You are a great teacher. You provide the underlying math and theory. Then you build the circuit and show the operation on lab grade test equipment. Simply brilliant.
i pressed like before watching the video..... the even harmonics disappeared when we added the 2nd diode because now the signal became a square wave like which only contains odd harmonics. i would urge you and even begg you shahraiar to write a book with all these interesting experiments and i promise you it will be the best selling of all times. million thanx for your excellent videos.
i think (and i truly do not know) you've got that backwards .. it lost the harmonics and became a square(ish) wave, not that it became it and lacks them. but why? is it because the +/- voltages of the even harmonics cause a phase cancellation to occur?
because the square wave is half wave symmetric signal..f(t)=-f(t+T/2)...there is a proof for this theory. www.allaboutcircuits.com/technical-articles/the-effect-of-symmetry-on-the-fourier-coefficients/
Great videos! From Signals and Systems course when the content of even vs odd functions are covered we all learn these things about how symmetrical distortion is related to odd order harmonics and asymmetrical distortion is related to even order. Wonderful demonstration of this around 19:00 to 20:00 with the inclusion of diodes in half vs full clamping orientation. Check out some guitar pedal circuits that utilize this, such as the Boss OD-1 and Ibanez Tubescreamer.
I'm a audio repair guy without any real technical training and this is the first video of yours that I have seen and I LOVE it! It's helped me to understand several concepts that I've run into in dealing with the audio domain and it's amplification and manipulation on various fronts. My math capabilities are quite rudimentary but I could follow your explanation rather well and I believe that to be a good sign! For me and for you :) So thank you for doing these. M
I have to say, you made a very good video demonstration together with the explaination. Honestly speaking, it woks much better for me to understand the theory than getting a lecture in the university.
@epatto No problem, I'll see what I can do. I have an array of new videos and tutorials which I will be doing during this month. So check back mid Feb! :)
I think the answer has to do with the odd terms for the current cancle out, since the diodes are back to back connected, so the even terms stay (and add up) and the odd terms cancle out. GREAT videos, I really enjoy them, and I hope you keep up the good work.
These are some of the best tutorials on RUclips. I would only suggest to use a multiple camera setup so they are also a bit shorter and more to the point. Right now there is a lot of camera movement involved breaking the flow. One or two mobile phones would do the trick, RUclips also has a video mixing capability which is pretty intuitive and fast. Keep up the nice work!
The even harmonics are attenuated because the diodes clip the signal to be more like a square wave which has only odd harmonics. Thanks for the videos SignalPathGuy, as long as you're making them I'll be watching, especially looking forward to the mixer one.
Thanks a lot for this, it really helps to see this demonstration and explanation on high quality equipment! If you ever get around to the mixer video I'll be celebrating :)
Me too. Only I would like to see the tutorial as part of a practical build similar to how Ben Eater designed his Breadboard Computer ruclips.net/video/HyznrdDSSGM/видео.html.
@3cu14rs Yes, you are right about the fact that a square wave does not have even-mode harmonics. But what about at the circuit level, what is happening there?
you said that the diodes only 'kick in' after a certain point, and then the second diode is flipped in the other direction allowing the opposite voltage, so is it that the even harmonics are causing perfect phase cancellation?
I would really like to see what's inside the spectrum analyzer. If you bought it broken, the fixing would be a great story as well! Keep up the good work!
2019 watching this video and the answer for the question in the 19:50 is because symmetric and Asymmetric clipping effects diffrently on the harmonic content of signal,one causes to produce even harmonics and the other will cause to produce odd harmonics.
The diodes are in parallel and opposite. Harmonics mean power of basic input function (third -> power of 3); power surely keeps the original sign if odd (while - * - = +). At every [half] cycle diodes have same harmonic values but all opposite in sign; so due to their positions, the difference is the result. Then even components are eliminated, while the odds...just pass (no add because diodes don't amplify). P.S.: harmonics gain a little bit in amplitude: maybe a sort of power redistribution?
