100 questions I had since childhood answered in one video. Well ...

@MegawattKS

3 ай бұрын

Sounds like an interesting childhood ! FWIW, I remember somewhere around junior-high school trying to add a BFO to a multi-band/SW radio by injecting a signal at IF. I think I read about that somewhere. Sadly it didn't work. Had to wait till later to learn about the key to mixing (multiplication rather than addition - or equivalently using a strong enough signal to create non-linearity). Thanks for the comment ! Brings back memories 🙂

Thanks for making this material. It's very interesting to watch and listen to 👍

keep em coming.. thank you again.

Thanks for the prototype board designs; I've ordered some.

@MegawattKS

2 ай бұрын

Awesome. Hope they work well for you. Thanks go to Maël for making these available on github (and doing the improved layouts there!). I'm curious - Did you go with Option 1: ExpressPCB or option 2: JLCPCB (or another company)? ecefiles.org/rf-circuit-prototyping/

@Erhannis

2 ай бұрын

@@MegawattKS I've ordered from JLCPCB before, so I went with them again.

Fantastic video. I like the intro the I/Q and zero-IF. I had no idea why or how that worked. This video has given me the foundation to glue the concepts together and perhaps take a deeper dive on the subjects. Thank you, sir!

@MegawattKS

2 жыл бұрын

Thanks. Good to hear that it was helpful. I've done IQ modulators (for QPSK), but never really written code for demod. Here's a decent introduction to some of what's involved (but it seems to be limited to AM - in general it can get pretty messy mathematically I think - especially for FM demods): www.allaboutcircuits.com/textbook/radio-frequency-analysis-design/radio-frequency-demodulation/understanding-quadrature-demodulation/

@emailuser3869

2 жыл бұрын

@@MegawattKS 73's W6MGV

Than-you so much for the very clear explanation of how mixers work. This was part of my syllabus for my ham licence here in the UK. Then I had to accept that was the way mixers work. Now I understand HOW they work. I have a Nano VNA and was weighing up whether a Tiny SA would be useful and more than a toy for ham radio construction. Seems you have answered that question too! Thanks again.

@MegawattKS

2 жыл бұрын

Glad it was helpful. Enjoy the TinySA if you take the plunge and get one. It doubles as a signal generator too, and even has some amplitude control and ability to generate CW, NBFM and WBFM signals. I keep thinking I should get a second one for that reason alone ! See specs here: tinysa.org/wiki/pmwiki.php?n=Main.Specification

👍Very interesting presentation, thank you.

Excellent explanation of the I/Q signals. I could have used this when I first learned it, took me forever to learn it the way I did.

@MegawattKS

2 жыл бұрын

Glad it helped! In the course we actually go into the quadrature nature in more depth, but glad to here this brief treatment was useful. I remember the first time I met this stuff - in a book on Fourier Analysis and "bandpass signals". That can get pretty deep, but is insightful. In the course, we generally stick with digital BPSK -> QPSK modulations to introduce the topic without having to get into "complex signal processing" stuff...

@fjs1111

Жыл бұрын

@@MegawattKS Thank you!!

Excellent ! Thank you.

@MegawattKS

10 ай бұрын

You are welcome!

very helpful

I don't have a _TINYSa,_ but I do have a _HackRF One (with PortaPack)_ and I *THINK* it has similar functionality

Thank you!!! I subbed for sure. 73!

@MegawattKS

2 жыл бұрын

Awesome, thank you!

Thenk you for your tutorials. For the receiver at 2:30. How does the demodulator work? Is the signal simply just amplified until it clips either at the top or bottom? And then averaged by the headphones?

@MegawattKS

8 ай бұрын

Yes. The last tube acts as both a rectifier for the AM signal being processed, as well as a current amplifier. If the tube is biased lightly on (I don't know the exact bias details needed for best receiver sensitivity, but I'd assume we want to be close to cutoff), then, as the RF signal out of stage 2 comes into the grid of V3 (the third tube), it pulls current through the headphones during the positive half cycle and no current during the negative half cycle. Essentially what you said. A higher amplitude of RF sinewave coming in leads to higher current on the positive half cycle, and hence more average current - so AM is converted into varying "DC" (an audio waveform) . As you noted, the pulsing RF current at the plate of V3 is converted to the audio acoustic waveform when averaged by the response time of the headphone transducer mechanical element 🙂

At 9:25, those question marks are 2x70 MHz and 2x70-20 MHz.

