I've been working with electronics for more than 10 years and by god RF still sounds like forbidden witchcraft out of a Lovecraftian book. Congrats dude your channel is amazing. I got here by hackaday! Always nice to see a maker channel creating original and fun content. Cheers!
Hey man! New subscriber, just dipped my toes in the microwave dark magic. I managed to build a 5.6GHz synthesizer with LTC6946 and SKY85712 on a board (FR4 substrate) designed by a fellow HAM. Having the board premade was easy enough to build the synthesizer to output +15dbm at 5.616 GHz. However, I mostly learned SMD soldering (very good lessons altogether) but less design. Excellent channel, you give me the courage to try and test for myself with microstrip technique. your approach is much more hands-on and I believe this is what I need for my development. In theory, this would work with lower frequency input signal, but the spurious would be much higher and probably harder to filter out. Greetings from Romania, Octavian
Your video is amazing. To create a multiplier in this manner is something I would have never thought about until I saw your video. That being said, there is one very important point you should have mentioned. If I was a lay person, I would just take your word for it without questioning; however, with some knowledge of digital signal processing, here are some additional sentence that need to be added: when a sine wave is changed from time domain to frequency domain, there is one fundamental frequency. When sine wave is over amplified and top and bottom parts of the sine wave get chopped off and turned into almost square wave and finally transform this almost square wave from time domain to frequency domain, there will be lots of frequencies that are multiple harmonics of the sine wave fundamental frequency. As you mentioned, 2.5GHZ, 5GHZ, 7.5GHZ, and so on. Just my two cents thought.
Brilliant video, thank you! I learnt a lot from this and will try to reproduce once I get a higher bandwidth SA. Also, really enjoying your Portuguese channel, even though I have to sadly use subtitles ☺️
Very nice job and explanation. The BFP420 is a very nice transistor for stuff like this with its 25GHz transistion frequency while still being solderable by hand. I use it for my experiments as well. Did you notice any artifacts by only grounding one of the emitter pins? I guess for the calculation of lambda you use a shortening factor for scaling to the FR4 medium? One interesting thing is to try and use some parasitic effects of your components in a useful way. Like the coiled up leg of your through hole resistor. In general I find through hole resistors quite nice for bias voltages, since most of them have some inductive component to them already. I totally agree with you about the experimenting approach. Not only for RF stuff but generally, I find it way more efficient to just try some stuff instead of only reading about it.
I didn't find any artifact, at leat for this frequencies. But more investigation could show something about that spurious tones? It may be related. I didn't use any factor 😅🤣 Sometime ago I backcalculated the effective Er to be about 4.2, using a notch 1/4 stub, I'm using this for all tests Yep!! The resistor are great fot biasing. That coil I made to act as an inductor to help with isolation.. some turns on the lead become a good inductor for GHz
excellent job again and good explanation as usual. thank you. how do you measure and tune those two quarterwave transformers (at 5GHz) to get the best RF isolation? do you do that when the transistor is in the circuit? do you inject a 5GHz tone to the DC bias? what do you measure to tune those two? or maybe you dont tune them at all?
FR4 is not considered a RF grade substrate.. So, there is a lot of variability, and you don't have proper characteristics. I recommend you to buy some and test. This is what I do. In the end, everything I buy is useful at some point. Don't expect to make a very precise design in FR4, tuning it a lot, to work when you get a new batch of pcbs.
Get it! I buy from local market, much worse that yours haha. I recommend you to etch a stub. Measure the frequency of the notch, and calculate the material Er backwards. Now you know the Er of your pcb for usage in yours designs. Get some 1.6mm and 1.2mm FR4 and GO TO THE BENCH hahaha Tip: you already thought too much. Just buy it. My approach (my big secret) to design is: try first and think later =] Imaging what would have happened if, before I record this video, I asked for someone about using copper tape over a PCB hahaha. I would never have done this video!
