Find the Q-factor for coils and understand weak power transfer

I built coils for wireless power transfer. Found inductance with a LRC meter. Put capacitors, hoping for a resonant frequency at 300kHz. Student assisted me with a NVA machine where a resonant frequency 800kHz was found. It plots power transfer over frequency. My capacitor values were pretty off. Needed a .9micro farad and used a 1micro farad. Still a large shift from 300 to 800kHz. It had like 80% transfer over 2cm distance. The yellow line is S21 at 800kHz.

I also purchased wurht coils https://www.we-online.de/katalog/datasheet/760308111.pdf. Saw a high Q factor around 200-300kHz. Placed capacitors for 270kHz resonance. And they worked best at 303kHz. It lit an led from centimeters away!.

When I place a led on my coil design it doesn't light at all. My coils show power transfer on the NVA machine yet can't light up a led. Wurth coils work great wirelessly lighting up a led.

Could it be the Q facotr? Never payed attention to Q until I bought the wurth coils which inside their data sheet show high Q as a function of frequency.

Looked up the equitation for Q and got Q series = 2 pi f L / R. This graph should look liner with frequency. The Q data graph looks more complicated(maybe R changes with f). I'm lost. Why don't my coils light up a led, even with them almost touching(NVA says 95 power transfer). Also how would I calculate Q and better understand Q for my coils. I believe either the NVA machine was wrong or my S21 has a power sent so week, that my coils could not light a led. Maybe if I run my design at resonance and high Q value. Any tips or resources would help.

• If you have finished with this question you should mark the answer as "accepted" as per this. This tells other people that the chosen answer is reliable. – Andy aka Sep 4 '20 at 8:50

Any tips or resources would help.

The Wurth coil as sold by Farnell: -

What do you think that the black support sheet is made from?

Answer - it's partially made of ferrite material and, it will highly concentrate the fields from your transmit coil and collect more energy in effect. If you want to prove this, get a magnet and feel the attraction force. It should be maximized at around the centre of the plate i.e. the ferrite is probably a disc placed centrally to attract the AC magnetic fields through the centre of the coil arrangement. You want magnetic flux to be drawn through the centre.

The 760308111 is a wireless power charging Transmitter Coil with pot core results in limiting the stray field in the design. This WE-WPCC coil features high permeability shielding coil

It's not great or perfect language but, I think you can see that it uses ferrite material to enhance the ability of the basic coil. The data sheet also says this: -

Direct mechanical impact to the product shall be prevented as the ferrite material of the core could flake or in the worst case it could break

Notice the words core and ferrite!

It has no ferrite material to enhance its ability to collect energy - that's your first problem and it's a show-stopper in my opinion. Of course, if your receive coil and transmit coil were equal size then it's less of a problem.

The second problem is that it only has 3 turns and is therefore not resonant at the same low frequency compared to the Wurth coil.

If you then made your transmit coil resonant at the higher frequency (to make it compatible with the resonant frequency of your homespun receive coil), there will be less current in the transmit coil because its impedance will be higher. Therefore less energy transferred. OK, if you are series resonating the transmit coil this may be less of a problem.

Given that this type of arrangement is a transformer, having fewer turns on the secondary also means less voltage produced by the receive coil.

In all of these types of experiment, you should be prepared to re-tune with capacitance and tweak the driving frequency and keep checking this. You only really need to look for voltage amplitude on the output coil so trying to measure S parameters is missing the direct point. Use a signal generator and use an oscilloscope to set it up, then use your other equipment to make precise measurements.

I might add that in all the power transfer coils I've built I only ever used a signal generator and oscilloscope and I got great results and built probably about ten systems for customers that are still in use today.

• You might need more turns but, I would advise you to get some ferrite material and be prepared to increase the number of turns on your receive coil. You have the facts in front of you - the Wurth device works and your homespun one doesn't. – Andy aka Aug 9 '20 at 13:42
• If you research MPT you will find this can be done where the gap can transfer power almost up to coil diameter over 11kW in 3rd Gen devices. However your load must match the transferred impedance exactly and then you only get 50% efficiency at MPT. With reduced load, greater efficiency. So choosing resonant frequency must also factor impedance of √(L/C)=Zo then mutual coupling L dominates the equation. – Tony Stewart EE75 Aug 9 '20 at 16:07
• I only saw two transmit windings (as per what I said to Tony about keeping the number of turns down) - your transmit coil has 10 or more turns and inductance rises with turns squared hence, the inductance will be high and it needs much more voltage to get current flowing. You'll also get far better results with ceramic capacitors rather than polyester. The better the dielectric (NP0 is the best), the more resonance you get and the more transmit current you get. It didn't cover what operating frequency it used so I can't add much more. – Andy aka Aug 9 '20 at 20:45
• Try to aim for about 1 or 2 uH is my advice. I'm sure there are some on-line calculators that will give you theoretical numbers. Higher inductance means a higher drive voltage. – Andy aka Aug 9 '20 at 21:36
• Remember the ferrite in the Wurth coil "draws in" much more flux than your receive coil. Anyway, we have to draw this discussion to a conclusion because we'll get told off for extended comments. – Andy aka Aug 9 '20 at 21:44