PowerVelocity controller review

ElectricGod · Sep 28, 2017

Alan B said:
Another more important clue that gives us information about the control type is the answer to the question "does the controller have real time synchronous sensing for at least two (or all three) of the motor phase currents". The Trap and sinusoidal controllers generally only have battery current sensing using a shunt which is located on the battery side of the capacitor bank, thus it is filtered and cannot give real time information relating to motor currents. FOC requires accurate and prompt motor current information (of at least two phases) to determine fields and calculate rotor position. They also need a lot more information about the motor such as resistance and inductance whereas the simpler controllers don't need that. FOC has a lot of tuning parameters for things like this as well as the multiple feedback loops that it must operate. So by looking at the parameters one can guess a lot about the control algorithm.

Thanks for keeping this thread polite and informational. We all learn more when threads are well mannered.

I can tell you for a fact that none of the phases have any current sensing on them for 100% of the product line. For example looking at the phaserunner or VESC variants, they have shunts on each phase. So it sounds like you may be right...that the 3232 is not a basic FOC style MCU. Awe damn...now I need to edit my posts in this thread to change anywhere I posted that it was FOC. These controllers have 2 sets of caps in them. There are the main batt+ and - busses that have caps on them and then at each power buss per phase is another set of filter caps. The shunts "live" between the battery power buss and the phase buss. That sounds like they would be more reactive than battery side shunts, but maybe we are talking about the same thing.

Talking about tuning points...the ASI I have requires all of this sort of information. Some of it, it detects for itself, some you have to enter or fiddle with endlessly until it "works". The inconsistency in the controller programming for motor poles was ridiculous. I don't think I ever set an inrunner or outrunner motor to it's actual pole count on that controller.

gman1971 · Sep 28, 2017

Timing vs FOC... that is the question. Advancing timing is what is already done on my 49 dollar el-cheapo 40A square wave Cyclone controller, works good, but is no FOC... again, there is the confusion, I agree.

I am certain that a true FOC sensorless is my Kontronik Kosmik ESC, that one has no hall sensors anywhere, they are simply not needed; and the thing is butter smooth and efficient, there is nothing hot even after hovering a giant 16 lbs RC scale helicopter for 1 hour in the air. The motor coils are about 150F after 1 hour flight using ~800W, the controller remains at ambient temperature during the entire flight, inside a closed fuselage without ventilation, ... again.. that is what I am after, and I think I am not the only one, a true sensorless FOC controller that works on more than 14S.

I also own a host of Kontronik controllers, from Jazzs, to Jives, you name it, and nothing comes close to them in terms of sensorless smoothness, starting without hiccups and efficiency; simply put, nothing on those controllers ever gets hot, no matter what you throw at them; as for needing halls to start up, I think there is something magical about my Jazz controller then, b/c its in one of my electric RC cars, and it simply blows away all my sensored controllers in every dept, and stays cool even running all out when the others need a heat sink the size of texas... again, I just don't know how they do it, but these controllers are IMO the gold standard in sensorless motor control.

G.

Alan B said:
Wow, a lot has gone by while I was out doing errands.

As fechter has said, there are different ways of increasing the speed of a motor and different ways of viewing this. Field Weakening which FOC does by applying current generating a field in opposition to the permanent magnets can also be looked at as shifting the timing which has been done for many years with trap controllers and would also be used with Sinewave commutated ones. I suspect the terminology "field weakening" should apply to the FOC variants, and "timing advance" should be used with encoder driven systems (such as hall sensors). It is another area where the terms can imply something and mislead.

Which CPU is used in a controller gives some idea of the computing and I/O capability available, but the firmware in the controller can vary tremendously. For example some fairly low powered CPUs have special I/O that allows them to run FOC. Clearly they can also run Sinewave commutation, or trap as well. So the CPU alone doesn't tell us much.

Another more important clue that gives us information about the control type is the answer to the question "does the controller have real time synchronous sensing for at least two (or all three) of the motor phase currents". The Trap and sinusoidal controllers generally only have battery current sensing using a shunt which is located on the battery side of the capacitor bank, thus it is filtered and cannot give real time information relating to motor currents. FOC requires accurate and prompt motor current information (of at least two phases) to determine fields and calculate rotor position. They also need a lot more information about the motor such as resistance and inductance whereas the simpler controllers don't need that. FOC has a lot of tuning parameters for things like this as well as the multiple feedback loops that it must operate. So by looking at the parameters one can guess a lot about the control algorithm.

Thanks for keeping this thread polite and informational. We all learn more when threads are well mannered.

Alan B · Sep 28, 2017

Thanks for the information It seems pretty consistent with what we expected.

gman1971 · Sep 28, 2017

Correct, that is the advantage of mid-drives, you get to use the gears to shift the powerband. I am achieving top speed through a combination of gearing and aerodynamics (trying to keep it open cockpit too) I am at a point in which the aero is so good that my controller efficiency starts to play an important role, hence why I am so eagerly looking for a true FOC controller.

