Hello, Sotiris here! 👋

 🚀Today I’m going to be talking about current measurement of a PFC inductor and we will check out a ferrite core inductor, testing it under real conditions, as usual.

Ok, you have your PFC prototype in the lab and it’s working 🤗!

👉 You have your isolated transformer and passive probe (a x100 one) to see the output voltage.

👉 You also care about the mosfet’s Vgs to see if you did a good job with signal integrity. Again, using a classic x10 passive probe.

👉 And then you want to see the ™️-trademark-™️ inductor current of the specific topology. You want to see the inductor high/low frequency ripple, verify that it is inside limits etc.

▶️ How do you do it ?  -> “Just use your current probe”!

🤔 “But, I don’t have one”. “How much for a low end one”?

👉 “A 300kHz one will go for about 500-800💲 …”

😳 Is there something cheaper? …Like a lot cheaper to use ?

▶️ That was also of my question when testing the PFC. Yeah, I could have had access to expensive toys like branded current probes, but the idea of having 2-3 current probes at the same time for tests, kept bugging me. What then? 2K worth of probes just to see some low bandwidth current signals ...

⚡ How much BW(bandwidth) do I need for a PFC inductor current signal, anyway?

   👉 Going back to the basics 

• In my case Vout=390V+-5%, so as a worst case we’ll take VL=420V and L=565uH. These number gave me a 0.75A/μs as a maximum current rate of rise.
• Ok, so if I have an inductor with a 1.5A ripple (peak-peak) the current rise or fall time will be 2μ

So from the well known:

But the current waveform in a PFC CCM inductor goes like this:

👉 We observe that the current ripple is at its max value in the middle of the half cycle and then it drops.

So for example, when the current ripple would be at 0.5A the rise/fall time would be at 0.6μs increasing the BW needed to >500kHz. The higher the ripple current value the lower the BW it requires. So low power PFCs will be harder to measure in that sense.

⚠️The point here is to know the BW you need to have to get a decent measurement! 🍺

🔑 What about a 35💲 custom probe?

❓Interested ❓

   What I built, was a very simple pcb with LA 55-P capable of making isolated measurements up to 600V at 50A with a 200kHz bandwidth. That’s the specs of this current transducer. Check the datasheet yourself. All you need to do is supply +-12V and that’s it. The only thing you may need to play with is R3. The “turns ratio” is 1:1000 with a 50mA max output. So, setting R3=100R will is translated to:

• +-50A à +-5V
• The input is just a wire passing through the hole in the transducer – completely isolated – very low capacitive coupling into the scope.

Custom +-50A current probe

Test inductor current waveforms

💥 Now it’s time for test of the week. Let’s test a ferrite core inductor 💥  

     👉 Inductor design project 😊    

▶️    How did it perform under 0.5kW PFC load for 1hr+ (natural cooling)? 🧐

🧐   We have a 30% error. This is the elephant in the room here. Having checked all the setup equipment, I can say that this how the inductor behaved. Again, as with the previous newsletter there is some research going on to find the source of the problem. 💡

🔥 Temperature rise of the magnetics is kind of the holy grail to get right. All parameters come into play and finally we end up with this one number❗ So it’s understandable to me, that there will be prediction errors as our models evolve and account for more parameters. 🔥  

👉 If you have questions or an idea let me know at sotiris.zorbas@frenetic.ai

WHAT’S COMING NEXT WEEK?

👉 Next week focus is going to be material selection. Now these tips are meant to raise some red flags when you make your own design. The usual tips and tricks will be there as well!

We have a technology that allows you to go from concept design to manufacturing in a few weeks. We will be showing it in PCIM 2022 next week (Hall 6, Stand 356)

You can have a personalized demonstration if you hurry up. 

Here is the link: Optimize the process of your power electronics

❌ If you like to optimize only losses, Frenetic is not appropriate for you. 

✅ If you want to optimize the complete process, You are Frenetic. 

 👋  Until next week 👋 

Sotiris Zorbas, MSc 

Power Εlectronics Εngineer