Confirming the Delayed Lenz Effect

[conradelektro]

Hi Conrad,

Would like to suggest two things.

One is using a diode D in series with the positive supply rail going to the H-bridge and connecting a puffer capacitor C across the supply rails of the H-bridge. (I modified your original schematic to show what I mean.) The reason for these small modifications is that the collapsing field of the coil in your original schematic goes back to the power supply and most probably dissipates across its inner resistance but inserting a diode prevents the 'flyback' pulse seeing the power supply and can only go into the puffer capacitor.

The second suggestion would be (after testing the first one) to use Schottky diodes in parallel with the built-in body diodes of all the 4 MOSFETs, (one diode for each MOSFET).

rgds,
Gyula

@Gyula:  the Diode in series with the poitive rail in front of the H-bridge and the electrolytic puffer capacitor as you indicated on my schematics really helped. As you said the spikes went back to my laboratory power supply.

First I only put the diode in the circuit and I saw the spikes on the scope (measured over the coil). Then I added a 4700 µF electrolytic puffer cap and I saw how it absorbed the spikes.

I still have to get a good Schottky Diode, at the moment I use a BYW29EX-200 (8A, 0.9 V forward Voltage). Since I use little power with the relay coil also a 1N4007 and a 1N4003 worked well.

Concerning your second suggestion (Schottky diodes in parallel with the built-in body diodes of all the 4 MOSFETs): My guess is that these Schottky Diodes should be able to conduct several Ampere?

The only Schottky Diode I could get easily from Farnell which allows 1A forward current and has 70 V reverse Voltage and 0.41 V forward Voltage is this one:

http://at.farnell.com/avago-technologies/hsms-2800-tr1g/diode-smd-schottky/dp/1056832RL

There seem to be a lot of Schottky Diodes for 100 V reverse Voltage and several Ampere forward current, but the forward Voltage will be 0.8 Volt.


I have the 1N5711 Schottky Diode with a very low forward Voltage of less than 0.4 Volt, but it is only for small signals (15 mA max, best for 1 mA or less).

I also have the BAT754C (200 mA, forward Voltage down to 0.2 Volt for very small signals).

Greetings, Conrad

[gyulasun]

Hi Conrad,

Thanks for the details, have you noticed any reducement in input current draw with this modification?

Diode type BYW29EX-200 are fast ones but due to the high forward current rating the voltage drop is close to a normal Si diode at the current levels your motor draws.

Will return with some diode types (Farnell is good for you or you have other order choices?)

Gyula

PS No need for high current diode types because the relay coil has its own 200-300 Ohm DC (plus its own inductive reactance) so even a some hundred mA rated diode would be good.

[conradelektro]

Hi Conrad,

Thanks for the details, have you noticed any reducement in input current draw with this modification?

Diode type BYW29EX-200 are fast ones but due to the high forward current rating the voltage drop is close to a normal Si diode at the current levels your motor draws.

Will return with some diode types (Farnell is good for you or you have other order choices?)

Gyula

PS No need for high current diode types because the relay coil has its own 200-300 Ohm DC (plus its own inductive reactance) so even a some hundred mA rated diode would be good.

@Gyula: Yes, the current draw of the circuit was reduced by a few mA (e.g. 18 mA instead of 20 mA) and the rpm went up by a few percent (e.g. 75 Hz instead of 70 Hz) for the same supply Voltage and current. Thank you for this very good suggestion, it made the circuit much safer for use with a laboratory power supply.

I would like that the circuit can handle up to 1 Amp current draw (e.g. 10 to 20 Watt) for future lower DC resistance coils.

At the moment I am building a new more solid mechanical set up with two drive coils and ball bearings which have an axle diameter of 5 mm (outer diameter 16 mm, a hint given by TinselKoala). See the attached circuit diagram and drive coil principle (at the right of the schematics). I want to reach more than 10.000 rpm (even if it means more power draw, e.g. 1 Watt to 2 Watt).

The new set up goes a bit in the direction of your suggestion to use both poles of a coil (although I am using two separate coils, as a start).

Farnell is good for me, also CONRAD. Mouser would cost me 20.-- Euro for shipping (instead of 5,40 Euro with Farnell and CONRAD).

Greetings, Conrad

[gyulasun]

The only Schottky Diode I could get easily from Farnell which allows 1A forward current and has 70 V reverse Voltage and 0.41 V forward Voltage is this one:

http://at.farnell.com/avago-technologies/hsms-2800-tr1g/diode-smd-schottky/dp/1056832RL

Hi Conrad,

Be careful with the above diode because the 1A forward current rating is valid for 1 usec pulse mode (from datasheet, page 2). But of course you may test them especially if paralleling them too...

I found these types, maybe the first would be good for your lower resistance coils and the second for the higher ones:

Schottky SB3100 100V 3A  VF=0.3V at 100mA  and 0.38V at 1A from Fig.2 data sheet:  http://www.micropik.com/PDF/SB350.pdf 
http://www.conrad.at/ce/de/product/160222/Schottky-Diode-SEMIKRON-Semikron-SB3100-Gehaeuseart-DO-201-IF-3-A-IFAV-3-A-URRM-100-V

Schottky RB160L-90TE25 90V 1A VF=360mV at 100mA and 0.6V typical at 1A  http://www.produktinfo.conrad.com/datenblaetter/125000-149999/140674-da-01-en-SCHOTTKY_DIODE_RB160L_90TE25_SOD_106.pdf
http://www.conrad.at/ce/de/product/140674/Schottky-Diode-ROHM-Semiconductor-RB160L-90TE25-Gehaeuseart-SOD-106-IF-1-A

It sounds good to me using 2 coils. In this case the input current draw will probably double for the same input voltage level if you are to compare it to the single coil case because the impedances of the coils will be in parallel i.e. the resultant impedance gets halved.

Just I am curious, whether you have taken a  scopeshot from the waveform across the coil after the modification (diode +cap)?

Thanks, Gyula

[synchro1]

One power coil counter wound in series wuth it's sister, would only need  one transistor. Like Mopozco's TROS motors.