Again, I wasn't trying to start a pissing match, or claim the PV controllers are crap, I was honestly disappointed it wasn't true FOC... so, I am sorry if I came too strong. I am also learning a lot in this thread thanks to Alan.

G.

ElectricGod said:
I see how you get to 65mph...you are using the bikes gears to get there. Otherwise if it wasn't for gearing you would have long since been limited by the motor Kv.

ASI told me like a year ago that they are working on a 4kw (18 fet) and 8kw (24 fet) controller. I had bought one of their 12 fet controllers

ElectricGod · Sep 28, 2017

gman1971 said:
Correct, that is the advantage of mid-drives, you get to use the gears to shift the powerband. I am achieving top speed through a combination of gearing and aerodynamics (trying to keep it open cockpit too) I am at a point in which the aero is so good that my controller efficiency starts to play an important role, hence why I am so eagerly looking for a true FOC controller.

Again, I wasn't trying to start a pissing match, or claim the PV controllers are crap, I was honestly disappointed it wasn't true FOC... so, I am sorry if I came too strong. I am also learning a lot in this thread thanks to Alan.

G.

ElectricGod said:

I see how you get to 65mph...you are using the bikes gears to get there. Otherwise if it wasn't for gearing you would have long since been limited by the motor Kv.

ASI told me like a year ago that they are working on a 4kw (18 fet) and 8kw (24 fet) controller. I had bought one of their 12 fet controllers

Click to expand...

True FOC is going to cost you a good bit. I have a 12 fet ASI and it cost me more than the 24 fet PV controller! It may be great at FOC, but geez the cost!!! It's just my opinion, but a 5% performance gain at nearly 2X the price isn't compelling at all. I can think of a lot of other things I can do with my $$ than buy an over priced controller.

gman1971 · Sep 28, 2017

Yeah, I know man; the Kosmik ESC was near 1k, but having 1 hour flight times was soo worth it; even though I usually fly no more than 15 minutes before I am tired and land... having the ability to decide when I land rather than being forced to land was worth a lot to me. The same kinda applies to the trike, I've already optimized it pretty well in most areas, without a wind tunnel ($$$$$$$) I can't get much better aero as it stands now, the chainline now runs entirely on ball bearings vs. tubes (not ceramic bearings, BTW) so the loss is very little there, there isn't much left to optimize for cheap, so I am at the point in which gains are going to be minimal for a larger and larger amount of investment (either time or money, or both) The next one would be to redo the entire body in CF, will cut the weight by 30 lbs probably, but that is a large $$$$ and time investment as well... the controller for as crazy as it sounds is the lowest hanging fruit ATM in my plate....

G.

ElectricGod said:
gman1971 said:

Correct, that is the advantage of mid-drives, you get to use the gears to shift the powerband. I am achieving top speed through a combination of gearing and aerodynamics (trying to keep it open cockpit too) I am at a point in which the aero is so good that my controller efficiency starts to play an important role, hence why I am so eagerly looking for a true FOC controller.

Again, I wasn't trying to start a pissing match, or claim the PV controllers are crap, I was honestly disappointed it wasn't true FOC... so, I am sorry if I came too strong. I am also learning a lot in this thread thanks to Alan.

G.

ElectricGod said:

I see how you get to 65mph...you are using the bikes gears to get there. Otherwise if it wasn't for gearing you would have long since been limited by the motor Kv.

ASI told me like a year ago that they are working on a 4kw (18 fet) and 8kw (24 fet) controller. I had bought one of their 12 fet controllers

Click to expand...

Click to expand...

True FOC is going to cost you a good bit. I have a 12 fet ASI and it cost me more than the 24 fet PV controller! It may be great at FOC, but geez the cost!!! It's just my opinion, but a 5% performance gain at nearly 2X the price isn't compelling at all. I can think of a lot of other things I can do with my $$ than buy an over priced controller.

gman1971 · Sep 28, 2017

What is the model # for the ASI controller that is FOC?

Thanks.

G.

amberwolf · Sep 29, 2017

I think all of them are, but I haven't checked them all out.
http://accelerated-systems.com/

But the Grin Tech Phaserunner is also an ASI controller, and it's FOC. They've also got the BAC2000 from ASI.

gman1971 · Sep 29, 2017

amberwolf said:
I think all of them are, but I haven't checked them all out.
http://accelerated-systems.com/

But the Grin Tech Phaserunner is also an ASI controller, and it's FOC. They've also got the BAC2000 from ASI.

Thanks.

Yeah, would love to use the Phaserunner, but unfortunately won't work on 20S

gman1971 · Sep 29, 2017

I stand corrected the Phaserunner will indeed work on 20S, 90V max, 20S LiPo is 85V fully charged.

G.