Thanks again. I am still learning about things like ER. How do you measure the ER of a trace? I have a NanoVna and other ways to measure impedance. I am researching ER to try to understand what ER is.@@AllElectronicsChannel
Very Good Video as allways. What is the Maximum harmonic That you can pick easyly with an analog multiplier? I have Seen some Videos uses the 9 Harmonic or the 11 very interéting topic
Thank you! What matters is the rise or fall time of the distorted wave.. You can grab higher and higher harmonics with faster transitions. With the transistor amplifier I believe up to the 5th harmonic you still have nice power. I have a video on the channel using a PIN diode step characteristic to extract the harmonic 10.
Could the spurs you see also come from higher order harmonics that are passed and may be beyond the range of your spectrum analyzer but not rejected properly?
Actually when you use this technique, it generate odd harmics 3th 5 th and etc the even harmonics amplitude is less this is much more efficient for generating odd harmonics not even
I'm confused, how did you get an even harmonic from a square wave? Is it possible that the transistor is creating a triangle-ish wave, or so many harmonics are bouncing off each other that you eventually get 5Ghz.
@@AllElectronicsChannel I wanted to know how I could design it on AWR, not a lot of material on the internet to guide for that, your video is the closes I have come
![[#2024MAMA] BIGBANG (빅뱅) - 뱅뱅뱅 (BANG BANG BANG) + FANTASTIC BABY | Mnet 241123 방송](http://i.ytimg.com/vi/EQsYfiPR9uM/mqdefault.jpg)
Support the channel being a Patron!
www.patreon.com/allelectronics
I've been working with electronics for more than 10 years and by god RF still sounds like forbidden witchcraft out of a Lovecraftian book. Congrats dude your channel is amazing. I got here by hackaday! Always nice to see a maker channel creating original and fun content. Cheers!
Thank you!! Welcome to the channel 😁😁😁
Thank you for making microwave accessible to hobbyists!
😀😀 Welcome to the channel!
Hi Grégory, felow brazilian here, studying microwave electronics from Belgium! Nice videos, excellent channel!
Thank you man! Welcome to the channel!
Hey man! New subscriber, just dipped my toes in the microwave dark magic. I managed to build a 5.6GHz synthesizer with LTC6946 and SKY85712 on a board (FR4 substrate) designed by a fellow HAM. Having the board premade was easy enough to build the synthesizer to output +15dbm at 5.616 GHz. However, I mostly learned SMD soldering (very good lessons altogether) but less design.
Excellent channel, you give me the courage to try and test for myself with microstrip technique. your approach is much more hands-on and I believe this is what I need for my development.
In theory, this would work with lower frequency input signal, but the spurious would be much higher and probably harder to filter out.
Greetings from Romania,
Octavian
Be welcome! That is nice to hear. We are always learning, the thing is to go to the bench and try, try, try!
Your video is amazing. To create a multiplier in this manner is something I would have never thought about until I saw your video. That being said, there is one very important point you should have mentioned. If I was a lay person, I would just take your word for it without questioning; however, with some knowledge of digital signal processing, here are some additional sentence that need to be added: when a sine wave is changed from time domain to frequency domain, there is one fundamental frequency. When sine wave is over amplified and top and bottom parts of the sine wave get chopped off and turned into almost square wave and finally transform this almost square wave from time domain to frequency domain, there will be lots of frequencies that are multiple harmonics of the sine wave fundamental frequency. As you mentioned, 2.5GHZ, 5GHZ, 7.5GHZ, and so on. Just my two cents thought.
Thanks man!
Cool! Thanks for teaching!
Brilliant video, thank you! I learnt a lot from this and will try to reproduce once I get a higher bandwidth SA. Also, really enjoying your Portuguese channel, even though I have to sadly use subtitles ☺️
Very nice job and explanation.
The BFP420 is a very nice transistor for stuff like this with its 25GHz transistion frequency while still being solderable by hand.
I use it for my experiments as well.
Did you notice any artifacts by only grounding one of the emitter pins?
I guess for the calculation of lambda you use a shortening factor for scaling to the FR4 medium?
One interesting thing is to try and use some parasitic effects of your components in a useful way.
Like the coiled up leg of your through hole resistor.
In general I find through hole resistors quite nice for bias voltages, since most of them have some inductive component to them already.
I totally agree with you about the experimenting approach.
Not only for RF stuff but generally, I find it way more efficient to just try some stuff instead of only reading about it.
