Sharing ideas on how to make a more efficent motor using Flyback (MODERATED)

[Magluvin]

That's interesting Mags

Maybe when the PWM kicks off the motor cores go above the max saturation point which would actually weakens the motors torque output.
If you can change your gear ratio so your motor can turn at a higher RPM, it may deliver max torque at full throttle.
To me it sounds like your motor is bogged down with too much of a load and that may be corrected with a gear change.

Does it make any sense to you?

Luc

Hey Luc

Below is a pic that is very close to what im running on my bike.  So there are no gears. The motor is direct drive in the wheel. 

Im also reposting the previous scope shots. The one with no lines of pwm is when the throttle is pulled all the way on. I would think that not having any off time would be more motive force in the motor or to say the most torque. But when riding and I pulled the throttle back a bit, i go faster. More go.  So Ill have to put the tiny scope on the bike and ride to see if loading the motor with me on it will still show no pwm at full on throttle and see if it kicks into pwm when I back it off for more boost.  If thats still the case when riding then I think we have something to aim for in all this. 

The pwm on and off time periods would be mostly on and very short off when just backing off of full throttle. Seems to be around 15  to 16 kz on the scope. Varies a bit for some reason. So if it is what I think then I would want to experiment with different freq also to see if there is more to get by varying that parameter.

All this time im pulling the throttle all the way back when I wanna go, burning the most current from the batteries when I could have been saving one energy use by limiting the throttle and also going faster.  Seems like a good deal to me.     So Im looking further into it.

Mags

[gotoluc]

Hey Luc

Below is a pic that is very close to what im running on my bike.  So there are no gears. The motor is direct drive in the wheel. 

Im also reposting the previous scope shots. The one with no lines of pwm is when the throttle is pulled all the way on. I would think that not having any off time would be more motive force in the motor or to say the most torque. But when riding and I pulled the throttle back a bit, i go faster. More go.  So Ill have to put the tiny scope on the bike and ride to see if loading the motor with me on it will still show no pwm at full on throttle and see if it kicks into pwm when I back it off for more boost.  If thats still the case when riding then I think we have something to aim for in all this. 

The pwm on and off time periods would be mostly on and very short off when just backing off of full throttle. Seems to be around 15  to 16 kz on the scope. Varies a bit for some reason. So if it is what I think then I would want to experiment with different freq also to see if there is more to get by varying that parameter.

All this time im pulling the throttle all the way back when I wanna go, burning the most current from the batteries when I could have been saving one energy use by limiting the throttle and also going faster.  Seems like a good deal to me.     So Im looking further into it.

Mags

Yes, that won't work on a hub motor!... unless you changed the rim to a smaller size

I guess you don't have amp metering capabilities while you drive right?... or else you would of noticed the extra current wasted?

Is your Hub motor one of those that say you can use it with 24 to 36 volt input?... what is your input voltage?
I don't believe motors can be that flexible in voltages as transportation weight changes, added wind resistance, tire pressure and all kinds of things can bog down the motor to a point the cores are over max flux capability. I think that's what you're experiencing.

So you're still thinking of taping into flyback capture?

Luc

[Magluvin]

I have a WattsUp meter that will work good for those measurements. My system is 48v and the WattsUp can do 50 amp, 100 amp peak.

You might have something with the saturation bit.  More power, less output. But maybe not.

Mags

[verpies]

If I understand the circuit correctly, the discharge is to activate a MOSFET to dump the flyback cap?

Yes, the "Discharge C" is signal is for dumping the energy accumulated in the "flyback cap" into a resistor heater or a 2^nd coil (a.k.a. "assistance coil")

... if so, why would we want to do this during Energize L on time?   Would it not always be after L is opened and flyback has charged the cap?

That's exactly what is happening with this pulse sequencer: The "flyback cap" is dumped after "L is opened and the flyback has charged the cap".
What you don't realize in this scheme, is that the flyback has already charged the cap in the previous cycle.

This way,  the energy stays in the "flyback cap" for almost the entire cycle and is dumped right before the next recharge, which facilitates good measurement of the peak voltage, when an averaging voltmeter is connected across this "flyback cap".

Of course, if you do not care about the accuracy of the voltmeter readings, you can change the time when the "flyback cap" is dumped, to any other time you want, just by rewiring this pulse sequencer a little ...as long as the "dumping" is not happening while the "flyback cap" is recharging (...like Itsu has done accidentally in the previous version).

Speaking of flyback, what will activate the MOSFET to capture flyback so I can get rid of the diode to make it more efficient?
would pin 11 be fast enough to do this?

Yes, pin 11 is fast enough. Actually it might be too fast and not give the main coil's MOSFET enough time to turn off.

The problem with substituting MOSFET for the "flyback diode" is elsewhere.  Namely for that purpose, you'd have to use 2 MOSFETs connected in series back-to-back, because of their internal body diodes. 
Providing suitable gate drive signals to such MOSFETs with non-isolated drivers and a single power supply is a challenge.

  also, keep in mind that I've got lots of MIC4451YN (attached pdf) which I think have two out that are inverted. Could this also work and should I use them to trigger all the mosfets?

Yes, big MOSFETs should not be driven directly from the 4047 chips. They should be driven thru some stronger driver chips. 

The MIC4451 drivers are suitable for this purpose but be warned that they are inverting chips, which means that all the output signals from the Pulse Sequencer should be inverted too, e.g.: as in the schematic below.

[verpies]

@Itsu

In the circuit you are using now, the "Discharge C" signal should be inverted (compared to Gotoluc's diagram), because the L2 requires a precharge period in order to build up energy in L2 which is used later to generate a gate pulse for Q3.  This gate pulse begins when the precharge period ends and your MOSFET driver interrupts the grounding current to L2 (when point F goes high).
Additionally in your circuit, the width of Q3's gate pulse is not determined by the width of the "Discharge C" signal but by the inductance of L2.

Note, that in the Pulse Sequencer circuit, that I had posted for Gotoluc, the width of the "Discharge C" signal is determined by U4, which is suitable for discharging the "flyback cap" by a relay. 
So don't be tempted to use Gotoluc's Pulse Sequencer with your motor circuit as is (without minor changes).  I will post these changes if you ever build Gotoluc's Pulse Sequencer (...or Gotoluc builds your motor circuit)