ElectricGod · Sep 29, 2017

gman1971 said:
Yeah, I know man; the Kosmik ESC was near 1k, but having 1 hour flight times was soo worth it; even though I usually fly no more than 15 minutes before I am tired and land... having the ability to decide when I land rather than being forced to land was worth a lot to me. The same kinda applies to the trike, I've already optimized it pretty well in most areas, without a wind tunnel ($$$$$$$) I can't get much better aero as it stands now, the chainline now runs entirely on ball bearings vs. tubes (not ceramic bearings, BTW) so the loss is very little there, there isn't much left to optimize for cheap, so I am at the point in which gains are going to be minimal for a larger and larger amount of investment (either time or money, or both) The next one would be to redo the entire body in CF, will cut the weight by 30 lbs probably, but that is a large $$$$ and time investment as well... the controller for as crazy as it sounds is the lowest hanging fruit ATM in my plate....

An hour worth of flight time is a giant battery pack! I'm sure the controller helped in some small way to extend flight times. Still a giant battery pack is the only reason why you can fly for an hour. If people are forced to land due to ESC overheating, that's due to an under powered or cheap ESC or both. I used to fly electric heli's and planes. Honestly, I'm surprised you could get that large of a pack in a heli. In a plane I can see it happening since they are much more efficient fliers.

I bet your trike is quieter with no tubes for the chain to rattle inside of all the time. I use ceramic balled bearings in just about everything. All steel bearings are cheap and common and are better than no bearing, but the difference between ceramic balls and steel balls is quite significant. I have a stack of ceramic balled bearings on my bench right now so that I can pull the steel bearings out of several motors and replace them. The friction reduction between ceramic balled and all ceramic bearings IMHO isn't worth the cost, but going from shielded steel to sealed ceramic balled bearings is a significant frictional improvement. As a result, every bearing that is steel and shielded or sealed, gets replaced with sealed ceramic balled bearings. It all helps! Do I get 40mph on steel bearings and 45 mph on ceramics? No, but it helps a little. I've got quite a few new shielded bearings that I replaced with sealed ceramic bearings. You can feel the difference just spinning the bearing in your fingers. IMHO ceramic balled bearings are worth it.

CF...this is on my list of things to get into. Making your own CF components isn't overly difficult. It's just an air compressor/vacuum pump, bags, CF sheets, epoxy and shapes to mold around. The challenge is joining CF to metal and having it be strong. You can't really thread into CF so that means brass or stainless threaded inserts for things that need to bolt on. Also CF doesn't take getting bashed into things particularly well. A friend of mine has 4 CF bikes right now that all weigh right at 15 pounds. He's also had 5 more before them that were involved in various wrecks and had to be retired due to frame cracking. Once CF gets a crack in it, you are going to be doing major rework if it's a structural component. It's just an opinion, but I'd use CF for things where I don't expect any physical impact to happen and then aluminum for everything else.

If your trike is already aluminum, I'd leave it alone and go with a bigger motor, pack and controller if your speed and acceleration are not good enough yet. Of course getting away from a motor that uses planets (cyclone) is an excellent idea too. There's lots of inrunners and outrunners out there that have no need of planetary gears to be used on an EV. I get why people use the cyclone. It's already geared down to human pedaling cadences so it makes implementation easy, but there are other options too that are better. I'm looking at building a quad bike based on the Utah Trikes quad set up. My motor will drive the back wheels directly. It's possible I might use a 2 or 3 speed transmission. I won't use the human powered components at all for electric drive. It is highly likely that it will have a PowerVelocity controller on it since they do work quite well on inrunners and outrunners and not just hub motors.

Another direction I am going in soon is the micro EV. I've already picked out the machine I want to build onto. It's a Lehe K1S. Front and rear suspension and disk brakes come standard. I'm thinking a 6 or 12 fet PV controller, a C80100 outrunner and run it all at 66 volts. It might be a good home for my ASI controller as well. I'm quite sure that with single stage gearing I can exceed 40mph and accelerate as well as my GTI. My Currie on this exact setup and 75 pounds will do that. This project would be about keeping the EV weight at 40 pounds max, achieve 40mph+ and have 15-20 miles of range of pure electric power.

Before either of these things happen, I'm going to program the ASI controller for my Currie/C80100 outrunner and see if I find it giving me the "OMG!! This is so much better than my PV controller!" feel. I always do a standard tune on any controller that will work and then tweak the settings for optimal performance. The 12 fet PV will get a bit more of that over this weekend. I'll get the ASI set up to be a plug in replacement that I can strap down to the top of the Currie back deck. The PV 12 fet will have been optimally tuned and seen at least 2 full battery charges by then or about 50 miles. I'll then install the ASI in place of the 12 fet PV and see what happens. I'm not the least bit hung on one opinion or the other, but I have to say that the ASI for 2X the price is going to need to really impress me. A 5% improvement for 2X the cost isn't compelling IMHO. In either case, I will continue to promote the PowerVelocity product line as a great, reasonably priced, middle of the road controller option.

