Thank you.
Very helpful and well explained, Alan. Thanks!
Thanks for the explanations! There are so much specific details in RF electronics..
Your tutorials are the best I've ever seen on RUclips! 73 de DL1WW!
This is amazing! Who on Earth could have possibly disliked this video?
The ‘never Tesla’ crowd ?
You did a great work, making sketches and explaining how the mixer works. In these 10min I have learned more, then from reading books for 3 hours.
Thanks!
Another amazing video. Thank you Alan. If you are open to requests, I would love to see videos about mixers that use as few magnetics as possible with the goal of reducing PCB size and cost.
A trifilar transformer like this for say 2-5GHz is the size of a 1210 capacitor if you wind it yourself. It’s even smaller - a small grain of rice - if you buy it and it costs pennies. Kinda hard to beat with discrete parts. The only way to beat it is with transistors on a chip: those are usually “free” compared to anything discrete. So if you really want low cost, use a dedicated mixer IC, or make an ASIC for your application that will replace other discretes with free transistors :)
When pricing discrete parts, the cost is in raw materials and processing of each unit: a discrete run-of-the-mill transistor costs about as much as a discrete inductor or transistor of a similar size and volume. When costs matter that much, you’re talking hundreds of thousands or millions of units. At such scales, small magnetics cost about as much as small anything else, to a first order. Even the costs of inserting the component start to matter at such scales.
For anything made at hobbyist scale, the cost of labor wipes everything else. PCB and other stuff is essentially free, and it costs more to ponder about it than to just put it on the board and move on if it does the job to spec you need :k)
I love how detailed and in depth your rf videos are, your'e the only one who makes decent videos about this its awesome! keep it up.
I actually followed this all the way through.....a testament to your presentation. Thanks!
This is the best explanation about mixers I found over the YT! Thanks!!!
I feel like it's worth mentioning that the non-linearity of the diodes is what makes the "cos(a+b) + cos(a-b)" (addition and subtraction of the input frequencies) components appear in the output signal.
Otherwise, what we've created is just the sum of two signals sin(a) and sin(b), which can be achieved without the use of diodes or switching at all.
Thanks to you w2aew, RF is becoming less voodoo magic to me. Thanks :D
The explanation on the diode circuit diagram is amazing and in-depth. Classroom courses pale in comparison.
The regular electronics hobbyist could not justify the expense of owning such a nice mdo piece of gear (around $14K tag price) I would like to thank you, Alan, for your generosity on producing videos like this one. The 50K frames in this video are priceless for many of your followers.
+Frank Acosta I do realize that I often use equipment that is outside the reach of the hobbyist. I am very lucky to be working for Tektronix and have access to this fantastic equipment. I hope that the use of this equipment serves to help viewers understand the topics and concepts being presented.
+w2aew It does....I am a design engineer with a big mechanical background and a "working" background on the electrical side of things. These videos have really expanded my scope of understanding. Especially in the industry I work in. It opens my eyes to a lot of the things my companies technology does.....metal detection.
I was actually thinking: "that scope would be a waste in my possession because I wouldn't know how to use half it's features"
Thank you for such a well thought through explanation. I'm still trying to wrap my head around AC waveforms and the math, but you really helped to visualize it.
Super. I got back to working on electronics after looking at one of your "back to basics" video, and I am still an audio frequency guy but I love learning about RF which is a domain I touched really very lightly during my studies. Thanks for these fantastic videos.
Lots of info, learned something new today. I'm sure it will be helpful in the future.
You're doing Gods work.
I'm subscribed to the synth-diy mailinglist and there's so much material that's common between these two areas. Besides the diode nonlinearity, which in audio-synth territory gets kinda fixed by an ideal diode circuit (fancy balanced opamps), it's basically a ring modulator!
Crystal clear explanation as always, much appreciated for uploading!
Finally, I understood this circuit!
Thanks W2AEW!
These videos are super helpful with the COVID19 quarantine preventing us from going into the lab (and with professors who aren't capable of making videos) - thanks a lot! You're a life saver!
By far the best, clearest explanation of how the DBM works. I kinda' understood them, but now I feel as I DO understand them. THANK YOU!! You are the man!!!Ron K7RJ
Thank you for the nice comments! I hope you find some of my other videos just as helpful. Be sure to check out the video index .pdf file that is linked on the main channel page (lower right corner of the banner graphic).