Maybe the one-diode clamp distorts the sine asymmetrically, therefore it behaves as a kind of sum of some kind of a half rectified sine/saw plus a square wave, and a full clipped sine is symmetrical so it behaves like an almost square wave? Thanks for share all of your knowledge, Its something very Valuable nowadays!!!
Hi Shahriar Very impressive video. You explain some very complex concepts in a clear and concise manner while maintaining a presentation that is very engaging.Regarding your question I will take a stab at it. The even order products introduced by each diode are 180 degrees out of phase. So when they are summed together they cancel each other out. Am I right ?
@28:06 If the start frequency was at 0 MHz instead of 1 MHz, would we see a spike at the Beat Frequency (ie at 100KHz) and would it's amplitude be the same as that of the two tones? Thanks a lot for your videos. Much appreciated.
@TheSignalPathBlog A harmonic is defined as a multiple of the fundamental K. The first diode and the introduction of a sin wave creates the non-linear output due to the e^x attitude of how a non-linear diode operates (as a sharp change in definition is non-linear). Since the diodes are the same and therefore considered balanced, the even order harmonics are canceled by only allowing odd harmonics due to the diodes I to V non-linearity being closely matched during manufacturing.
Correct If I am wrong!!! When the first diode is for all odd term is active that mean is Sin90 (+90) that mean +1 and when second one you add parallel with output is active then it get Sin(-90) that mean -1 they both cancelled each other.
when you add one diode there comes 1st 2nd 3rd 4th....nth harmonics created but when when you add second diode you add also 1st 2nd and 3rd ...nth harmonics,,,but even(2,4,6..) harmonics produced from both diodes cancels out and odd harmonics gets added(in your spectrum analyser we seen gain of odd harmonics after two diodes addition.)
A bit late, but I think the answer is related to the symmetry of the waveform relative to the time axis when the two diodes are present. With one diode, there is no symmetry, since the signal is just going to be clamped in one side. When two diodes are used, the signal is clamped in both sides. When calculating the Fourier series coefficients for a signal like this, the even coefficients are 0, it comes out of the calculation. I guess it's what is happening here. When Shahriar puts the diode, it's even possible to see the even harmonics slowly going to 0. I think this is due to the temperature of the diode slowly increasing?
Here' are two questions not related to the non-linear portion of your demonstration. I'm simply exploring the circuit with the mixed signals with just the resistor divider portion of the circuit. Regarding that brass fitting which combines the output of the two signal generators... Does that have circuitry inside? Or is that just a simple coupler and you're merely superimposing one signal atop the other? Now to my main question: I'm puzzled why I only saw two beautiful pure peaks on the spectrum analyzer when you were mixing two signals through the plain vanilla voltage divider. Forgive me if this sounds stupid. But I thought heterodyning generates sum and difference signals of two pure sine waves (let's say each with zero THD). My question is: How come your spectrum analyzer showed only two pure peaks before you applied the clipping diodes? Why didn't I see any sum/difference signals? Shouldn't I have seen a 100 KHz peak and a 3.1MHz peak? One step further: Let's say that they were there and that you were somehow hiding them on the analyzer to exclude those ranges... Well, once you have a 100KHz sideband signal, would that not (again) heterodyne with each of the two fundamentals, generating signals at 1.4MHz (visible), 1.5 (hidden within the fundamental peak), 1.6 (hidden again) and 1.7 (visible)? I'm genuinely troubled by seeing just two pure peaks on the spectrum analyzer. Responders please be gentle. I understand that I don't know what I'm talking about because those two pure peaks are proof that I'm overthinking this. I'm hoping to learn why/how heterodyning is not happening with this circuit.
Hi very nice videos, thanks! I have a question for you... at time 29.24 we can se the intermodulationproducts from your two-tone test. If one do the math, the upper and lower IM-products should have the same amplitude if the amplitude of the two inputed signal is the same, right? But as i noticed all the upper IM-products have more energy then the lower IM-products, its not symmetrical. Its mostly seen when watching the IM3-product. So my question for you is why does this happen? My idea is that the two diodes are not identical, they are missmatched, so the taylor expansion for each of them wont have the same coefficients. But if that is the case, what would happend if you switch place on the diodes in the design. Would it be that all the lower IM-products will have higher amplitude? Im curious to know!