So far I've used the 12 fet controllers the most since I have 3 of them and fried one due to my own errors. The 18 fet version is running strong, but has yet to see real load testing and tweaking since that EV isn't finished yet. It will initially run at 82 volts, but I may bump that to 130 volts. OR I will swap out it's IRF4115 mosfets for AOT290's and keep it at 82 volts. If I go to AOT290's, I will be able to bump phase and battery amps significantly and that will exceed the HLD/LR big block inrunners capabilities that I'm using on the XB-502 moped. I have my eye on it's big brother already. My battery/BMS solution is already capable of 160 amps continuous and 32,000mah capacity. I have just received the 24 fet PV controller which will probably get used in the quad bike and will run an Alien Power 12090 50kv outrunner. I'm expecting something like 15kw out of the motor. It will be quite strong! In that build, I will also have the option to compare the 24 fet Sabvoton MQCON controller. Again we will see if FOC is a mind blow or not. 5% improvement isn't going to get me to switch for most uses.

gman1971 · Sep 29, 2017

Its a large heli, so you can fit a large battery, but lets just say the controller amounted to a good amount of that flight time.

Yes, the trike is dead silent in terms of chain noise, is just the C3000W motor that has the F-1 whine when you punch it and rip through the gears... which I kinda like too... makes me feel I am doing 160!

but the FOC controller will make it even more silent so even more stealth...

better!! although I won't get my F-1 sound tho

It is indeed good to know the ceramics make a noticeable difference, thanks for that tip; I wasn't sure if it would be worth the $$$$ at such low RPM...

Yeah, the efficiency game is a tedious one to play, but the aero was the biggest win for the lowest amount of money... just a few dollars worth of plastic...

heat gun and molds did the trick. Why add more power when you can just lower the requirements to get there, right? I might take the plunge and get some ceramics bearings... worth a try, and nowadays they are fairly cheap

CF is also on my list to do, but this is just for a shell, I have no plans on doing anything structural on home made CF; and while I got really good at thermoforming panels, and love the simplicity of it and the no-mess process... you end up adding more weight than you probably should, but is easy to repair and nearly indestructible...

No aluminum here, I went with a steel frame, and that was a MUST in the initial design choices when I set out to build the trike; I would never dare to use any commercial grade AL alloy on a 60+ mph trike without doing a long FEA study.. as AL fatigues with vibration and steel doesn't (so FEA is much less of a requirement so over-engineering works in this case) and its a lot springier too; one can abuse the snot out of the frame without it ever cracking due to fatigue... that was another reason why I built the body around it, to prevent water and other crap from getting in.

Motor... trike can (theoretically) already hit ~70 mph on ~6kW, so there is no need to get a bigger motor, trust me. The trike also runs 90A peaks at shunt, (unknown phase amps since I shunt modded the thing), so its stupid quick too, faster than cars if that answers the question; but what I need now is need a more efficient controller, so it will require less amps to do the same speed, which equals longer range. Me wants the combination of a "Chevy Volt efficiency", "Go-Kart performance" and "F-1 looks" long though mutually exclusive...

5% efficiency isn't a priority when you're dealing with already super efficient vehicles (compared to gas guzzling cars) but when want to push 50 miles out of 1 kWh averaging ~30 MPH speed (not cruise, average) all motor, every percent you can shave counts, kinda like the ceramic bearings.

You seem to know your way around these controllers really well, what programming language do you use for the MCU? C++?

G.

ElectricGod said:
An hour worth of flight time is a giant battery pack! I'm sure the controller helped in some small way to extend flight times. Still a giant battery pack is the only reason why you can fly for an hour. If people are forced to land due to ESC overheating, that's due to an under powered or cheap ESC or both. I used to fly electric heli's and planes. Honestly, I'm surprised you could get that large of a pack in a heli. In a plane I can see it happening since they are much more efficient fliers.

I bet your trike is quieter with no tubes for the chain to rattle inside of all the time. I use ceramic balled bearings in just about everything. All steel bearings are cheap and common and are better than no bearing, but the difference between ceramic balls and steel balls is quite significant. I have a stack of ceramic balled bearings on my bench right now so that I can pull the steel bearings out of several motors and replace them. The friction reduction between ceramic balled and all ceramic bearings IMHO isn't worth the cost, but going from shielded steel to sealed ceramic balled bearings is a significant frictional improvement. As a result, every bearing that is steel and shielded or sealed, gets replaced with sealed ceramic balled bearings. It all helps! Do I get 40mph on steel bearings and 45 mph on ceramics? No, but it helps a little. I've got quite a few new shielded bearings that I replaced with sealed ceramic bearings. You can feel the difference just spinning the bearing in your fingers. IMHO ceramic balled bearings are worth it.

CF...this is on my list of things to get into. Making your own CF components isn't overly difficult. It's just an air compressor/vacuum pump, bags, CF sheets, epoxy and shapes to mold around. The challenge is joining CF to metal and having it be strong. You can't really thread into CF so that means brass or stainless threaded inserts for things that need to bolt on. Also CF doesn't take getting bashed into things particularly well. A friend of mine has 4 CF bikes right now that all weigh right at 15 pounds. He's also had 5 more before them that were involved in various wrecks and had to be retired due to frame cracking. Once CF gets a crack in it, you are going to be doing major rework if it's a structural component. It's just an opinion, but I'd use CF for things where I don't expect any physical impact to happen and then aluminum for everything else.

