Thank you for the walk through of your build! The passive components I've used in power applications almost always need their own snubbing circuit to dampen their self-resonance. At first I was reluctant to add them but after I saw how well they performed I was less reluctant. I apologize for the terrible pun.
What exactly would you be interested in? I usually do not cover in too much detail how I build things since I don't know how to make it interesting... rather I focus on the end result
It would also be interesting as a project to make an In Line filter like Line Reactor DV/DT as surge protection and filter like the ZeroSurge and Surgex products. Especially since the patent is expired and available and it could be a good opportunity to make an excellent project and to learn. Thank you!
I think that the way your coax cables are 'peeled back' and connected to your PCB (at 10:12) isn't conducive to making accurate measurements with your SA. Ideally you'd want as little of the centre core exposed as possible. I think using a u.FL connector and a cable assembly would be better. I often have to probe around circuits with a make-shift cable probe attached to an SA but I have probably no more than 1mm exposed at the end. All the rest of the cable remains at 50R. But in your case a u.FL connector and a proper cable could be used as this is a permanent assembly, not just some random probing around here and there.
I agree with you, having part of the cable exposed is an impedance discontinuity that is not helping.. But if I would build this project again from scratch, rather than using the u.FL connector, I would skip the cable altogether - so have both the tested line connectors and the BNC mounted directly on a pcb together with the handful of necessary components. A new box would be needed, but it would make the circuit and wiring far better controlled.
Firstly thank you for all your efforts to make these videos, I learned most of the electronics from you only. I have a question. Is it possible to build a LISN for the load specification (800V DC, 60A, and 250kW). Please help me out.
This should not be a limitation - you need to make sure the series inductors are not heating up and have a stable inductance at the required 60A; and that all filter capacitor are rated above 800V; finally, what I saw in some LISN designs is the addition of a transient suppressor and attenuator on the 50R signal output; this bit needs to be built so that you never send too large signals to the analyzer and break it.
To be honest, I'm not sure what type it is. It was a random capacitor I already had, so I don't know the exact order code. Is there any specific marking present on capacitors to indicate the material?
@@FesZElectronics Now you got me xD I really don't have much experience with film caps. I know PET have MKC or MKT marking, PP on the other hand have MKP code. I think all X2 caps I saw were PP, don't take it as a rule. For the small packages I have no idea if there is any other shorter marking.
Now that you mention it, I looked trough my box of capacitors and indeed some of the larger ones have an "MKS 4" marking on them relating to the series I guess (they are WIMA capacitors); but the ones that I used in the LISN have no such marking - they only have manufacturer/capacitance/voltage rating/date code? = WIMA/0.1/100-/F6
Single channel filters was my first thought. Magnetic shielding is hard beyond -40dB. Tin plated steel would be better, as would separate boxes, don't be focused on looks. There is nothing wrong with using coaxial connectors for DC. Type N or the old UHF pl259 would do. Electrostatic shielding is easier than magnetic shielding. Inductors and wires couple less at 90° angles, so starting out with parallel inductors is a losing proposition. Seeing a face is not needed for explaining tech stuff.
I agree with the single channel approach - I think that would be best while keeping the construction simple; I know this is quite common in commercial application. Regarding the DC trough coax connectors, I find this approach a bit odd - maybe I did not understand what you where suggesting; I mean from the LISN to the power supply sure - it will help to reduce injected noise; but from the LISN to the test circuit, you want to use realistic wires, using coax will change the line to line impedance as well as the line to gnd impedance. Maybe you can give some more details on this approach and your practical experience.
I’d suspect those open-core inductors before anything else. Didn’t know you could buy ferrite sheet, try wrapping those inductors with the ferrite sheet, maybe a flux band too. Or use based toroids for everything.
![EMC tutorials - [1/3] Building an LISN](http://i.ytimg.com/vi/OEvkRW5vZNA/mqdefault.jpg)
![EMC tutorials - [1/3] Building an LISN](/img/tr.png)
It is so helpful to watch your process of figuring out what helps and what does not. Thank you.
Thank you for the walk through of your build! The passive components I've used in power applications almost always need their own snubbing circuit to dampen their self-resonance. At first I was reluctant to add them but after I saw how well they performed I was less reluctant. I apologize for the terrible pun.
