I like the way you explain things, combining theory and practical demos - that's exactly how I learn best. W2AEW does that too, he's another good RUclips electronics educator - thankfully you are mainly covering different topics!
i did considerable EMI tesdting a few decades ago on switch mode power supplies and found the layout of the PCB had a major impact on the EMI produced by a power supply.
Thank you for the wonderful video because spectrum analyzers are very expensive so this is more accessible to me. Can you make a video on how to shield the components with low cost materials as much as possible without redesigning the pcb?
I cannot comment on the usage of a Tiny SA since I never used one.. maybe I'll try to acquire it in the future. Regarding "calibrated readings" - even with a commercial probe - even if the response is more reliable - there is a better correlation between field strength and measured value, the nearfield measurement in general is highly position dependent - you move the probe a bit to the side and the measurement completely changes.
Great vid! You are a natural talent in explaining how things work. Question: The spikes you measure in the near-field, are they also present, say, 10 meters away as RF? Is there a voldtage threshold that needs to be exceeded for near-field to also radiate out in far-field?
If the near filed noise will be visible in the far field has a lot to do with the casing and the overall device construction. Usually its good practice to measure using far field methods to observe the actual problems and then use near filed to find the sources of the problems on the device; going the other way is not that easy since there is no "threshold" or obvious link.
Do you need actually need to measure radiated emissions at such low frequencies? Anyway, at this frequency most of the radiation will come from the cables connected to the circuit rather than the circuit its self; maybe just magnetic fields coming from some poorly shielded inductors could cause problems.
In my experience, this will work only under very limited circumstances. You can inject noise into small parts of a circuit using near field probes, but you can't generate noise over the entire circuit usually.
Hello, FeeZ Electronics! I like your tutorials about the simulator S.P.I.C.E. LTspice, and all your educational content abount analogic electronics.Okay.I have some questions for you: 1.What are the limitations in analysing numerically calculations of program LTspice in the matter of analogic and digital electronics? 2.From your professional point about modelling and simulation S.P.I.C.E. electrical and electronic behavioural predictions, if you made a comparasion between programs LTspice VIII and Micro-CAP12,which are difference between them? 3.I want to know how can I learn better analog electronics? Thank you, Dragos.
Hello Dragos! 1 - For analog electronics the main limitation on accuracy (in any simulator) refers to how accurate the simulation model is - how much of the parasitics and non-ideal behaviors are simulated. Also in HF circuits modeling parasitics of the PCB starts to be a problem. For digital electronics, the main issue is getting models that represent the various components - I mean a logic gate is easy to find, but a processor or something similar is a completely different story. 2 - I never used MicroCap but I guess I should try it. In general any spice simulator (LTspice, OrCad Spice, MicroCap etc) should be just as good. Simulation speed might be different based on algorithm implementations but as long as things are set up in the same way, the results should be identical. By setup I don't just mean the simulation circuit and models, I also mean simulator parameters - accuracy, number of digits used, extra default parasitics etc. 3 - Practice. Simulate then build circuits to try and understand how they work. Once you think you understand the circuit, try to improve it - that will confirm correct understanding. Hope this helps! Fesz
Hello Fesz, Thank you for the video. I was wondering about how can we use the results we get from this type of pre-compliance testing, I mean they are basically qualitative and not quantitative tests, so how can we translate what we got from them to the far-field tests and know for example, whether a spike should be an alarm that we will fail EMC around that frequency or not. Edit: Another question, do you have any commercial E-field probes that you can recommend ?
In my experience, the use of near-field probing is to optimize designs on the bench (see if variant A is more noisy than variant B); the other use case is for debugging after a first round of EMC measurements in the lab - for example the unit is generating a spike at whatever frequency, then the exact location on the PCB can be determined using the near-field probe; subsequently any optimizations can evaluated in reference to the initial board. You can make the assumption that if you fail by 10dB and the optimizations you make are reducing noise by more than 10dB at the problem frq you should be fine, but this is very relative. Probing alone will not be enough to ensure that a qualification is passed but it will help to evaluate solutions on the bench rather than in the lab. I don't know what to recommend, it depends on your budget - probes will go from a couple hundred usd to a few thousand.