I didn't find any artifact, at leat for this frequencies. But more investigation could show something about that spurious tones? It may be related.
I didn't use any factor 😅🤣
Sometime ago I backcalculated the effective Er to be about 4.2, using a notch 1/4 stub, I'm using this for all tests
Yep!! The resistor are great fot biasing. That coil I made to act as an inductor to help with isolation.. some turns on the lead become a good inductor for GHz
Thank you for sharing this amazing design and information 👍
Thanks!
You just blew my mind...
peace.
excellent job again and good explanation as usual. thank you. how do you measure and tune those two quarterwave transformers (at 5GHz) to get the best RF isolation? do you do that when the transistor is in the circuit? do you inject a 5GHz tone to the DC bias? what do you measure to tune those two? or maybe you dont tune them at all?
I didn't tune them!! Performance don't seem to be much sensitive about them
Very nice. Can you make more of such experiments ?
Yep!
Question. What fr4 double sided pcb would you recommend? I looked at some sites and I am not sure which pcb to get. Thank you for your videos.
FR4 is not considered a RF grade substrate.. So, there is a lot of variability, and you don't have proper characteristics. I recommend you to buy some and test. This is what I do. In the end, everything I buy is useful at some point.
Don't expect to make a very precise design in FR4, tuning it a lot, to work when you get a new batch of pcbs.
Thank you for the reply. Any suggestions as to what to start with? I was looking at pcbs from Mouser.@@AllElectronicsChannel
Get it! I buy from local market, much worse that yours haha.
I recommend you to etch a stub. Measure the frequency of the notch, and calculate the material Er backwards. Now you know the Er of your pcb for usage in yours designs.
Get some 1.6mm and 1.2mm FR4 and GO TO THE BENCH hahaha
Tip: you already thought too much. Just buy it. My approach (my big secret) to design is: try first and think later =]
Imaging what would have happened if, before I record this video, I asked for someone about using copper tape over a PCB hahaha. I would never have done this video!
Thanks again. I am still learning about things like ER. How do you measure the ER of a trace? I have a NanoVna and other ways to measure impedance. I am researching ER to try to understand what ER is.@@AllElectronicsChannel
Er is a propertie of the material, tale a look with calm at my last message
Very Good Video as allways. What is the Maximum harmonic That you can pick easyly with an analog multiplier? I have Seen some Videos uses the 9 Harmonic or the 11 very interéting topic
Thank you! What matters is the rise or fall time of the distorted wave..
You can grab higher and higher harmonics with faster transitions.
With the transistor amplifier I believe up to the 5th harmonic you still have nice power.
I have a video on the channel using a PIN diode step characteristic to extract the harmonic 10.
Could the spurs you see also come from higher order harmonics that are passed and may be beyond the range of your spectrum analyzer but not rejected properly?
Yeep!! Very good hypothesis
Actually when you use this technique, it generate odd harmics 3th 5 th and etc the even harmonics amplitude is less this is much more efficient for generating odd harmonics not even
Yes, you are right. But the waveform is not so symmetrical as you are imagining, a lot of energy goes to even harmonics.
I'm confused, how did you get an even harmonic from a square wave?
Is it possible that the transistor is creating a triangle-ish wave, or so many harmonics are bouncing off each other that you eventually get 5Ghz.
Take a look here! th-cam.com/video/acWZLW5EE3U/w-d-xo.html
@@AllElectronicsChannel Thanks
Take a look at 'Steve Kim's comment also because this will clearly explain your question.
I have been striving to design a 1.2GHz to 2.4GHz, any pointers?
Hi friend, I think that this design I pretend may work great.. give it a try
@@AllElectronicsChannel I wanted to know how I could design it on AWR, not a lot of material on the internet to guide for that, your video is the closes I have come
Never used AWR..
The brabo of brabos
is that the language of my fellow br?
nice!!!
BAM, I learned a few new things already. subbed.
Thanks HaD for mentioning this content. I am ready to learn the magic of RF, by just doing it :D
would the correct term be: "Giggity" - Glen Quagmire
🤘🤘
Odd...lav mics usually sound GOOD...
It didn't record at that section 😪