Did they name a school after you yet?
No, not yet.
Thanks for the nice and simple explanation, i don't do RF stuff but i always wondered what these mixers are that i see everywhere.
Exactly what I was looking for, tyvm for posting.
Terrific video. It must take you a long time to get all these diagrams and oscilloscope setups working, but they really make things clear. Thanks.
Sometimes the preparation work for producing a video like this can be several hours, or even a few days of spare-time tinkering.
Exceptional! Truly enjoyed it.
Thanks for this. Love it. I never really understood diode mixers. Now I do!
wow Alan, that was awesome. I knew how mixing worked, but to see the ways you did this it really made me think.. Also -- that was really cool to see a custom math expression on a scope, I've never seen custom expressions before, only built-in
An amazing explanation! Than you very much and keep up the good work!
this explanation is very nicely done
Your videos have definitely saved me this term. With COVID19 shutting down the labs and everything's moved online, it's been really hard for me to learn just from my professor's lectures and no lab work. I was really struggling til I found your videos. Thanks so much!
Very happy to hear that!
It was great to visualize the results from multiplying the cosine waves at different frequencies. Thx
Great explanation, thanks Alan...
Bud...
Thank you sir, couldnt be explained better.
Very well explained! Many Thanks !!!
Excellent explanation, thank you very much.
Thanks so much. I came here from your other mixer video. I've got a good handle on this now, at least for this type of mixer. I'm more of a digital domain, embedded guy but am getting the urge to try some RF projects.
very good explained. Thank you !!!
Jim, AC8NT my Extra Class instructor showed this and three more of your videos tonight. I am one of those is having/ has been having a difficult time understanding electronic theory. I am struggling. Now that the former Advanced and former Extra class material is combined I will be attempting to pass Extra in 5 weeks. This is my 2nd time in Jim's class, last time was 18 months ago, and I am still having difficulty grasping the ET concept. Thankful for guys like you and Jim though. Terry KB8AMZ
Great stuff Alan.
Good refresher. Thank you.
Thank you very much for detal video about diode ring mixer!
Learn a lot from your video, thanks !
Thank you for great explanation.
Hi Alan! I have recently found your videos through making my own. I Just started making videos, and guys like you are a great inspiration to keep going. We have a lot in common I see. I have a scope collecting problem "ha ha" ....... Well, test gear collecting in general.... And so on.
Great Videos! And keep it up. My subscription list is still small, but your now in it!
Thanks! I've subscribed to your channel too.
Learned a lot! Thanks!!
excellent, Alan. Thanks
Thanks very much for an informative video. Sometimes it takes one key piece of information to turn the light on in the ole noggin and you did it. I have been wondering for years just WHERE the information in an AM signal (sidebands) was located with respect to the overall signal. Watching it on a spectrum analyzer is one thing, but to see in on a scope trace is another. You answered a question that has been bothering me for years. I know that you can see the relation in the math, and that "stuff" may be found in other quadrants of the graphs, but your explanation finally showed me HOW it is related in the time domain. Great stuff. Now if I can translate this to a better understanding of SSB. . . . .
Very interesting video Alan !! I noticed that my IFR-1000S has similar circuits that seem to mimic your Diode Ring Mixer video !! 73's Gary Grove
Wow..great explained...!!!
Thank you for nice explanation
The ‘Math Expression’ in Tek MDO3104 is very neat; a scope married to an Excel spreadsheet (or to the more highfalutin: MATLAB or Mathematica). Cheers, Mark
this channel is Amazing
Thank you. Be sure to check out the index of all of my videos:
www.qsl.net/w2aew//youtube/W2AEW_video_index.pdf
w2aew thank you
2:15 Dave's ruler, that's nice!
Excellent!
I tried building one of these to make a 38kHz DSB-SC subcarrier for my FM radio transmitter. It worked, but my whole project failed when it came to trying to make a nice 19kHz pilot tone (sinewave) while also having a perfectly in-phase 38kHz sinewave to modulate the L-R audio. It gets so complicated! Wish I'd chosen electronics when I was college age.
very helpful
One of the things that is a real grey area for me is winding coils and transformers for a circuit like this. Is there are good introductory material on choosing the right ferrite material, geometry and so forth? Perhaps you could discuss this topic in a new video, if appropriate?