The coefficients for the FT in case of diode 1 are all positive. The coefficients for the diode 2 have the magnitude as in diode 1 but alternate in sign. When you add all the coefficients from both diodes the resulting every other peak disappears and remaining peaks double in amplitude. I think. I would have to work it out.
When you summed the exponential series with the sine wave in it, and you got the powers of the sines, you suddenly got the fourier series with sin(kw), but you didn't show how the cosine terms disappear from it.
+thesignalpathblog why have you connected the two capacitors on the bread board ?? what is their use ?? And when you added diode to the circuit it was actually clipping the negative and the positive half cycles and not clamping it right ??
Those are the same thing. Clipping and clamping are the same thing in this circuit. The capacitors are just there to decouple the power supply (get rid of any noise the power supply might introduce to the circuit). They are not part of the circuit under examination.
The capacitors are not for noise from the power supply. They are providing the "pulse" power to the circuit. The leads from the power supply to the breadboard have inductance and so cannot deliver power pulses fast. The capacitors are capable of delivering those fast pulses.
like and subscribed! i want more videos :D Especially when you explain how sole circuit works . I like when you make a practical demonstration/implementation of the theory described in the video. Very educative ! Thank you!
In the initial expression for id for the first nonlinear circuit what is Vt? is is simply the value of Vin at a particular time t? if that is the case, then surely Vout could never be bigger than Vt, right?
+Zambia95 Vt is thermal voltage, basically the ratio of temperature (in energy units, kT) to charge of an electron. By the way, about this equation, we're in the land of being king of sloppy (not that it's a bad thing;p) by getting (V_out/V_t)^k/k! and then summing over i ;)
The circuit is non linear so once you have the linearized system consider the transfer function in both cases or rather in presence of diode and not. The even harmonic will disappear if and only if the transfer function is affected by pole zero cancellation.
sir i am from pakistan and your video has too much knowledge but sir there is problem that you speak fluently and i can`t understand some of the points from your videos so sir please speak slowly in your new videos because sir we have language problem.
You are a great teacher. You provide the underlying math and theory. Then you build the circuit and show the operation on lab grade test equipment. Simply brilliant.
i pressed like before watching the video.....
the even harmonics disappeared when we added the 2nd diode because now the signal became a square wave like which only contains odd harmonics.
i would urge you and even begg you shahraiar to write a book with all these interesting experiments and i promise you it will be the best selling of all times.
million thanx for your excellent videos.
Already a book like this -- called the Art of Electronics.
Note if you do get this (and you should) get the latest printing of the 3rd (and last) edition. Later printings have more of the errors fixed.
i think (and i truly do not know) you've got that backwards .. it lost the harmonics and became a square(ish) wave, not that it became it and lacks them. but why? is it because the +/- voltages of the even harmonics cause a phase cancellation to occur?
because the square wave is half wave symmetric signal..f(t)=-f(t+T/2)...there is a proof for this theory.
www.allaboutcircuits.com/technical-articles/the-effect-of-symmetry-on-the-fourier-coefficients/
Agreed.
Make some money my Brother.
O.M.G.!!! This channel is pure GOLD.
I'm speech-less. Thank you so much!
I agree - so good!
Outstanding video for both RF Engineers and also for Amateur Radio Operators.
Great videos! From Signals and Systems course when the content of even vs odd functions are covered we all learn these things about how symmetrical distortion is related to odd order harmonics and asymmetrical distortion is related to even order. Wonderful demonstration of this around 19:00 to 20:00 with the inclusion of diodes in half vs full clamping orientation. Check out some guitar pedal circuits that utilize this, such as the Boss OD-1 and Ibanez Tubescreamer.
I'm a audio repair guy without any real technical training and this is the first video of yours that I have seen and I LOVE it! It's helped me to understand several concepts that I've run into in dealing with the audio domain and it's amplification and manipulation on various fronts. My math capabilities are quite rudimentary but I could follow your explanation rather well and I believe that to be a good sign! For me and for you :)
So thank you for doing these. M
I have to say, you made a very good video demonstration together with the explaination. Honestly speaking, it woks much better for me to understand the theory than getting a lecture in the university.