If your trike is already aluminum, I'd leave it alone and go with a bigger motor, pack and controller if your speed and acceleration are not good enough yet. Of course getting away from a motor that uses planets (cyclone) is an excellent idea too. There's lots of inrunners and outrunners out there that have no need of planetary gears to be used on an EV. I get why people use the cyclone. It's already geared down to human pedaling cadences so it makes implementation easy, but there are other options too that are better. I'm looking at building a quad bike based on the Utah Trikes quad set up. My motor will drive the back wheels directly. It's possible I might use a 2 or 3 speed transmission. I won't use the human powered components at all for electric drive. It is highly likely that it will have a PowerVelocity controller on it since they do work quite well on inrunners and outrunners and not just hub motors.

Another direction I am going in soon is the micro EV. I've already picked out the machine I want to build onto. It's a Lehe K1S. Front and rear suspension and disk brakes come standard. I'm thinking a 6 or 12 fet PV controller, a C80100 outrunner and run it all at 66 volts. It might be a good home for my ASI controller as well. I'm quite sure that with single stage gearing I can exceed 40mph and accelerate as well as my GTI. My Currie on this exact setup and 75 pounds will do that. This project would be about keeping the EV weight at 40 pounds max, achieve 40mph+ and have 15-20 miles of range of pure electric power.

Before either of these things happen, I'm going to program the ASI controller for my Currie/C80100 outrunner and see if I find it giving me the "OMG!! This is so much better than my PV controller!" feel. I always do a standard tune on any controller that will work and then tweak the settings for optimal performance. The 12 fet PV will get a bit more of that over this weekend. I'll get the ASI set up to be a plug in replacement that I can strap down to the top of the Currie back deck. The PV 12 fet will have been optimally tuned and seen at least 2 full battery charges by then or about 50 miles. I'll then install the ASI in place of the 12 fet PV and see what happens. I'm not the least bit hung on one opinion or the other, but I have to say that the ASI for 2X the price is going to need to really impress me. A 5% improvement for 2X the cost isn't compelling IMHO. In either case, I will continue to promote the PowerVelocity product line as a great, reasonably priced, middle of the road controller option.

So far I've used the 12 fet controllers the most since I have 3 of them and fried one due to my own errors. The 18 fet version is running strong, but has yet to see real load testing and tweaking since that EV isn't finished yet. It will initially run at 82 volts, but I may bump that to 130 volts. OR I will swap out it's IRF4115 mosfets for AOT290's and keep it at 82 volts. If I go to AOT290's, I will be able to bump phase and battery amps significantly and that will exceed the HLD/LR big block inrunners capabilities that I'm using on the XB-502 moped. I have my eye on it's big brother already. My battery/BMS solution is already capable of 160 amps continuous and 32,000mah capacity. I have just received the 24 fet PV controller which will probably get used in the quad bike and will run an Alien Power 12090 50kv outrunner. I'm expecting something like 15kw out of the motor. It will be quite strong! In that build, I will also have the option to compare the 24 fet Sabvoton MQCON controller. Again we will see if FOC is a mind blow or not. 5% improvement isn't going to get me to switch for most uses.

ElectricGod · Sep 29, 2017

gman...my experience with trapezoidal vs sinusoidal in inrunners was a very small sound difference. I have 2 HLD/LR big block inrunners. The sound difference between them is very slight. One is running on a PV sinusoidal controller and the other on a trapezoidal Kelly controller. You probably won't notice much if any motor noise difference. Also your cyclone has planets in it and they account for probably 60% of your motors noises anyway. Typically on a planet free inrunner, the chain/sprocket noises are much louder than the motor. On an outrunner sinusoidal is slightly more noticeable than a pure inrunner. On hubs is where you really hear the difference between trapezoidal and sinusoidal. Since the sound difference between an inrunner on trapezoidal and sinusoidal is so small anyway, probably going with an FOC controller is not going to make any audible difference. On a motor with planets, definitely not.

gman1971 · Sep 29, 2017

ElectricGod said:
gman...my experience with trapezoidal vs sinusoidal in inrunners was a very small sound difference. I have 2 HLD/LR big block inrunners. The sound difference between them is very slight. One is running on a PV sinusoidal controller and the other on a trapezoidal Kelly controller. You probably won't notice much if any motor noise difference. Also your cyclone has planets in it and they account for probably 60% of your motors noises anyway. Typically on a planet free inrunner, the chain/sprocket noises are much louder than the motor. On an outrunner sinusoidal is slightly more noticeable than a pure inrunner. On hubs is where you really hear the difference between trapezoidal and sinusoidal. Since the sound difference between an inrunner on trapezoidal and sinusoidal is so small anyway, probably going with an FOC controller is not going to make any audible difference. On a motor with planets, definitely not.

Thanks... I really like the F-1 whine so that's all good !

G.