Here is one of Alan's videos on winding the coils for this: ruclips.net/video/a8ViWS61hsU/видео.html
awesome and subscribed
it is best video to well know RF
That 'scope is thirteen & a half grand (€) here in Ireland 😔☹
But at least I am blessed to have your teaching & demonstrations for free here on the Tube... for that I sincerely thank you !! 😎👍☘🍺
I am very lucky to work for Tektronix - so that helps!
Fantastic video! - Any chance you can do a similar video on the JFET ring mixer?
Another perfect video. It would be nice to talk a little about the transformer. How the number of turns was found? Any math involved or just trial and error? Thank you!
I have re designed the front end board in my old Yaesu FT-101 by Replacing the RF amplifier FET with a 3N211 or a 40673 which I have and using a MC1495 as a 1st recieve mixer as well as the 2nd recieve mixer to get rid of some of the birdies found in the rigs pass band. After what you tell me about dynamic range and higher intercept points I am now concidering using either a shottkey diode DBM or DDBM if I have enough LO drive.
Ted VE6AMR
Thank you so much for this video. I tried to find good explanation of how could diodes make the mixing, but never found any relevant one.
Thank you Alan!
Great.. sir please make a video on working of bfo. Also a video on total working of bitx20 by OM ashar farhan with tx and rx procedure. Thank you.
Question about theory: The diode ring mixer shown here seems very different from the Gilbert Cell mixer described elsewhere. The diode ring seems to “chop” one frequency with another frequency. That doesn’t seem like “multiplication”, but more of a harmonic creation mechanism. The undesired harmonics are eliminated with filtering, but with losses from the energy spent in suppressed harmonics. The Gilbert Cell seems like a true multiplication, where the product of 2 sinusoids is produced accurately, without creating a lot of extraneous harmonics. It is therefore less noisy. Is this a roughly true comparison?
Basically yes. The diode ring can be though of as a multiplication of one signal by an alternating +1 and 0, or +1 and -1, as opposed to a continuous multiplication. Even the Gilbert Cell can (and is often) be driven into limiting, thus giving a similar result to the way the diode ring mixer works.
Thank you best 73s
Nice!
That was a really good explanation. Thank you very much for this. The only (small) modification I would have made to this video would be to change the math to have your logic driven by the transition level of the diodes (assume .6v for the 1N5711) only after you show the IF signal at the mixer output. This would help demonstrate that the zero portions are due to diode transitions. I saw that you used the LO to demonstrate this. But, it would have been good to see that mathematical waveform approach the measured wave. Also, it seems that the measured signal lags behind the math waveform (see 8:52). But, it's hard to tell without taking the difference of the mathematical waveform and the measured signal. Can you confirm that this is happening and can you explain why this might be happening?
Thanks, again. Really, it was a great video. I could not have done better myself.
John Rinehart Thanks for the nice comments. I haven't taken that close a look at those small phase shifts, although it looks to me that the measured waveform is slightly leading the math waveform. I suspect that it is at least partially due to the difference in phase shift through the transformers (and I was measuring the signals on the primaries.
Awesome
please explain and make little circuit of how Tuning Meter Led indicator circuit works in old Radio ?
very informative video. Thank you Alan. Did the amperage in this system rising when reducing the frequency ? aka mix 100khz with 1 vpp and 50 khz 1 vpp equals 150 khz and 50 khz with twice ampere ?
I have been thinking about buying an entry level scope for hobby HAM work and basic learning, similar to what you are showing here. I am looking towards the Rigol brand but for giggles I thought "How much does the scope Alan has here, cost?" So I googled it.....and that led to about 90 seconds of nonstop laughter. :-D
You are a professional that does this for a living, and it shows through the cost of your test equipment. My hat is off to you my good man, you have IMPRESSIVE kit. :-) - N0BPS
PS now watching some of your other videos to understand bandwidth to see if the Rigol 1054z is a worthwhile scope for HAM radio use and general learning......there is also the 100Mhz hack for it...not sure if that is important or not.
+Brian Streufert the Rigol DS1054Z is generally a no brainer as a beginners scope. You'll never regret buying it.
However, it is in no way comparabale to Teks "middle" class scopes, they are simply fantastic instruments. And the MDO3000 series starts at just 3,5k$ ;-).
