I'm embarrassed to ask, because I've been around here long enough that I should already know. I just kept assuming I would stumble across the information if I waited long enough.
Do any sensorless controllers have regen?
- Thread starter spinningmagnets
- Start date
rojitor
100 kW
I am pretty sure lyen sold sensorless controllers with regen.
https://endless-sphere.com/forums/viewtopic.php?f=31&t=17683
I think he sold that model sensored and sensorless. It has regen.
https://endless-sphere.com/forums/viewtopic.php?f=31&t=17683
I think he sold that model sensored and sensorless. It has regen.
rojitor
100 kW
Paul is selling them:
https://em3ev.com/shop/12-fet-irfb4110-infineon-sensorless-controller/
https://em3ev.com/shop/12-fet-irfb4110-infineon-sensorless-controller/
Thanks
Yes. Both of the cheap 12FETs on my trike do. One was sent by Dogman near the beginning of the SB Cruiser thread,
https://endless-sphere.com/forums/viewtopic.php?f=2&t=67833&start=25#p1025351
http://www.ebay.com/itm/271700702048?_trksid=p2060353.m2749.l2649&ssPageName=STRK%3AMEBIDX%3AIT
and the other I've forgotten now where it came from but it's origin is probably also posted in that thread, or CrazyBike2's. (I seem to vaguely recall it was sent to me while I was at Bill's after the fire in 2013).
The one from Dogman actually has better than plain regen, it has "EABS", "active braking", that actively powers the wheel against it's direction of rotation to force it to slow down faster. This uses power rather than recovering any but it works to a lower speed and it works harder all teh way down to that speed. With the X5304 on there it would brake hard enough to brake-steer the trike to the right that it requires significant steering input to compensate for it. It's not as hard with only the HSR3548 in it's place.
The other one is just regen, but it's a two-stage regen. Above a certain speed it's kinda weak, but once it drops below a certain speed it grabs harder. Neither level is spectacular, but it's there, and on a regular bike would work just fine.
Unfortunately neither controller actually uses it's hall wire inputs, even though they both have hall wires installed and all the motors tested with them have working halls. They only work sensorless, rather than auto switching between modes, or when the "learn" wires are used, etc. Annoying.
And the first one above has all hall and phase wires as green wire, all the same color. Guess it's their way of preventing you from bothering to try different combos, and forcing you to use the learn wires? Or just cheaper to make with only one spool of wire needed, rather than three?
https://endless-sphere.com/forums/viewtopic.php?f=2&t=67833&start=25#p1025351
http://www.ebay.com/itm/271700702048?_trksid=p2060353.m2749.l2649&ssPageName=STRK%3AMEBIDX%3AIT
and the other I've forgotten now where it came from but it's origin is probably also posted in that thread, or CrazyBike2's. (I seem to vaguely recall it was sent to me while I was at Bill's after the fire in 2013).
The one from Dogman actually has better than plain regen, it has "EABS", "active braking", that actively powers the wheel against it's direction of rotation to force it to slow down faster. This uses power rather than recovering any but it works to a lower speed and it works harder all teh way down to that speed. With the X5304 on there it would brake hard enough to brake-steer the trike to the right that it requires significant steering input to compensate for it. It's not as hard with only the HSR3548 in it's place.
The other one is just regen, but it's a two-stage regen. Above a certain speed it's kinda weak, but once it drops below a certain speed it grabs harder. Neither level is spectacular, but it's there, and on a regular bike would work just fine.
Unfortunately neither controller actually uses it's hall wire inputs, even though they both have hall wires installed and all the motors tested with them have working halls. They only work sensorless, rather than auto switching between modes, or when the "learn" wires are used, etc. Annoying.
And the first one above has all hall and phase wires as green wire, all the same color. Guess it's their way of preventing you from bothering to try different combos, and forcing you to use the learn wires? Or just cheaper to make with only one spool of wire needed, rather than three?
They should. Regen doesn't need hall (or cos/sin) sensors to work.
I'm not 100% sure, but regen is basically this, the controller will short all 3 phase wires together, it will then run that through a rectifier setup, and back into the battery. Then there will be extras added to this circuit for controlling the amount of regen current, for making it variable, etc... I believe PWM circuits are used for that, but I haven't looked into that aspect of regen yet.
I'm not 100% sure, but regen is basically this, the controller will short all 3 phase wires together, it will then run that through a rectifier setup, and back into the battery. Then there will be extras added to this circuit for controlling the amount of regen current, for making it variable, etc... I believe PWM circuits are used for that, but I haven't looked into that aspect of regen yet.
If you short all phase wires together, current does not flow out of the motor, it recirculates only within the motor (and phase wires) and/or is cancelled out between phases, and the motor heats up as it internally dissipates all the power of stopping the vehicle (just like a mechanical brake). There's nothing to run thru a rectifier. it's not regen braking, it's more commonly called plug braking.
Regen braking in a cotnroller can be done primitively by turning the FETs on in a way that creates a 3-phase rectifier, but this only does anything at all if the generated voltage is higher than the battery voltage, meaning the motor is spinning fast enough. As soon as the speed drops enough to drop the voltage below the battery voltage, no current flows and no braking happens.
More commonly the controller switches the FETs in a way that causes the current to flow differently in the motor coils, and voltage to boost momentarily and repeatedly above the battery voltage, so that braking and regen continues down to a lower speed than is possible with simple rectification.
There's a few threads around that discuss details of various kinds of regen, but you'll have to poke around a bit to find them.
this one has some info
https://endless-sphere.com/forums/viewtopic.php?f=2&t=7891
There's others in this search
https://endless-sphere.com/forums/search.php?keywords=regen*+brak*&terms=all&author=&sc=1&sf=titleonly&sk=t&sd=d&sr=topics&st=0&ch=300&t=0&submit=Search
but most of them are not relevant.
Regen braking in a cotnroller can be done primitively by turning the FETs on in a way that creates a 3-phase rectifier, but this only does anything at all if the generated voltage is higher than the battery voltage, meaning the motor is spinning fast enough. As soon as the speed drops enough to drop the voltage below the battery voltage, no current flows and no braking happens.
More commonly the controller switches the FETs in a way that causes the current to flow differently in the motor coils, and voltage to boost momentarily and repeatedly above the battery voltage, so that braking and regen continues down to a lower speed than is possible with simple rectification.
There's a few threads around that discuss details of various kinds of regen, but you'll have to poke around a bit to find them.
this one has some info
https://endless-sphere.com/forums/viewtopic.php?f=2&t=7891
There's others in this search
https://endless-sphere.com/forums/search.php?keywords=regen*+brak*&terms=all&author=&sc=1&sf=titleonly&sk=t&sd=d&sr=topics&st=0&ch=300&t=0&submit=Search
but most of them are not relevant.
Similar threads