Alan B · Sep 30, 2017

People have reported noise reduction in geared motors from sine wave drive, so apparently part of the noise in gear reductions comes from torque ripple (at least in some cases), so changing from trapezoidal to sine wave drive can make the geared motor quieter. It makes sense that the torque ripple hammering on the gears could make them generate more noise, just as the side covers often do on DD motors. I have not tested this. I recall it from a thread on ES, perhaps it was the PhaseRunner thread.

It would be fun to have a controller that could change from trap to sine wave to experiment with this, and be able to control the noise level, thus having a "silent running" feature. This could be implemented with the three/four speed switch as an extra programming option, perhaps, so it would not require any new wires. So each speed option could have a sine/trap selection option in the programming software. This would also allow testing the speed, power and efficiency differences between trap and sine drive.

Powervelocity.com · Sep 30, 2017

The difference between sinusoidal and trapezoidal modulation is well pronounced with hub motors on hard tail frames under moderate to hard acceleration. With trapezoidal you get lots of vibrations transferred to the frame that makes the whole thing sound almost like a gas powered engine. Kind of cool I guess from that perspective for some people (not my thing though). With sinusoidal modulation, when you hit the throttle, it's like catching a wave that is smoothly and strongly pulling you forward. Can't hear the motor at all.

As for switching between sinusoidal and trapezoidal, that can be done on dual mode F6, F12, F18 models by just connecting or not connecting the hall sensors. Disconnecting halls enables sensorless mode with trapezoidal modulation, connecting sensors switches it back to sinusoidal. Don't even need to program anything. Technically, a switch can be fitted to do that on the fly for comparison/testing purposes.

Alan B said:
People have reported noise reduction in geared motors from sine wave drive, so apparently part of the noise in gear reductions comes from torque ripple (at least in some cases), so changing from trapezoidal to sine wave drive can make the geared motor quieter. It makes sense that the torque ripple hammering on the gears could make them generate more noise, just as the side covers often do on DD motors. I have not tested this. I recall it from a thread on ES, perhaps it was the PhaseRunner thread.

It would be fun to have a controller that could change from trap to sine wave to experiment with this, and be able to control the noise level, thus having a "silent running" feature. This could be implemented with the three/four speed switch as an extra programming option, perhaps, so it would not require any new wires. So each speed option could have a sine/trap selection option in the programming software. This would also allow testing the speed, power and efficiency differences between trap and sine drive.

Alan B · Sep 30, 2017

Great suggestion. A single switch on the hall sensor +5v line should do that.

Is that the case on the PV 12 FET I bought from you (9/2016), or is it too old? (I have not tested it yet).

madin88 · Sep 30, 2017

gman1971 said:
Timing vs FOC... that is the question. Advancing timing is what is already done on my 49 dollar el-cheapo 40A square wave Cyclone controller, works good, but is no FOC... again, there is the confusion, I agree.

that nailded it down!

I am certain that a true FOC sensorless is my Kontronik Kosmik ESC, that one has no hall sensors anywhere, they are simply not needed; and the thing is butter smooth and efficient, there is nothing hot even after hovering a giant 16 lbs RC scale helicopter for 1 hour in the air. The motor coils are about 150F after 1 hour flight using ~800W, the controller remains at ambient temperature during the entire flight, inside a closed fuselage without ventilation, ... again.. that is what I am after, and I think I am not the only one, a true sensorless FOC controller that works on more than 14S.

As an owner of some Jazz controller powered rc planes i know that Kontronik is great.
I just asked google about Kosmik controller and WOW, 120Mhz processor, current sensing on each phase, Sine wave with FOC control (or at least ready in terms of hardware for that) is impressive.
Well, but also the price :shock:

I can imagine that it would work quite well for powering light EV's (when it needs to control 2-3 times higher current when accerlerating or climbing hills), but the lack of program features for such (like regen, torque control) makes them not much interesting for such usage.

Powervelocity.com · Sep 30, 2017

The dual mode versions of 6F, 12F, 18F controllers became available a couple months ago, so your controller purchased in 2016 does not have that. I am still offering both versions - single and dual mode - and there is a slight difference in pricing.

This is a single mode/sinusoidal version:
http://powervelocity.com/index.php?id_product=54&controller=product

And this is the dual mode version:
http://powervelocity.com/index.php?id_product=56&controller=product

Alan B said:
Great suggestion. A single switch on the hall sensor +5v line should do that.

Is that the case on the PV 12 FET I bought from you (9/2016), or is it too old? (I have not tested it yet).

ElectricGod · Sep 30, 2017

madin88 said:
gman1971 said:

Timing vs FOC... that is the question. Advancing timing is what is already done on my 49 dollar el-cheapo 40A square wave Cyclone controller, works good, but is no FOC... again, there is the confusion, I agree.

Click to expand...

that nailded it down!

I am certain that a true FOC sensorless is my Kontronik Kosmik ESC, that one has no hall sensors anywhere, they are simply not needed; and the thing is butter smooth and efficient, there is nothing hot even after hovering a giant 16 lbs RC scale helicopter for 1 hour in the air. The motor coils are about 150F after 1 hour flight using ~800W, the controller remains at ambient temperature during the entire flight, inside a closed fuselage without ventilation, ... again.. that is what I am after, and I think I am not the only one, a true sensorless FOC controller that works on more than 14S.