I could not agree more... the Rigol is the best per buck out there. And, they don't cost much more than a used TEK with fewer features.
I looked around and considered used scopes among other things. I ended up with a Siglent SDS1102CML. I'm still learning to use it, but it seems to fit the bill so far.
I might add, digital scopes are wonder, and you should have one in your future, however if you can pick up a 20 MHz old analog scope for less than $50, you should consider it. They are dirt simple to use and will give you some experience using a scope before you plug onto the more confusing (but much more powerful) world of digital scopes. I have a cheapo 20 MHz scope that I love for its simplicity. I won't be without my 200 MHz giga sample huge memory USB run data box, but generally the ole trusty analog job does just fine.
Before you guys with your fancy dancy data acquisition wonders flame me, remember this guy is where we all were and just need practice going at waveforms in real time. Remember when we did that just for fun, and how much we learned about electronics before the computers got in the way.
Howdy.
Mesmerizing ...
Regards.
very helpful video ....alan. can u please simulate the same using any software.
thanks
perfect!
Great informational video! Your videos help a great deal when I'm trying to work out something of my own. Quick question: @ 6:42 when you are explaining the negative state of the diode mixer ring, I'm assuming the input is still LO, but in the video and in your copy of the notes, it's written as IF.
Yes, this is a mistake in my notes. Thank you for paying attention!
I've edited the PDF to correct it and uploaded it.
Ah wow! Thank you sir, appreciate it.
What kind of diodes can we use for making these double-balanced diode-ring mixers? And what determines the frequency window of operation of these mixers, if any? Great video as always. Thanks!
High speed schottky diodes are the most common. The design of the transformers (number of turns, core material, etc.) will largely determine the operating frequency range.
One aspect I have not seen addressed (yet) is drive level. The signal amplitude at the diodes I think should be at least 1.5 v, since 2 x silicon diode voltage drops of 0.7 v would be at least 1.4 volts to turn them on. But I do not know what the transformer ratio is, so I cannot guess at the LO amplitude. I also would like to play with IF filters, but do not know the proper drive level. My initial thought would be to drive a filter (mechanical, crystal, or piezo) with a sig gen at the filter's center frequency, slowly increasing the amplitude until I see some saturation evidenced by output ceasing to increase. At some point, I would expect damage from excessive drive, so I would like to stay below that. Radiofun232 comes close to addressing this with some of his lash-ups, but my questions remain unanswered.
I do address this topic in this video: ruclips.net/video/GvadQpkZ8l0/видео.html
hello..can you do video about antiparallel diodes mixer? that works on half frequency of local oscillator
Thank you for this great tutorial. But I have another question: How works a Polyakov mixer?
Polyakov mixer in his DC receivers works with large signal from LO (with frequency two times lower than the signal frequency), which turns on the diodes twice each cycle (one diode for +, other diode for -) and the RF signal, which is small, is sent to the output at these times. We have at the output the difference frequency between 2 times LO frequency and RF frequency.
Thanks for the video! I asked about a mixer last night and you provided. Afrotechmods check out the beach 40. Only one transformer.
Thanks for clearing concepts. Can you please say is any low frequency (kHz) QPSK modulator available or not?
is the leakage you pointed out in the spectrum the same reason we see "LO feedthrough" on spectrum analysers? excellent video!
Similar, yes.
nice to see dave jones EEVBLOG uRuler in the shot there ;)
Thanks for this tutorial! Why, however, does the LO alone set the direction of the current in your analysis? Why doesn't the RF signal affect the current? Do we need to assume that the LO's amplitude is much greater than the RF?
Yes, that is typically the case. When you look up commercial mixer specs, you see Level-7 mixers, Level-10, Level-13, etc. This refers to the power of the LO signal (7dBm, 10dBm, etc.).
Congratulations great video,
wanted to know if when you were measuring the signals used some component that is not described in the list
I could not reach the measured results
The parts I used are listed on the first page of the notes.
www.qsl.net/w/w2aew//youtube/Diode_ring_mixer_operation.pdf
thanks
***** At 5:15 the circuit on the right doesn't have an LO input but does have an IF input and an IF output. This is a mistake? or is there something I don't quite understand?
Good catch! Of course, the port on the left of the right-hand circuit should also say LO.