I remember *USING* (deliberately) image frequencies to receive _"out of band"_ signals, such as the 850 mHz *analog* cellular band, back when monitoring it was still legal and it still existed. I wonder if FCC regs still forbid the sale of scanners capable of (or _"easily modifiable")_ receiving that frequency range.

@MegawattKS

Ай бұрын

Wow. That brings back memories. Old UHF TV tuners could also receive in that range as I recall. That's where the original cell spectrum came from in the US (channels above 69). Good question about the scanner regs. I don't know. But even as they were rolling out the regs, cell phones switched to digital modulations so it became a non-issue anyway, I think.

@ChaplainDaveSparks

Ай бұрын

@@MegawattKS Analog service was slowly phased out even as digital service started to replace it. I was told that the very last analog site to be decommissioned was in Death Valley, CA. (Sort of fitting.) I remember when the old UHF TV tuners were analog (continuously tunable), until the FCC required manufacturers to provide channel detents. The weird thing now is the ATSC channel numbering system. I used to wonder why KABC channel 7 was so difficult to receive. Then I found out. They're about the only area channel to retain a *VHF* frequency. Most antennas (antennae?) are optimized for the UHF band.

Great course! I'm watching it again (and again) to get these concepts. I was curious what type of breadboard is that and are the capacitors and zero ohm resistors 0603 scale? Thanks.

@MegawattKS

Жыл бұрын

Thanks. You can find some info on the breadboards here. The 0 Ohm resistors are 0603 and the caps are 0805. The pads are 90 x 90 mil on 100 mil centers. ecefiles.org/rf-circuit-prototyping/

@W1RMD

Жыл бұрын

@@MegawattKS Thanks!

24:00 that what i previously thought about how SA might work by scanning frequency but than i found Many places that SA uses FFT algorithm so now my doubts is if there's two mathod of SA or both mathods are used tougher scanning and FFT ??

@MegawattKS

Жыл бұрын

Good discussion point. The TinySA and all classic RF spectrum analyzers work as shown, BUT - there was a shift about 15'ish years ago to a hybrid approach for high-end analyzers. It's always been too difficult to directly digitize at the front-end and use an FFT directly at RF (due to technology speed, power consumption, and spurious signal generation). But the hybrid approach is able to digitize at the IF and then do the FFT. For example, the traditional analog method may sweep a GHz range with a 10 MHz wide IF filter, and then digitize the 10 MHz bandwidth and then FFT that with a 1024 point transform to get down to 10 Hz resolutions. This provides _much_ faster sweep times for narrow resolution bandwidth settings. As ADCs got faster, this could be done on wider bandwidth IF chunks, opening the way for spectrum analyzers to also do "signal analysis" (e.g. look at signal constellations and other modulation features). But it costs more. Hence the TinySA uses the original non-hybrid approach (and doesn't have 10 Hz bandwidth capability as a result). For the record, I think some high-end scopes are starting to do some spectrum analysis with FFT directly at RF (no mixing to IF) since they already have 10 Gsps ADCs, but they're still going to suffer from spurious problems, bandwidth limitations, power consumption/etc...

@omsingharjit

Жыл бұрын

Ok , that's Very nice Explanation thank you .

thanks

@MegawattKS

2 жыл бұрын

You're welcome!

Is there a way to measure Mixer gain performance with NanoVNA? Rhode & Schwartz documents recommends using reference locked VNA and Signal Generator to perform analysis with VNA. Some method of quantification, would help with performance tuning the RF input matching network. LO signal power's influence on mixer output levels, higher LO power does increase mixer output levels, but also larger LO breakthrough too. What exact levels of IF output are expected for unbalanced inputs(50 ohm source) and output, with SA612? It appears Gain of SA612 works only at RF levels

@MegawattKS

2 жыл бұрын

You should not see gain compression before about -20 or -22 dBm input with a matching network in place. As to measurement, I don't know that the NanoVNA can do it. In the class, we just supply known input signals and look at the IF output on a spectrum analyzer. Then we plot P_if_out vs P_rf_in like in the datasheet (at specific LO levels). As I recall, once the LO is around 200 mVpp, the "conversion gain" starts to become constant at the datasheet's value of 15'ish dB. But that assumes the RF input match is good and both sides of the output are used through a proper match at IF. Hope that helps.