Great presentation. Years ago, but worth every second! 73
@epatto No problem, I'll see what I can do. I have an array of new videos and tutorials which I will be doing during this month. So check back mid Feb! :)
I think the answer has to do with the odd terms for the current cancle out, since the diodes are back to back connected, so the even terms stay (and add up) and the odd terms cancle out. GREAT videos, I really enjoy them, and I hope you keep up the good work.
Excellent discussion on distortion. Really quite outstanding.
These are some of the best tutorials on RUclips. I would only suggest to use a multiple camera setup so they are also a bit shorter and more to the point. Right now there is a lot of camera movement involved breaking the flow. One or two mobile phones would do the trick, RUclips also has a video mixing capability which is pretty intuitive and fast. Keep up the nice work!
The even harmonics are attenuated because the diodes clip the signal to be more like a square wave which has only odd harmonics. Thanks for the videos SignalPathGuy, as long as you're making them I'll be watching, especially looking forward to the mixer one.
Thanks a lot for this, it really helps to see this demonstration and explanation on high quality equipment!
If you ever get around to the mixer video I'll be celebrating :)
I would love to see a tutorial on mixers!
Thank you for a great video!)
Same
Me too. Only I would like to see the tutorial as part of a practical build similar to how Ben Eater designed his Breadboard Computer ruclips.net/video/HyznrdDSSGM/видео.html.
Excellent explanation of non-linearity. Thanks. 🙏
I have been binge watching your videos. Great topics and awesome explanations. Keep up the great work!
@3cu14rs Yes, you are right about the fact that a square wave does not have even-mode harmonics. But what about at the circuit level, what is happening there?
you said that the diodes only 'kick in' after a certain point, and then the second diode is flipped in the other direction allowing the opposite voltage, so is it that the even harmonics are causing perfect phase cancellation?
I would really like to see what's inside the spectrum analyzer. If you bought it broken, the fixing would be a great story as well! Keep up the good work!
Very good explanation. Thank you for your effort
@evhgl87 Thank you! Soon there will be more videos. Please check back.
2019 watching this video and the answer for the question in the 19:50 is because symmetric and Asymmetric clipping effects diffrently on the harmonic content of signal,one causes to produce even harmonics and the other will cause to produce odd harmonics.
One small feedback. You are awesome. Thank you.
Just found your videos and I enjoy them very much! Will be going through and watching all of them! :)
Did you do tutorial on mixers?
correct, a clamper consists of and integrator with a diode in parallel with the capacitor.
The diodes are in parallel and opposite. Harmonics mean power of basic input function (third -> power of 3); power surely keeps the original sign if odd (while - * - = +). At every [half] cycle diodes have same harmonic values but all opposite in sign; so due to their positions, the difference is the result. Then even components are eliminated, while the odds...just pass (no add because diodes don't amplify).
P.S.: harmonics gain a little bit in amplitude: maybe a sort of power redistribution?
Amazing ! I'm learning so much with this nice videos !
Excellent man...your teaching style is perfect theory/practical app!..could you do non-lin in IR dect's or other sensor's? the series is great!..peace
Maybe the one-diode clamp distorts the sine asymmetrically, therefore it behaves as a kind of sum of some kind of a half rectified sine/saw plus a square wave, and a full clipped sine is symmetrical so it behaves like an almost square wave? Thanks for share all of your knowledge, Its something very Valuable nowadays!!!
Hi Shahriar Very impressive video. You explain some very complex concepts in a clear and concise manner while maintaining a presentation that is very engaging.Regarding your question I will take a stab at it. The even order products introduced by each diode are 180 degrees out of phase. So when they are summed together they cancel each other out. Am I right ?
@28:06 If the start frequency was at 0 MHz instead of 1 MHz, would we see a spike at the Beat Frequency (ie at 100KHz) and would it's amplitude be the same as that of the two tones? Thanks a lot for your videos. Much appreciated.