Thank you for the walk through - fantastic!
It would be great if you could provide some details about how you built it 😊
What exactly would you be interested in? I usually do not cover in too much detail how I build things since I don't know how to make it interesting... rather I focus on the end result
It would also be interesting as a project to make an In Line filter like Line Reactor DV/DT as surge protection and filter like the ZeroSurge and Surgex products.
Especially since the patent is expired and available and it could be a good opportunity to make an excellent project and to learn. Thank you!
I think that the way your coax cables are 'peeled back' and connected to your PCB (at 10:12) isn't conducive to making accurate measurements with your SA. Ideally you'd want as little of the centre core exposed as possible. I think using a u.FL connector and a cable assembly would be better. I often have to probe around circuits with a make-shift cable probe attached to an SA but I have probably no more than 1mm exposed at the end. All the rest of the cable remains at 50R. But in your case a u.FL connector and a proper cable could be used as this is a permanent assembly, not just some random probing around here and there.
I agree with you, having part of the cable exposed is an impedance discontinuity that is not helping.. But if I would build this project again from scratch, rather than using the u.FL connector, I would skip the cable altogether - so have both the tested line connectors and the BNC mounted directly on a pcb together with the handful of necessary components. A new box would be needed, but it would make the circuit and wiring far better controlled.
very well done as usual !
Firstly thank you for all your efforts to make these videos, I learned most of the electronics from you only. I have a question.
Is it possible to build a LISN for the load specification (800V DC, 60A, and 250kW). Please help me out.
This should not be a limitation - you need to make sure the series inductors are not heating up and have a stable inductance at the required 60A; and that all filter capacitor are rated above 800V; finally, what I saw in some LISN designs is the addition of a transient suppressor and attenuator on the 50R signal output; this bit needs to be built so that you never send too large signals to the analyzer and break it.
@@FesZElectronics Thank you so much for the reply, got some clarity 🫡
It looks like you printed a revised core for the inductor. Did you update your Ultimaker Thingiverse files?
I don't think I made any further modifications to the design. What is the difference that you noticed?
Did you use PP or PET film capacitor? Normally PP have better high frequency response.
To be honest, I'm not sure what type it is. It was a random capacitor I already had, so I don't know the exact order code. Is there any specific marking present on capacitors to indicate the material?
@@FesZElectronics Now you got me xD I really don't have much experience with film caps. I know PET have MKC or MKT marking, PP on the other hand have MKP code. I think all X2 caps I saw were PP, don't take it as a rule. For the small packages I have no idea if there is any other shorter marking.
Now that you mention it, I looked trough my box of capacitors and indeed some of the larger ones have an "MKS 4" marking on them relating to the series I guess (they are WIMA capacitors); but the ones that I used in the LISN have no such marking - they only have manufacturer/capacitance/voltage rating/date code? = WIMA/0.1/100-/F6
Single channel filters was my first thought.
Magnetic shielding is hard beyond -40dB.
Tin plated steel would be better, as would separate boxes, don't be focused on looks.
There is nothing wrong with using coaxial connectors for DC. Type N or the old UHF pl259 would do.
Electrostatic shielding is easier than magnetic shielding. Inductors and wires couple less at 90° angles, so starting out with parallel inductors is a losing proposition.
Seeing a face is not needed for explaining tech stuff.
I agree with the single channel approach - I think that would be best while keeping the construction simple; I know this is quite common in commercial application.
Regarding the DC trough coax connectors, I find this approach a bit odd - maybe I did not understand what you where suggesting; I mean from the LISN to the power supply sure - it will help to reduce injected noise; but from the LISN to the test circuit, you want to use realistic wires, using coax will change the line to line impedance as well as the line to gnd impedance. Maybe you can give some more details on this approach and your practical experience.
@@FesZElectronics also look up MIL-STD-461F.
The impedance for a LISN flatten out at 50 ohms, above 1 MHz , so using 50 ohm connectors won't hurt.
I’d suspect those open-core inductors before anything else. Didn’t know you could buy ferrite sheet, try wrapping those inductors with the ferrite sheet, maybe a flux band too. Or use based toroids for everything.
thank