Completely agree. It's sad that I'm having to supplement several semesters of EE education with Khan Academy and UKposts (particularly w2aew), and I wasn't a bad student at all. I started college at the turn of the century, but still have to reinforce things like this all the time.

Well said Walid

My life doesn't always make sense, but these videos do!

They don't know this much, How could they teach like this?

@@Aemilindore That's a little unfair! The guys who teach it at Uni's have PhD's. They KNOW the theory very well, but what many CAN'T do is teach the practical aspects. Most of them have never practised as professional hands-on RF design engineers like Alan has.

@@hectorpascal Writing some equations on a board or presenting them on slides teach nothing to students. Professors do science. They must teach practicals. And frankly they want to. They just do not have the knowledge that Alan has in this field. Now, professors may know things about wireless that Alan doesn't. But about IQ, about modulation, Its Alan. A PhD is a qualification that says you know a very specific topic. Not an entire field. Alan's job is to engineer things. But this time, he takes a job of a professor. To teach. And about wireless communication, Alan does it the best. Imagine learning this from a university professor. It will be a series of trigonometry equations. HAHHA.

Thank you Lou - much appreciated - especially from a fellow AE (except I'm with Tektronix). Are you any relation to John K2OX (also Keysight AE)?

Ahhh...my brother John Eckert, no relation but we joked about it at S800 training every year. I retired in 2015 and enjoy hamming and staying healthy. TEK is lucky to have you!

I do have a VNA here. That is a good topic, I'll add it to my list.

Thanks Chris!

Thank you Scott. As for the math.. ...gee, I have to leave *something* for the viewer to do!

There is a fairly common technique that is found in low-cost SDR radios that uses a 4x clock and some dividers to create a pair of quadrature clocks. See an example in this video: ukposts.info/have/v-deo/endhpqOEfqaorpc.html

Using the python GUI Interfaces

I can understand why a SDR platform could be called a universal radio - because once you have the RF signal represented as IQ, you know everything you need to know to process that signal - at that point, it's just math....

Depends on the design of the mixers. Also can depend on whether the 20MHz squarewaves have a DC offset on them, or are they balanced around ground.

@@w2aew the square waves do have a DC offset on them. they are balanced though, also I'm generating the square waves from an fpga and its psuedo random.

@@slothluvschunk2301 If they have a DC offset (i.e. they run from 0V to xxV), then they are not balanced around ground.

@@w2aew thanks so much for the replies I really appreciate it. I meant that the dc offsets are equal to each other on both the I and Q lines. I keep thinking this is some sort of symbol clock leakage or possibly the bit stream is not psuedo random but repeating some pattern.

DC offset was about 50-100mV out of the 400-500mV IQ amplitude. Phase shift was 10-20 degrees off of 90.

Yes, if you apply an AM signal to a quadrature demodulator, both the I and Q output will vary in amplitude. Phase between the I and Q outputs will be 90 degrees, due to the 90 degree phase of the LOs in the demodulator. I hand soldered the ADE-1ASK onto a small SMT breakout board from QRPme (qrpme.com/?p=product&id=MEP). Very easy to do - just solder one leg at a time. You're going to need a couple of generators - one at least for the LO input, and possibly another two for the I and Q inputs respectively.

Thank you. I was thinking of using the Minicircuits ZFMIQ-10M quadrature mixer to obtain the 10 mHz quadrature signal and then using the ADE-1ASK mixers, one on the I output of the ZFMIQ-10M and the other on the Q output of the ZFMIQ-10M. I would use a Wavetek (older) signal generator as the 10 mHz input to the LO of the ZFMIQ-10M. I would apply the input signal (RF input) of a variable signal generator to each of the ADE-1ASK s. I did not quite understand how to get the I and the Q vector lines that you obtained on your oscilloscope and how the vector lines rotate with a FM input. I think that I need to study a little more to understand about demodulating SSB signals. What inspired me to learn about IQ modulation and demodulation is that we have had several hams talk about SDR radio and IQ modulation, but with no real good explanation of how IQ modulation really works. I know and have several books about SDR, but these books are on an engineering level with complicated mathematics. I am now working on sharpening my math skills and concentrating on trig and complex numbers. I would just like to explain IQ modulation on an intuitive level and I would love to tinker with IQ modulation and demodulation. Thanks again. Rick Phillips

I think you need a ZFMIQ-10D *Demodulator* to give you the I and Q signals, not a ZFM-10M.

Okay, Thank you.

