Lebowski
10 MW
building a triple stator axial flux at the moment 
-
Howdy! we're looking for donations to finish custom knowledgebase software for this forum. Please see our Funding drive thread
The true sine-wave (vectored) controller & electronics set
- Thread starter Adaptto E-Drives Lab
- Start date
KF, thanks, reading now..
Lebowski, you got a thread for that baby?!
In the meantime check this out: http://www.propulsionmarine.com/electric/10kw-water-cooled-electric-drive
It looks like the big motor is the PMS 150 or 156 in the top "comparison" pic..but what I'd like you guys to take a look at is the pic at the bottom of that page, the one that shows the magnet rotor.
Just by looking at it, would you say these PMS motors are meant for sin or trap commutation?
Edit: here is that pic and the one of the liquid cooling full size:
Lebowski, you got a thread for that baby?!
In the meantime check this out: http://www.propulsionmarine.com/electric/10kw-water-cooled-electric-drive
It looks like the big motor is the PMS 150 or 156 in the top "comparison" pic..but what I'd like you guys to take a look at is the pic at the bottom of that page, the one that shows the magnet rotor.
Just by looking at it, would you say these PMS motors are meant for sin or trap commutation?
Edit: here is that pic and the one of the liquid cooling full size:
Attachments
Hi Zom! You gotta come check out the track sometime. The nice thing is it is made on a mound of sand so it is dry within a day of a hard rainfall! And it's not just sand, it's black and a bit sticky so it doesn't push around under thin tires like you would expect, but if you fall into it, your.. say, elbow.. will go into it and save you (me, several times!) from injury.
Did I read "CNC spindle"? What are you working on?!
Later,
Jay
Did I read "CNC spindle"? What are you working on?!
Later,
Jay
They seem to have the controller up on their store now, but it's just a pre-order: http://velomastera.ru/shop/show_cat.php?catid=5&grid=1
Looks like the final cost is about $800 USD (24,000 rubles). Steep, but it's a VFC... and with continuous regen and throttle kick-down and an integrated BMS and one-stop control of all those limits... mmyes, it's what I need!
I can see it now. A dude on a Harley with Screaming Eagles is idling next to me at a light... he begins laughing, right as the light turns green... I'm in low-power mode! I'm only pushing 3kw! But my 6kw motor can take even greater peak power, and my A123 pack can provide it! I twist the throttle all the way down! TWELVE KILOWATTS push me forward as I squeal ahead, thanks to my superior torque response and power/weight ratio!
I win, and not just because I'm completely silent and using only 20-60 wh/mile!!
Looks like the final cost is about $800 USD (24,000 rubles). Steep, but it's a VFC... and with continuous regen and throttle kick-down and an integrated BMS and one-stop control of all those limits... mmyes, it's what I need!
I can see it now. A dude on a Harley with Screaming Eagles is idling next to me at a light... he begins laughing, right as the light turns green... I'm in low-power mode! I'm only pushing 3kw! But my 6kw motor can take even greater peak power, and my A123 pack can provide it! I twist the throttle all the way down! TWELVE KILOWATTS push me forward as I squeal ahead, thanks to my superior torque response and power/weight ratio!
I win, and not just because I'm completely silent and using only 20-60 wh/mile!!
24-98v so they are going to rate their controller 2v away from the limit of the irfb4468??? Scary stuff 10awg wires might be OK for interment use....xenodius said:They seem to have the controller up on their store now, but it's just a pre-order: http://velomastera.ru/shop/show_cat.php?catid=5&grid=1
Looks like the final cost is about $800 USD (24,000 rubles). Steep, but it's a VFC... and with continuous regen and throttle kick-down... mmyes, it's what I need!
I can see it now. A dude on a Harley with Screaming Eagles is idling next to me at a light... he begins laughing, right as the light turns green... I'm in low-power mode! I'm only pushing 3kw! But my 6kw motor can take even greater peak power, and my A123 pack can provide it! I twist the throttle all the way down! TWELVE KILOWATTS push me forward as I squeal ahead, thanks to my superior torque response and power/weight ratio!
gensem
100 kW
Just to keep this alive.
Any news on the final controller version?
Any news on the final controller version?
parabellum
1 MW
It is told, controllers will be available at end of this month, and BMS in ~ 3weeks.