Question: I've built a diode mixer for AM modulation and compared it with the Gilbert Cell modulator. Does the diode version always suppress the carrier? or can it be adjusted produce a traditional AM signal like the Gilbert Cell?
You can sort-of do this by adding a DC offset to the baseband signal applied to the IF input.
Hey! do you have a video where you explain about the harmonic terminations to avoid reflections? thanks!
I'm sorry, I don't...
Hello Alan, great video as usual! But I have some questions to it:
1. You have your LO amplitude set to 400 mV is it enough to switch on 2 diodes? When I see a diode I automatically think of 0.7 V voltage drop. I think it's hard wired in my brain after school :)
2. I understand that the RF signal amplitude must be below voltage that turns the diodes on is that right? So that only LO controls the diodes. Did I get it right?
3. Is the LO amplitude picked so low intentionally so that there is not clamping from top and bottom?
Hello Marek. I'm using 1N5711 low power schottky diodes in the ring. The Vf for these schottky diodes is about 1/2 of that a regular junction diode. The LO is used to modulate the diodes. The RF level is generally kept well below this to minimize distortion. Best (lowest) conversion loss occurs when the diodes are fully switched on/off, but the mixing happens even if this isn't the case. As long as the LO and RF are combined in a non-linear manner, the mixing happens - it's just that the conversion loss will be higher. When the LO power is high enough to fully switch the diodes, there is a fair amount of harmonic distortion visible on the LO drive waveform. For this video, I didn't want to get into that level of detail, so I kept the LO drive lower to keep the video length reasonable. Thanks for the good questions.
Marek The mixed two signals on diode should be less than 0.6v( for silicon diode) to have less current. That makes the load to both transformer light, hence less signal distortion to signal source. Also, less power consumption by the circuit. Diode has non liner v-I curve, in all range from 0.x volt to couple volt. That how the diode makes the added voltage of the two sources to multiplied current, measured by the resistor in serial . The diode ring also called ‘Gilbert cell’ question from stackoverflow, electronics.stackexchange.com/questions/307776/why-do-diode-rings-multiply
Excellent explanation of the operation of a mixer.
Regarding to it, I have a question to you.
I need to compare two input standard signals of 10MHz. I could use a good interval counter to do it, but I need to have a better resolution.
Then we decided to use a frequency mixer in order to obtain a frequency difference of some Hz ( 300-500Hz) at the output.
My question is how to calculate the construction of the two baluns (how many turns and the diameter of the wires) so they can work on that frequency with the same circuit that you show in the video.
daniel perez General practice is to make the inductive reactance of the winding to be 4 or 5 (or more) times the impedance of the circuits driving it. For 50 ohm applications (typical), the windings should results in a reactance of 200-250 ohms or so. The number of turns can be determine by using one of many online calculators like the one shown here:
toroids.info/
***** many thanks for your quick response! It is very useful to us and we will build our mixer immediately in accordance with the data that you have provided us.:)
I find mixers interesting. I realize that this is an entry-level, rudimentary mixer, but...
Could this general type of mixer be biased into either the positive or negative, (and then capacitively/AC coupled) to reduce the crossover distortion causing the odd harmonics?
This is a good video, as most of yours are. Thank you. (Generally, I have been so impressed by Amateur radio licensees/operators knowledge of RF electronics that it prompted me to get my General.)
Adding DC bias can negatively effect the balance of the mixer, and thus negatively impact the LO and IF/RF rejection. Sometimes a *small* amount of DC is added as an adjustment to improve the balance.
@@w2aew , thanks for the reply.
Hi , Thank you very much for the nice video and your information, but I was confused and about the value of coil , you made mixer with 10 turns of the coil with core T37-6 , and the other with 10 turns FT37-43 , which are very different in T37-6 =0.3 UH FT37-43= 35 UH the value of coils in the mixer It has no effect ? I am also happy if you have MCL SBL-1 mixer , measure the values of bases 1 and 2 or 7and 8 thank you in advance .
Thanks for the video! I am wondering, at 4:42 , could the 'LO' may well be a square wave? (since the mathematical expression simply detects the zero crossing) If so, the output doesn't seem to look so bad, why do we care so much about getting a low distortion sinusoidal oscillator? Or is that more relevant in circuits with 'better' mixers?
It does depend on the mixer design, but in this case, a squarewave LO would work