Most likely yes; the main advantage of the near field probe is that they will show you the exact emission point - the probe needs to be very close to the noise source to pick it up, otherwise you will not see the spike. The far field measurement picks up the general noise output of the complete unit, with near field probes you can usually localize the exact circuit/component responsible.
@@FesZElectronics thank you for you insight. I did some near field probing recently and came to the conclusion that the FFT on my cheap Rigol scope isn't really up to the task. I'll have to look for a proper spectrum analyzer.
I'm not sure if there is a limit. But on one side, with increasing frequency, its more and more difficult to stay in the "near field" - at 6GHz the wavelength is 5cm so the near field ends at around 0.79cm. So its difficult to measure the magnetic and electric fields separately. On the other hand, a 6GHz spectrum analyzer is really expensive, and anything with more BW is even more expensive.
I like the way you explain things, combining theory and practical demos - that's exactly how I learn best. W2AEW does that too, he's another good RUclips electronics educator - thankfully you are mainly covering different topics!
Look for IMSAIGuy, he's pretty good at explaining things too.
I know these things in a general way but your examples make things crystal clear -- many thanks.
Thank you for the wonderful video! Keep up the great work!
i did considerable EMI tesdting a few decades ago on switch mode power supplies and found the layout of the PCB had a major impact on the EMI produced by a power supply.
I came to a similar conclusion - a "good" schematic is never enough, a big contribution to a design comes from the layout.
Excelente vídeo...!! Gracias por compartir...!! Espero con ansias el siguiente vídeo...
one of the bests electronics channel! thanks!
Love your channel. The pace is ideal.
Thanks so much!
Great video and explanations, thank you
amazing approach to the topic, where did you study electronics?
He is a guy.
Thank you for the wonderful video because spectrum analyzers are very expensive so this is more accessible to me. Can you make a video on how to shield the components with low cost materials as much as possible without redesigning the pcb?
@FesZElectronics
Great demo and easy to follow explanation. 🎉 How about a video using Tiny SA. Does commercial probes provide a calibrated reading ?
I cannot comment on the usage of a Tiny SA since I never used one.. maybe I'll try to acquire it in the future. Regarding "calibrated readings" - even with a commercial probe - even if the response is more reliable - there is a better correlation between field strength and measured value, the nearfield measurement in general is highly position dependent - you move the probe a bit to the side and the measurement completely changes.
Thanks. Love your videos.
Awesome video. What's the brand of oscilloscope are you using?
Its a Hantek DSO5072; but most modern day DSO oscilloscopes have the same functions implemented.
great video again!
Glad you enjoyed!
Great vid! You are a natural talent in explaining how things work. Question: The spikes you measure in the near-field, are they also present, say, 10 meters away as RF? Is there a voldtage threshold that needs to be exceeded for near-field to also radiate out in far-field?
If the near filed noise will be visible in the far field has a lot to do with the casing and the overall device construction. Usually its good practice to measure using far field methods to observe the actual problems and then use near filed to find the sources of the problems on the device; going the other way is not that easy since there is no "threshold" or obvious link.
Thanks. Yes, I agree. Reason I asked is that radiated low frequencies, say 1-100 kHz, are not so easy to measure using an antenna 10 m away.
Do you need actually need to measure radiated emissions at such low frequencies? Anyway, at this frequency most of the radiation will come from the cables connected to the circuit rather than the circuit its self; maybe just magnetic fields coming from some poorly shielded inductors could cause problems.
Nice Video ! Can this E field probe be used to check the radiated immunity performance of PCB ?
In my experience, this will work only under very limited circumstances. You can inject noise into small parts of a circuit using near field probes, but you can't generate noise over the entire circuit usually.
Thank you!