That's great news, I just read through their forums a couple days ago as best as I could, didn't see anything. I hope somebody here who is getting one before me will post a hands-on review.
It really is expensive, but if you compare to the cost of a controller, BMS, and CycleAnalyst... well, to have it all together in one efficient and silent package, programmable on the fly, it's definitely worth it to me.
It really is expensive, but if you compare to the cost of a controller, BMS, and CycleAnalyst... well, to have it all together in one efficient and silent package, programmable on the fly, it's definitely worth it to me.
boostjuice
10 kW
It's interesting to finally see the guts of the MaxE. It appears to have a standard linear FET layout (18 x TO-247 it appears) so little innovation there. So it seems it's pwr circuit advantages are predominantly based on the clever firmware and phase current monitoring/feedback.
And to see the connectors: http://velomastera.ru/shop/admin/pictures/501036b-5.jpg
I could be wrong but it appears that it is possible to purchase 2 controllers and use the display with 2 Max-E's-- I know it was said dual Mini-E's will be an option, but dual Max-E's would open up the door for 6-phase motors like the Hubmonster and Midmonster, though it's less necessary to have smooth commutation with those motors... would there be an advantage in a 6-phase motor?
Regarding field-oriented control and sinusoidal control, I found this summary PDF helpful, though it fails to distinguish between direct and indirect field-oriented control. http://www.maccon.de/fileadmin/FTPROOT/Field-Oriented-Control.pdf
I don't know a whole lot beyond the principles of these approaches, but if this controller utilizes indirect FOC for the sensorless mode and direct FOC for the sensored... well, this is even greater than I thought. Either way I don't think it matters terribly, as the RPM's we will hit with our hub motors shouldn't get too high and sensorless operation is a last-resort.
It should be said that while sinusoidal is far better than trapezoidal for a hub motor, halls aren't a perfect measure of position so it won't be perfect. But it'll obviously be much better.
I could be wrong but it appears that it is possible to purchase 2 controllers and use the display with 2 Max-E's-- I know it was said dual Mini-E's will be an option, but dual Max-E's would open up the door for 6-phase motors like the Hubmonster and Midmonster, though it's less necessary to have smooth commutation with those motors... would there be an advantage in a 6-phase motor?
Regarding field-oriented control and sinusoidal control, I found this summary PDF helpful, though it fails to distinguish between direct and indirect field-oriented control. http://www.maccon.de/fileadmin/FTPROOT/Field-Oriented-Control.pdf
I don't know a whole lot beyond the principles of these approaches, but if this controller utilizes indirect FOC for the sensorless mode and direct FOC for the sensored... well, this is even greater than I thought. Either way I don't think it matters terribly, as the RPM's we will hit with our hub motors shouldn't get too high and sensorless operation is a last-resort.
It should be said that while sinusoidal is far better than trapezoidal for a hub motor, halls aren't a perfect measure of position so it won't be perfect. But it'll obviously be much better.
full-throttle
1 MW
The caps are mounted on the opposite side of the board to reduce stray inductance.boostjuice said:
Connectors will be different (better) on production lot.xenodius said:
bio said:
As far as I know, a precise estimation would be very difficult to determine. Sinusoidal controllers have the best advantage at very low speeds, e.g. starting. At high speeds it would probably be less noticeable. I'd expect to get more torque and less waste heat off of starts with a sinusoidal controller.
I would also love to hear an update. I've been alternating between this controller and a large 3-phase BLDC and back to a CA v3, dual Lyen controllers and a 6-phase BLDC. And if I get this electronics package, I wont be needing a BMS, which will also determine which batteries I choose. I can't start yet, but everything depends on this controller. I have to say that analog regen sounds like one of the most useful features... as well as a kick-down (2-stage) throttle.
An aside directed at bio: I found out after my last comment that the maximum theoretical efficiency of a 3-phase BLDC with trapezoidal control is 92%. For a sinusoidal controller, it would be 100%. So you could expect, at most, an 8% increase in efficiency. Which is actually no small thing-- that's a lot. I bet you could significantly reduce starting waste heat (and therefore increase power) with a sinusoidal controller, but I wouldn't expect radical increases in efficiency for maintaining cruising speeds. That said, if you can reduce starting waste heat you may be able to comfortably increase cruising power a little.