Click to expand...

As an owner of some Jazz controller powered rc planes i know that Kontronik is great.
I just asked google about Kosmik controller and WOW, 120Mhz processor, current sensing on each phase, Sine wave with FOC control (or at least ready in terms of hardware for that) is impressive.
Well, but also the price
I can imagine that it would work quite well for powering light EV's (when it needs to control 2-3 times higher current when accerlerating or climbing hills), but the lack of program features for such (like regen, torque control) makes them not much interesting for such usage.

VESC's are small and light and could be used in an RC heli for 30% the cost or less of these crazy expensive RC ESC's. I'd go that route first.

ElectricGod · Sep 30, 2017

Alan B said:
Great suggestion. A single switch on the hall sensor +5v line should do that.

Is that the case on the PV 12 FET I bought from you (9/2016), or is it too old? (I have not tested it yet).

My latest 12 fet is the only controller I have that supports sensorless operation. A switch that cuts 5 volts to the halls would do the trick. I haven't tried it, but I'm curious if I can kill the halls on the fly and have things still run. These controllers have supported loss of a single hall for a long time now and still run, but that's at less motor performance. What happens when all 3 halls are lost while under load? I've never tried that. It would be easy to set up a test rig, where I pull the 5 volt line on a running motor and see what happens.

ElectricGod · Sep 30, 2017

The past few days, I've been messing with the 24 fet controller I just got. I have a few minor issues that certainly are NOT show stoppers.

1. My controller has the older MCU firmware that doesn't support sensorless operation. Unfortunately we don't control the MCU firmware so that it can be changed.
2. The end caps have no silicon seals.
3. I found a couple of minor wiring items that I thought needed a little rework.

The small wires are not bound together particularly well and are not immobilized at all. THe grommet doesn't stop them from sliding in and out of the controller despite being as tight as it will go. Slowly over time vibration and movement will fatigue the wires and eventually cause a wire to break off from the board. I zip tied the wires together much more so that movement is halted.

The second item is probably adequate, but I'm whiny about connectors being better than sufficient for the job. I pulled the 5 power screw terminals and replaced them with beefier ones. Of course all the 8 awg wires external to the controller also use these screw lugs. The wires are soldered in place and not crimped. The original screw lugs were also soldered to the wire ends inside the controller.

These are a few pics of the original lugs that came on the controller and what I replaced them with. The new ones are a good bit thicker in every way and the hole in the lug is 6mm, like the screw and not 10mm like the original lugs. More surface contact and thicker conductors is always a good thing. Closing up the controller with the slightly longer lugs in it was a bit of a tighter fit, but still doable.

Screw%20lug%20thickness%20comparison.jpg

I don't know what I think about this. Notice how the 3 phase wires are all different lengths. Does this matter? The shortest wire is 2" long and the longest about 5". Should I make them all 5" long?

gman1971 · Oct 1, 2017

madin88 said:
gman1971 said:

Timing vs FOC... that is the question. Advancing timing is what is already done on my 49 dollar el-cheapo 40A square wave Cyclone controller, works good, but is no FOC... again, there is the confusion, I agree.

Click to expand...

that nailded it down!

I am certain that a true FOC sensorless is my Kontronik Kosmik ESC, that one has no hall sensors anywhere, they are simply not needed; and the thing is butter smooth and efficient, there is nothing hot even after hovering a giant 16 lbs RC scale helicopter for 1 hour in the air. The motor coils are about 150F after 1 hour flight using ~800W, the controller remains at ambient temperature during the entire flight, inside a closed fuselage without ventilation, ... again.. that is what I am after, and I think I am not the only one, a true sensorless FOC controller that works on more than 14S.

Click to expand...

As an owner of some Jazz controller powered rc planes i know that Kontronik is great.
I just asked google about Kosmik controller and WOW, 120Mhz processor, current sensing on each phase, Sine wave with FOC control (or at least ready in terms of hardware for that) is impressive.
Well, but also the price
I can imagine that it would work quite well for powering light EV's (when it needs to control 2-3 times higher current when accerlerating or climbing hills), but the lack of program features for such (like regen, torque control) makes them not much interesting for such usage.

Exactly, that is why it seats on my RC helicopter and not my Trike... is not meant to be used on EVs... sadly... all Kontronik ESCs are just good stuff.

G.

gman1971 · Oct 1, 2017

ElectricGod said:
madin88 said:

gman1971 said:

Timing vs FOC... that is the question. Advancing timing is what is already done on my 49 dollar el-cheapo 40A square wave Cyclone controller, works good, but is no FOC... again, there is the confusion, I agree.

Click to expand...

that nailded it down!