Excellent content.
@22:10 - What does he mean when he says that 30 dB is only 5 BITS?
What would be the simplest way to achieve the opposite? Keeping the even harmonics and attenuating the odd harmonics.
@TheSignalPathBlog A harmonic is defined as a multiple of the fundamental K. The first diode and the introduction of a sin wave creates the non-linear output due to the e^x attitude of how a non-linear diode operates (as a sharp change in definition is non-linear). Since the diodes are the same and therefore considered balanced, the even order harmonics are canceled by only allowing odd harmonics due to the diodes I to V non-linearity being closely matched during manufacturing.
Absolutely beautiful video!
How did you connect a 50Ohm SA to the 1KOhm resistor divider?
Correct If I am wrong!!! When the first diode is for all odd term is active that mean is Sin90 (+90) that mean +1 and when second one you add parallel with output is active then it get Sin(-90) that mean -1 they both cancelled each other.
ohhhhh MAN!!!! now i get ittt
wow thanks a lot for making this video. this is extremely relevant content for what im doing. cant thank you enough.
Sir any video on RF transmitter receiver module working circuit diagram for RC
Excellent , I can not thank you enough. PLEASE keep up the good work
Hello, can you provide details on the combiner you used on this video? By the way, excellent Job!
@TheSignalPathBlog Mind clarifying what you mean with circuit level?
Sorry, i dont understand, why use the buffer and what is the effect of the second diode.
Extremely interesting video, Thank you very much!
when you add one diode there comes 1st 2nd 3rd 4th....nth harmonics created but when when you add second diode you add also 1st 2nd and 3rd ...nth harmonics,,,but even(2,4,6..) harmonics produced from both diodes cancels out and odd harmonics gets added(in your spectrum analyser we seen gain of odd harmonics after two diodes addition.)
Diode switching caused the harmonics so Recovery time of the diodes... u need schotkey diodes? And maybe hi freq coupling cap?
A bit late, but I think the answer is related to the symmetry of the waveform relative to the time axis when the two diodes are present. With one diode, there is no symmetry, since the signal is just going to be clamped in one side. When two diodes are used, the signal is clamped in both sides. When calculating the Fourier series coefficients for a signal like this, the even coefficients are 0, it comes out of the calculation.
I guess it's what is happening here. When Shahriar puts the diode, it's even possible to see the even harmonics slowly going to 0. I think this is due to the temperature of the diode slowly increasing?
These videos are terrific!
2-nd diode makes the output odd symmetric, thus only odd cosine harmonics are needed.
Here' are two questions not related to the non-linear portion of your demonstration. I'm simply exploring the circuit with the mixed signals with just the resistor divider portion of the circuit.
Regarding that brass fitting which combines the output of the two signal
generators... Does that have circuitry inside? Or is that just a simple
coupler and you're merely superimposing one signal atop the other?
Now to my main question:
I'm puzzled why I only saw two beautiful pure peaks on the spectrum analyzer when you were mixing two signals through the plain vanilla voltage divider.
Forgive me if this sounds stupid. But I thought heterodyning generates sum and difference signals of two pure sine waves (let's say each with zero THD). My question is:
How come your spectrum analyzer showed only two pure peaks before you applied the clipping diodes? Why didn't I see any sum/difference signals? Shouldn't I have seen a 100 KHz peak and a 3.1MHz peak?
One step further: Let's say that they were there and that you were somehow hiding them on the analyzer to exclude those ranges... Well, once you have a 100KHz sideband signal, would that not (again) heterodyne with each of the two fundamentals, generating signals at 1.4MHz (visible), 1.5 (hidden within the fundamental peak), 1.6 (hidden again) and 1.7 (visible)? I'm genuinely troubled by seeing just two pure peaks on the spectrum analyzer.
Responders please be gentle. I understand that I don't know what I'm talking about because those two pure peaks are proof that I'm overthinking this. I'm hoping to learn why/how heterodyning is not happening with this circuit.