Hello, FeeZ Electronics!
I like your tutorials about the simulator S.P.I.C.E. LTspice, and all your educational content abount analogic electronics.Okay.I have some questions for you:
1.What are the limitations in analysing numerically calculations of program LTspice in the matter of analogic and digital electronics?
2.From your professional point about modelling and simulation S.P.I.C.E. electrical and electronic behavioural predictions, if you made a comparasion between programs LTspice VIII and Micro-CAP12,which are difference between them?
3.I want to know how can I learn better analog electronics?
Thank you,
Dragos.
Hello Dragos!
1 - For analog electronics the main limitation on accuracy (in any simulator) refers to how accurate the simulation model is - how much of the parasitics and non-ideal behaviors are simulated. Also in HF circuits modeling parasitics of the PCB starts to be a problem. For digital electronics, the main issue is getting models that represent the various components - I mean a logic gate is easy to find, but a processor or something similar is a completely different story.
2 - I never used MicroCap but I guess I should try it. In general any spice simulator (LTspice, OrCad Spice, MicroCap etc) should be just as good. Simulation speed might be different based on algorithm implementations but as long as things are set up in the same way, the results should be identical. By setup I don't just mean the simulation circuit and models, I also mean simulator parameters - accuracy, number of digits used, extra default parasitics etc.
3 - Practice. Simulate then build circuits to try and understand how they work. Once you think you understand the circuit, try to improve it - that will confirm correct understanding.
Hope this helps!
Fesz
❤︎
Hello Fesz,
Thank you for the video. I was wondering about how can we use the results we get from this type of pre-compliance testing, I mean they are basically qualitative and not quantitative tests, so how can we translate what we got from them to the far-field tests and know for example, whether a spike should be an alarm that we will fail EMC around that frequency or not.
Edit: Another question, do you have any commercial E-field probes that you can recommend ?
In my experience, the use of near-field probing is to optimize designs on the bench (see if variant A is more noisy than variant B); the other use case is for debugging after a first round of EMC measurements in the lab - for example the unit is generating a spike at whatever frequency, then the exact location on the PCB can be determined using the near-field probe; subsequently any optimizations can evaluated in reference to the initial board. You can make the assumption that if you fail by 10dB and the optimizations you make are reducing noise by more than 10dB at the problem frq you should be fine, but this is very relative.
Probing alone will not be enough to ensure that a qualification is passed but it will help to evaluate solutions on the bench rather than in the lab.
I don't know what to recommend, it depends on your budget - probes will go from a couple hundred usd to a few thousand.
@@FesZElectronics A great answer as always, Thank you.
let's say far field measurements come back with a high spike at 200 MHz. Will the near field probes (E or H) also see a spike at 200 MHz ?
Most likely yes; the main advantage of the near field probe is that they will show you the exact emission point - the probe needs to be very close to the noise source to pick it up, otherwise you will not see the spike. The far field measurement picks up the general noise output of the complete unit, with near field probes you can usually localize the exact circuit/component responsible.
@@FesZElectronics thank you for you insight. I did some near field probing recently and came to the conclusion that the FFT on my cheap Rigol scope isn't really up to the task. I'll have to look for a proper spectrum analyzer.
That cheapie Hantek looks better than my crappy Rigol DS1102E.
what is the maximum frequency the E-field and H-field probes can scan? is it 6GHz or more
The ones I built, I have no idea, but probably not that much. On the other hand, commercial versions are available that go up to 6GHz or more.
@@FesZElectronics i'm asking in general
I'm not sure if there is a limit. But on one side, with increasing frequency, its more and more difficult to stay in the "near field" - at 6GHz the wavelength is 5cm so the near field ends at around 0.79cm. So its difficult to measure the magnetic and electric fields separately. On the other hand, a 6GHz spectrum analyzer is really expensive, and anything with more BW is even more expensive.
@@FesZElectronics ya now clear about it thank you
@@FesZElectronics is it 0.79cm or 0.79m