I am certain that a true FOC sensorless is my Kontronik Kosmik ESC, that one has no hall sensors anywhere, they are simply not needed; and the thing is butter smooth and efficient, there is nothing hot even after hovering a giant 16 lbs RC scale helicopter for 1 hour in the air. The motor coils are about 150F after 1 hour flight using ~800W, the controller remains at ambient temperature during the entire flight, inside a closed fuselage without ventilation, ... again.. that is what I am after, and I think I am not the only one, a true sensorless FOC controller that works on more than 14S.

Click to expand...

As an owner of some Jazz controller powered rc planes i know that Kontronik is great.
I just asked google about Kosmik controller and WOW, 120Mhz processor, current sensing on each phase, Sine wave with FOC control (or at least ready in terms of hardware for that) is impressive.
Well, but also the price
I can imagine that it would work quite well for powering light EV's (when it needs to control 2-3 times higher current when accerlerating or climbing hills), but the lack of program features for such (like regen, torque control) makes them not much interesting for such usage.

Click to expand...

VESC's are small and light and could be used in an RC heli for 30% the cost or less of these crazy expensive RC ESC's. I'd go that route first.

The VESC wasn't readily available in 2007 when I got my first Kontronik Jazz ESC. But even that, when you have 8 grand, and half a year of your life tied in something like that which flies, I much rather trust Kontronik proven track record than some open source tinker toy that has a lousy track record of blowing up (my friend who builds eSkates has one that blew up, so not particularly confidence inspiring for my flying helicopter). Also, the VESC doesn't have dual high power BEC that you can run pretty much a pair of HID headlights from... it doesn't have a proven rock solid RPM governor for controlling rotor speed, a must, the smooth start on the Kontroniks is on another league vs anything else I've seen; without these BECs, you need to add something extra to run your electronics, and something extra also means more wires, more wires = more weight = more loss = less flight time... flying machines are on another league vs. driving machines where gravity isn't a concern.

G.

ElectricGod · Oct 1, 2017

gman1971 said:
ElectricGod said:

madin88 said:

gman1971 said:

Timing vs FOC... that is the question. Advancing timing is what is already done on my 49 dollar el-cheapo 40A square wave Cyclone controller, works good, but is no FOC... again, there is the confusion, I agree.

Click to expand...

that nailded it down!

I am certain that a true FOC sensorless is my Kontronik Kosmik ESC, that one has no hall sensors anywhere, they are simply not needed; and the thing is butter smooth and efficient, there is nothing hot even after hovering a giant 16 lbs RC scale helicopter for 1 hour in the air. The motor coils are about 150F after 1 hour flight using ~800W, the controller remains at ambient temperature during the entire flight, inside a closed fuselage without ventilation, ... again.. that is what I am after, and I think I am not the only one, a true sensorless FOC controller that works on more than 14S.

Click to expand...

As an owner of some Jazz controller powered rc planes i know that Kontronik is great.
I just asked google about Kosmik controller and WOW, 120Mhz processor, current sensing on each phase, Sine wave with FOC control (or at least ready in terms of hardware for that) is impressive.
Well, but also the price
I can imagine that it would work quite well for powering light EV's (when it needs to control 2-3 times higher current when accerlerating or climbing hills), but the lack of program features for such (like regen, torque control) makes them not much interesting for such usage.

Click to expand...

VESC's are small and light and could be used in an RC heli for 30% the cost or less of these crazy expensive RC ESC's. I'd go that route first.

Click to expand...

The VESC wasn't readily available in 2007 when I got my first Kontronik Jazz ESC. But even that, when you have 8 grand, and half a year of your life tied in something like that which flies, I much rather trust Kontronik proven track record than some open source tinker toy that has a lousy track record of blowing up (my friend who builds eSkates has one that blew up, so not particularly confidence inspiring for my flying helicopter). Also, the VESC doesn't have dual high power BEC that you can run pretty much a pair of HID headlights from... it doesn't have a proven rock solid RPM governor for controlling rotor speed, a must, the smooth start on the Kontroniks is on another league vs anything else I've seen; without these BECs, you need to add something extra to run your electronics, and something extra also means more wires, more wires = more weight = more loss = less flight time... flying machines are on another league vs. driving machines where gravity isn't a concern.

G.

The original VESC worked pretty well, but then the newer VESC 6 is taking advantage of lessons learned on the original model so that they are more robust and reliable. When the VESC 6 becomes more readily available, I'll buy at least one to "play" with.

Controllers burning out...anybody that has played with enough hardware and tried to push that hardware sufficiently has burned things out. I've destroyed a couple of controllers and a motor that way myself! I'm not sure I would call that "bad design" as much as I would call it abuse and exceeding the design limits of the device. I readily accept that when I blew up those 2 controllers and the motor that I was most definitely exceeding any reasonable limits for those devices. I bet your friend with the blown VESC, probably caused it's demise.

PowerVelocity controller review

10 MW

10 kW

100 GW

10 kW

10 MW

10 kW

10 kW

Administrator

10 kW

10 kW

10 MW

10 kW

10 MW

10 kW

100 GW

100 kW

100 GW

10 MW

100 kW

10 MW

10 MW

10 MW

10 kW

10 kW

10 MW

Similar threads