I used a Tektronix TSG4106A vector signal generator to get the I and Q signals. The 2 LOs were generated using a Tek AFG3252B 2-channel arbitrary function generator.

w2aew thank you soo much Alan 👍🏻😄

While the paper showed that I applied COS and SIN to the I and Q respectively, the actual signals that I applied were actually reversed. I had put an annotation on the video in UKposts, but YT removed the annotation feature and thus that note was lost.

@@w2aew Thanks for the response, although now that has thrown me somewhat, as I was expecting that the IQ mixer *WOULD* be spectrally inverting, whereas you’re saying that it isn’t. My reasoning for expecting spectral inversion is as follows:- imagine you have at complex IQ baseband a phasor with a low +ve frequency, starting with zero phase at 1+j0 (i.e. I=1&Q=0) and then gradually increasing in phase towards 0+j1 (i.e. I=0&Q=1) - i.e. gradually transitioning from pure I to pure Q. If you then think about what you would get out of the IQ modulator, it should start as cos(wt), when I=1&Q=0, and then gradually shift phase to end up at sin(wt) when I=0&Q=1. So at the end of this thought experiment the RF carrier phase is lagging by 90-degrees (w.r.t. the LO) what it was at the start (since sine lags cosine by 90-degrees), so the phase has been going in a -ve direction w.r.t. to the LO, or in other words the frequency at the RF output of the IQ modulator is slightly *LESS* than the LO. Hence a +ve frequency at complex IQ baseband has translated to a -ve frequency (w.r.t. the LO) out of the IQ modulator, or in other words the IQ modulator has created a spectral inversion at RF. At around 2 minutes into the video you describe an IQ modulator which you created yourself, which conforms to the paper diagram which you drew. Can you confirm that this homebrew IQ modulator is in fact spectrally inverting, as I have reasoned above?

@@damianbevan5967 yes

@@w2aew That's great. Thanks.

@@w2aew Here's a nice tech note from Maxim, which I've just learned about, on the topic of inverting vs. non-inverting IQ modulators, which nicely sums up the whole issue: www.maximintegrated.com/en/design/technical-documents/tutorials/5/5499.html

I have a few videos on impedance matching. Here's one: ukposts.info/have/v-deo/gZeWhn57gGOmzJc.html

Because it changes the bias on the diodes in the mixer, thus changing (lowering) the input impedance of the LO port.

Modulation is used when you are going to transmit your signal through a long distance. If you want to send you baseband signal(i.e. signal without modulation with an antenna, you would need an antenna maybe a few kilo meters long). In this video, he demonstrates basically a few linear modulation methods. Pros: Simple to build. Cons: Less resistance to noise in channel.

QPSK is an expection. It's a non-linear modulation.

That seems pretty good, the rest can be done with filtering...

@@w2aew Wow, thanks Alan, I appreciate it so much! This is such an amazing hobby, where people like you are actually accessible, and even reply to newbs like me. Thanks!

Usually there is a unique data pattern in the transmitted data that is used by the receiver to establish the base phase of the signal. The receiver performs an IQ down-conversion and looks for the sync pattern in the n-number of possible constellation phase rotations.

Looks to be some kind of modulated light source, and a receiver that looks for a reflected signal at the same frequency. If I even get my hands on one, I can do a video on it.

w2aew Thanks for quick response.

I think maybe you mis-heard me. At the end of the video, I am talking about frequency *offset*, not frequency _upset_, when referring to the sideband frequency with respect to the carrier.

I'll add that to my long list.

Well, if you're going for your PhD, you'll have to teach me! I only have a BSEE... ;-), and about 40yrs experience.

Sorry about that - my site host changed the way these are indexed. Here's the correct link (also corrected above): www.qsl.net/w2aew/youtube/IQ_modulator_basics.pdf

I've covered that a little on one or two of my SDR videos, like this one: ukposts.info/have/v-deo/endhpqOEfqaorpc.html

Yes, there will just be more harmonics involved.

What's a Q36?

@@w2aew THE Q36 Space Modulator, of course. :-) By the way, thank you for all your videos. I'm nearing retirement and it is my intention to pick up my love for analog electronics. Your videos have helped remove a lot of the cobwebs from my days as a physics/electronics engineering major in college.

I did. I mentioned it right at the beginning of the video. Link is in the description. Here it is: ukposts.info/have/v-deo/oI9omGOkaJuZ0I0.html If you'd rather read about it, here's a blog I wrote: www.tek.com/blog/what%E2%80%99s-your-iq-%E2%80%93-about-quadrature-signals%E2%80%A6