Perfect. Thank you. M0XYM

@MegawattKS

2 жыл бұрын

Thanks for the feedback ! 73s

I am so interested in your video that I can't help replicating the entire project. Now I am replicating this RF protoboard, but I am curious how the IPEX(U.FL) was soldered to these pads with 2.54mm (Judging from the way DIP-8 IC was soldered, I guess it is 2.54mm? ) pitch? Since the largest IPEX connector (IPEX gen 1) has only 1.475mm pitch between signal pad and ground pad.

@MegawattKS

2 жыл бұрын

Thanks for the interest ! On the board used for the LNA, I followed the footprint in the u.fl part's datasheet. Then on the general RF protoboard I deleted the pad at the edge of the array and placed the footprint I had used with the LNA board into that space. So the 'wings' of the ground pads reached to the other pads :-)

@sullivanzheng9586

2 жыл бұрын

@@MegawattKS Got you. I actually just submitted to JLCPCB my own design of RF protoboard (generally followed your design) with 8 u.fl. footprints surrounding each protoboard area. That coincides with your design. I will open source the Gerber file on github later on. Again thanks a ton for these videos. These videos really gave me a way to kickstart my journey in RF design. Esp the episode of SA602/612 mixer. I got a bunch of SA612 in my inventory but the datasheet is not friendly to someone who just started (lots of discussions on balanced-unbalanced-tuned signals, not just the basics of mixing). Your video taught me the minimal (non-trivial) circuit for SA602/612

@MegawattKS

2 жыл бұрын

Wonderful. We used basic versions of boards like these for many years in the class, but had to drill holes and inset wires at each pad where we needed a backside-groundplane connection in the circuit. With the boards in the video it's tricky to bridge the via to the ground-plane where a connection is needed - but its a lot easier than what we did with the prior boards that had no vias predrilled. The limitation we never have solved well is how to prototype with the boards when a SOP or SSOP chip is used. We've done it successfully with tiny adapter boards that bring the pins out to a 100-mil pitch, and in Spring 2019 we had a board for the final project (that used SA615 IF subsystem) where we had a dedicated footprint on the board for that chip. On the plus-side, soldering 0805 or 0603 discretes and even SOT23 transistors is quite easy as you seen in the videos. Glad they're helpful. Thanks for your feedback !

@sullivanzheng9586

2 жыл бұрын

@@MegawattKS Very very interesting. I have been working on several versions of RF protoboards before. Since I used JCLPCB (2 layers and 4 layers) to manufacture professional PCBs, via holes have never been a problem for me. Before using JCLPCB/PCBWay, I also DIY PCB at home using both dry film and toner transfer tech stacks. But finally I gave up since JCL has much better precision and 4-layer board is just impossible for home brewing. (and they offer every hobbyist 2 free PCB prototyping coupons every month with free shipping to China...So I am designing and manufacturing PCB essentially at 0 cost. And the turnaround time is only 4 days including shipping.) Regarding SSOP, TSSOP and many fine pitch footprints -- yes they are big headaches for me as well. My workarounds are: 1) try to use SOP or DIP or SOT-86 ("RF cross") when possible 2)For many ICs that I can't refuse to try (there are so many, such as wideband LNA/RF gain blocks e.g. TQP3M9009/BGA2869 , ceramic filters, down converters, frequency synthesizers....etc.) I will just put pre-defined circuits for those ICs in pre-allocated areas on the protoboard. 3) For some frequently used circuit topology such as pi bridge, 3~6-order shunt-first Chebyshev filter based on 0402 SMT LC, I will also dedicate specially defined areas on protoboard. 4)There are always sufficient UFL and a few SMA connector footprints for intra-board and inter-board RF connections. So the final protoboard is a mix of pre-designed IC/filter zones + free style improvisation areas with 2.54mm pitch pads and though holes to the backside and some with solder bridge jumper to ground plane so you can easily solder some of the single pad to ground as you need (this might be a dangerous design for high freq since such solder bridge jumper may have substantial parasitic capacitance to ground, therefore I only designed solder bridge for some of the pads). The entire 4 layer PCB has a ground plane right below the surface layer to ensure stable clean ground reference and 50 ohm characteristic impedance for microstrips in those pre-defined IC zones.

Where do you get ground plane perf boards?

