I wonder what other "Internet people" you have seen do stuff like this? Yours is the only channel I know that takes such deep dives into understanding, simulating and testing analog electronics down to the physics. Excellent work!
Hi, Can you talk about Vusec aspect of power inductors, specifically used for pulse generation. Stored energy in the inductor. Which type of cores to use for fast saturation and fast desaturation, keeping in mind cores for low hysterysis loss.Many thanks.
Hey ! Nice video ! Since you're talking about the parameters of cores, can you add a frequency measure in the next video please ? Like the F.Bmax product, then you will be able to make measures on cores that you don't know the specs, anyway thanks for this great video !!! ;)
I would have used a second operational amplifier to buffer the VCC/2 - with an effective source impedance of 3.4 kΩ it has quite a high impedance and might shift quite a bit under different load conditions. But certainly a nice success!
Thanks FesZ for the video. May I know please how to model the current transformer with the core for example, primary winding turns Np =1 & secondary winding turns Ns =100? How to simulate the effect of fast transient current pulses (which is high frequency for example 100 kHz) and core saturation. Thanks
Thanks! This is much more more educational than a dry textbook stuff. Question: if I understand it correctly, the circuit you've built is for demonstration purposes only, right? If you have to post-process the data on a computer, integration could have been done on that computer too. Thus the actual setup should just be a sig gen, a shunt resistor, core under investigation, two coils on it and a digital scope. Am I right?
Yes, you are right - the integrator is not mandatory as long as you apply post-processing of the data; however, without it, you cannot make the fancy diagram on the oscilloscope :D
It would be interesting to repeat this while demonstrating the sensitivity to winding methods. Your example shows roughly separated primary and secondary windings. What would bi-filer winding produce. Better coupling should be apparent. Also tests at various frequencies,...??
Hi FesZ, I have a Hypothesis that the slope of the B/H curve is quite different at low excitation levels such as those near 1µV, where weak signal radio signals are. Do you have any way to measure the B/H of a core at 1µV, 100mV and 1V and see if they have different shapes. At very low signal levels, I propose the B/H curve will be pushed over to the right, or more vertical that horizontal. I think core AsubLs should be derated for very low signals and they would need more turns to get the shunt inductance up. Do you have thoughts?
Hi FesZ, I have a hypothesis, It is this, the AsubL of a core becomes non-linear at very low levels, 1microvolt (-107 dBm) or even down to -110 or -115 dBm, because the slope of the B/H curve is mostly Horizontal at low H (low signal levels) and it becomes vertical at more normal, higher levels. This leads me to think that we should derate a core's AsubL when calculating turns for very low level signals. Can you verify this. To maybe help clarify, if I put 5 turns on a core and measure the inductance at say 1V excitation level and get 70uH. Then if I measured inductance with a 1µV excitation level my hypothesis would suggest I will get much lower than 70µH. This all applies to getting the shunt reactance for a transformer correct at the lowest frequency of operation. Thanks for any light you can shed.
Hello! I guess you have a limit for measurements with this setup by the output current limit of 200mA (if it is correct) of this signal generator. So some amplification/current source may be needed for some bulkier inductor measurements. Also measurement frequency can be important. What are the safe values, how these can be calculated?
I just the gain so that the waveform fills in the entire screen without touching the edges - so to make sure I get the best use of the available resolution.
Damnit Fez, will you stop being so damned interesting. I have other things to do - but this is too bloody good and I have to watch!
I wonder what other "Internet people" you have seen do stuff like this? Yours is the only channel I know that takes such deep dives into understanding, simulating and testing analog electronics down to the physics. Excellent work!
The area formula look's like an application of Green's theorem
I'd probably smooth the data first
Cool video
This is pure gold. Very helpful knowledge.
Excellent video and tutorial!
Exhausting, but fascinating. 🙂
Hi, Can you talk about Vusec aspect of power inductors, specifically used for pulse generation. Stored energy in the inductor. Which type of cores to use for fast saturation and fast desaturation, keeping in mind cores for low hysterysis loss.Many thanks.
Hey ! Nice video !
Since you're talking about the parameters of cores, can you add a frequency measure in the next video please ? Like the F.Bmax product, then you will be able to make measures on cores that you don't know the specs, anyway thanks for this great video !!! ;)
I would have used a second operational amplifier to buffer the VCC/2 - with an effective source impedance of 3.4 kΩ it has quite a high impedance and might shift quite a bit under different load conditions.
But certainly a nice success!
Many thanks! 👍 I will try this at home 😊.
Dude you're like coming from another planet or something
Thanks FesZ for the video. May I know please how to model the current transformer with the core for example, primary winding turns Np =1 & secondary winding turns Ns =100? How to simulate the effect of fast transient current pulses (which is high frequency for example 100 kHz) and core saturation. Thanks
Thanks! This is much more more educational than a dry textbook stuff. Question: if I understand it correctly, the circuit you've built is for demonstration purposes only, right? If you have to post-process the data on a computer, integration could have been done on that computer too. Thus the actual setup should just be a sig gen, a shunt resistor, core under investigation, two coils on it and a digital scope. Am I right?
Yes, you are right - the integrator is not mandatory as long as you apply post-processing of the data; however, without it, you cannot make the fancy diagram on the oscilloscope :D
It would be interesting to repeat this while demonstrating the sensitivity to winding methods. Your example shows roughly separated primary and secondary windings. What would bi-filer winding produce. Better coupling should be apparent. Also tests at various frequencies,...??
Hi FesZ, I have a Hypothesis that the slope of the B/H curve is quite different at low excitation levels such as those near 1µV, where weak signal radio signals are.
Do you have any way to measure the B/H of a core at 1µV, 100mV and 1V and see if they have different shapes. At very low signal levels, I propose the B/H curve will be pushed over to the right, or more vertical that horizontal. I think core AsubLs should be derated for very low signals and they would need more turns to get the
shunt inductance up. Do you have thoughts?
Hi FesZ, I have a hypothesis,
It is this, the AsubL of a core becomes non-linear at very low levels, 1microvolt (-107 dBm) or even down to -110 or -115 dBm, because the slope of the B/H curve is mostly Horizontal at low H (low signal levels) and it becomes vertical at more normal, higher levels. This leads me to think that we should derate a core's AsubL when calculating turns for very low level signals. Can you verify this. To maybe help clarify, if I put 5 turns on a core and measure the inductance at say 1V excitation level and get 70uH. Then if I measured inductance with a 1µV excitation level my hypothesis would suggest I will get much lower than 70µH. This all applies to getting the shunt reactance for a transformer correct at the lowest frequency of operation. Thanks for any light you can shed.
Hello! I guess you have a limit for measurements with this setup by the output current limit of 200mA (if it is correct) of this signal generator. So some amplification/current source may be needed for some bulkier inductor measurements. Also measurement frequency can be important. What are the safe values, how these can be calculated?
FesZ, how do you know to set the ratio of gains the vertical and horizontal output to the scope? They B/H curve could look odd with the wrong gains.
I just the gain so that the waveform fills in the entire screen without touching the edges - so to make sure I get the best use of the available resolution.
Applause!