The efficiency of trapezoidal control at higher speeds can be evinced in RC setups... they use simple trapezoidal controllers but have very high efficiencies in the 94% range.
An aside directed at bio: I found out after my last comment that the maximum theoretical efficiency of a 3-phase BLDC with trapezoidal control is 92%. For a sinusoidal controller, it would be 100%. So you could expect, at most, an 8% increase in efficiency. Which is actually no small thing-- that's a lot. I bet you could significantly reduce starting waste heat (and therefore increase power) with a sinusoidal controller, but I wouldn't expect radical increases in efficiency for maintaining cruising speeds. That said, if you can reduce starting waste heat you may be able to comfortably increase cruising power a little.
The efficiency of trapezoidal control at higher speeds can be evinced in RC setups... they use simple trapezoidal controllers but have very high efficiencies in the 94% range.
adrian_sm
1 MW
xenodius said:
These statement appear to contradict each other.
Would you mind linking to your references? Sounds like interesting reading.
It's only true at low speeds-- essentially, when starting. I have no citation. Just a back-of-the-envelope. But I linked a PDF a few posts back that compares trapezoidal, sinusoidal, and field-oriented controllers... It should clarify why the trend occurs.
PDF: http://www.maccon.de/fileadmin/FTPROOT/Field-Oriented-Control.pdf
PDF: http://www.maccon.de/fileadmin/FTPROOT/Field-Oriented-Control.pdf
adrian_sm
1 MW
Hmmm. That doesn't match my experience at all. I have been conducting a lot of dyno testing of RC motors, RC controllers, and ebike controllers recently. At low speeds the RC style trapezoidal controllers are not efficient at all, and create a huge amount of waste heat.
Thanks for the link. But the descriptions they give of the trapezoidal control seems to neglect motor inductance and the effect on phase current. They seem to think the phase current has a step function perfectly matching the switching, which is not the case. Perhaps this is closer to the truth for stepper motors that they are more concerned with, or they are just over simplifying to suit there needs.
Thanks for the link. But the descriptions they give of the trapezoidal control seems to neglect motor inductance and the effect on phase current. They seem to think the phase current has a step function perfectly matching the switching, which is not the case. Perhaps this is closer to the truth for stepper motors that they are more concerned with, or they are just over simplifying to suit there needs.
xenodius said:adrian_sm said:
I'm sorry, I probably didn't explain myself well, but that is exactly what I meant. At low speeds, especially for RC style systems, trapezoidal control will be quite inefficient. I assume they're oversimplifying.
I agree that RC controllers with trapezoidal control have issues at low speed because they do not control precisely enough the phase current and the phase advance. Unfortunately, it's at low speed and low PWM that the phase current is potentially the highest and phase delay the largest. Provided the phase current is controlled precisely with a PI loop running at the PWM frequency and the phase advance is dynamically adjusted depending on phase current and motor speed then the only difference with a sine-wave control approach is the precision of the vector direction, with the trapezoidal control only using 6. This will amount for some inefficiency but not that much.
adrian_sm
1 MW
I did a few runs on the dyno at low speeds, and found the higher PWM rate definitely kept the controller cooler. I just assumed that the low inductance, low resistance, and lower BEMF all colluded to create big current spikes. But what your saying is that the timing is possibly off at these low speeds. Very interesting, and makes sense since the RC styles ESCs are really not expected to run at high load and low speed, so they probably don't care at all if they have the right timing then.
So given these cheaper trapezoidal controllers I am playing with have no phase current measurement, then the last thing for me to optimise appears to be the timing or advance. Guess I'll add that to the todo list.
Oh, here is an example of how the predicted equilibrium temperature fr the ESC under load was affected by the motor speed, and the PWM frequency, under the same load conditions.
Sorry if this is all too for off topic. But this may help explain a few things I was not quite sure off. Cheers.
So given these cheaper trapezoidal controllers I am playing with have no phase current measurement, then the last thing for me to optimise appears to be the timing or advance. Guess I'll add that to the todo list.
Oh, here is an example of how the predicted equilibrium temperature fr the ESC under load was affected by the motor speed, and the PWM frequency, under the same load conditions.
Sorry if this is all too for off topic. But this may help explain a few things I was not quite sure off. Cheers.
Similar threads