@MegawattKS

Жыл бұрын

The one shown in the video is a custom one that I designed and had fabbed though www.expresspcb.com/ . For something like 60 bucks, you can get three 2.5x4" boards made, which allows for a total of 6 x 3 = 18 of these. I tried a few variants (hole sizes, gaps to ground plane, etc.) and what you see in the video is the one that worked best for this. It would be nice if someone marketed these. In the university course, we used to get ones from a small company in AZ (a bit cheaper than doing one's own - even using ExpressPCB). But they stopped making them. And they were a bit of a pain anyway because you had to drill a hole to insert a via wire for ground connections. (Perhaps good for students to learn to use a drill press :-) ) I tried what you see in the video and it seems to work quite well, not requiring drilling - though as you saw, the ground connections are still a bit awkward to do because of the difficulty of bridging due to solder surface tension...

@MegawattKS

Жыл бұрын

Here is the closest I found commercially. But I like mine better :-) These do allow for use of smaller footprint parts, but the ground locations aren't very flexible in terms of layout options... www.amazon.com/SP3UT-SMTpads-3U-Thin-Sided-Unplated-Ground/dp/B0040Z8FC6/ref=sr_1_42?crid=3RUU97R2BDXDX&keywords=rf+pcb+pads+ground+plane&qid=1662231961&sprefix=rf+pcb+pads+ground+plane%2Caps%2C84&sr=8-42

you should have build a playlist for this

@MegawattKS

Жыл бұрын

Sorry. There is a playlist, but I may not have enabled it properly so it is easy to find in KZread. I will see if I can improve it's visibility. (Where do you normally look to find playlists when watching videos?) In the mean time, here it is: kzread.info/head/PL9Ox3wpnB0kqekAyz6blg4YdvoEMoJNJY

@MegawattKS

Жыл бұрын

Thanks. I may have it set properly now. The playlist was not set as an "official series" so it didn't move into the next in the series at the end of a video. I think that is enabled now. 🙂

@MegawattKS

Жыл бұрын

For anyone interested in related series playlists, there is an associated website here: ecefiles.org/ It has links to the 4 main playlists as well as uploaded PDFs of the slides used.

@dalepres1

17 күн бұрын

@@MegawattKS Thank you so much for sharing the slideshows and the videos. So many share one or the other and both definitely makes for the bests. As a long time ham, then a career in electronics in the Navy, followed by a career in two-way radio, I was once really good at anything RF but when my next career was in IT, I forgot so much that I'm studying again in my retirement to relearn theory, hopefully to the degree I knew it in the 70s, 80s, and 90s. I'm looking forward to watching more videos, reviewing more slideshows, and picking up a lot of solid electronic theory and, based on what I just saw in the section on Mixers, I should have my expectations met.

@MegawattKS

7 күн бұрын

@@dalepres1 Thank you so much for the kind and encouraging words. Glad you found the slides on the companion website and that the combination is helpful! I also spent some time in software development (XSPICE) before coming back to RF - my original passion. Your path through all of this is certainly an asset in today's complex world. Best wishes in retirement. A well-earned new chapter in life. 73s

Sir. What happens when LO and RF frequencies are same?

@MegawattKS

Жыл бұрын

Interesting question. If both are simple sinewaves, then mathematically, the output is DC and depends on the phase relationship between the two. Of course the frequencies are never exact in the real world (unless they come from two outputs of a power-splitter or something), so the "DC" is not perfectly constant. The phase changes over time and so does the DC output. Now, If the LO is a sinewave and the RF is a modulated signal, then the spectrum of the signal is "brought down to baseband". This is what's done in "direct conversion" receivers used in cellphones. At that point, the "I" and "Q" outputs are digitized and it is possible to demodulate using DSP code.

@irfan_antennas

Жыл бұрын

@@MegawattKS Thank you so much.

@irfan_antennas

Жыл бұрын

@@MegawattKS I am working on finding the strength of the transmitted signal from a source at various distances? Is there any way that could help me?

@MegawattKS

Жыл бұрын

@@irfan_antennas Not sure. But if you're interested in signal strength, I would recommend looking at the "TinySA" spectrum analyzers. That's their job - measure signal strength and frequency (across a wide band of frequenies). There's one that goes to about 900 MHz, and one that goes to about 6 GHz: www.tinysa.org/wiki/

The IMSAI Guy channel is a great one for anyone interested a wide variety of electrical engineering topics. IMSAI comes from the IMSAI computer (en.wikipedia.org/wiki/IMSAI_8080) which was featured in his early videos.

@MegawattKS

2 жыл бұрын

Agreed. I've watched a number of his videos !