Hi very nice videos, thanks! I have a question for you... at time 29.24 we can se the intermodulationproducts from your two-tone test. If one do the math, the upper and lower IM-products should have the same amplitude if the amplitude of the two inputed signal is the same, right? But as i noticed all the upper IM-products have more energy then the lower IM-products, its not symmetrical. Its mostly seen when watching the IM3-product. So my question for you is why does this happen? My idea is that the two diodes are not identical, they are missmatched, so the taylor expansion for each of them wont have the same coefficients. But if that is the case, what would happend if you switch place on the diodes in the design. Would it be that all the lower IM-products will have higher amplitude? Im curious to know!
The coefficients for the FT in case of diode 1 are all positive. The coefficients for the diode 2 have the magnitude as in diode 1 but alternate in sign. When you add all the coefficients from both diodes the resulting every other peak disappears and remaining peaks double in amplitude. I think. I would have to work it out.
Excellent Video ! Keep them coming our way !
Perfect!
When you summed the exponential series with the sine wave in it, and you got the powers of the sines, you suddenly got the fourier series with sin(kw), but you didn't show how the cosine terms disappear from it.
Congratulations for the the very didactic quality of this vid! :bd
+thesignalpathblog why have you connected the two capacitors on the bread board ?? what is their use ??
And when you added diode to the circuit it was actually clipping the negative and the positive half cycles and not clamping it right ??
Those are the same thing. Clipping and clamping are the same thing in this circuit. The capacitors are just there to decouple the power supply (get rid of any noise the power supply might introduce to the circuit). They are not part of the circuit under examination.
The capacitors are not for noise from the power supply. They are providing the "pulse" power to the circuit. The leads from the power supply to the breadboard have inductance and so cannot deliver power pulses fast. The capacitors are capable of delivering those fast pulses.
like and subscribed! i want more videos :D Especially when you explain how sole circuit works . I like when you make a practical demonstration/implementation of the theory described in the video. Very educative ! Thank you!
As a math guy, I really enjoyed your video. #Subbed :)
Will there be a tutorial on mixers any time soon?
Great, thanks for this.
In the initial expression for id for the first nonlinear circuit what is Vt?
is is simply the value of Vin at a particular time t?
if that is the case, then surely Vout could never be bigger than Vt, right?
+Zambia95 Vt is thermal voltage, basically the ratio of temperature (in energy units, kT) to charge of an electron. By the way, about this equation, we're in the land of being king of sloppy (not that it's a bad thing;p) by getting (V_out/V_t)^k/k! and then summing over i ;)
Phase induced destructive cancellation of the positive and negative going harmonics...
I think that diods on the schematic are in reverse direction, but nevertheless great video!
Nice explained. Thanks.
Why you don't do a network analyzer or spectrum analyzer teardown? It probably very interesting.
Pheraph do a repair of one of them.
Nice video. You deliver well. Clear. Thanks :)
This one's amazing :-)
Quality content, thanks!
Got a circuit you wanna build? Test it out here first. Track down: 'Circuit Solver' by Phasor Systems on Google Play.
The circuit is non linear so once you have the linearized system consider the transfer function in both cases or rather in presence of diode and not. The even harmonic will disappear if and only if the transfer function is affected by pole zero cancellation.
Thanks!
good teaching
great video
Take a look at scotty sprowls project, it cover spectrum analyser operation. You can find it by asking google "msa scotty"
By putting the two diodes you are creating a simetrical square wave that cant have even harmonics and all should be zero.
@evhgl87 BTW Thank you very much for your channel and offering your knowledge to others, you are very kind and i cannot wait to learn more from you!
God bless you
the index variable of the summation should be "K" not "i"
Maybe he neglected to state i=k . Lol
Laplace transform😄
👍
Non-lineraty is bad thing why
Your a genius why aren't u using better audio recording equipment???
There is a difference between clipping and clamping. You are not clamping the signal. You are clipping the signal. FYI...
sir i am from pakistan and your video has too much knowledge but sir there is problem that you speak fluently and i can`t understand some of the points from your videos so sir please speak slowly in your new videos because sir we have language problem.