PowerVelocity controller review

Status
Not open for further replies.

ElectricGod

10 MW
Joined
Nov 1, 2015
Messages
3,404
Location
3 OH 3
Hi Everyone,

I've reviewed a couple of other controllers (ASI BAC2000 and Sabvoton MQCON), but this is the first time I've reviewed an entire product line. The only thing I lack is the 6 fet controller, but I've made suggestions on that too.

I've been working with Vadym (PowerVelocity) for a good while now on the controllers he sells here and on his web site. We've worked through a variety of bugs in the software and gotten what is a decent Chinese controller beaten into shape into a much better controller line-up. Vadym has done most of the work and I've been testing what he produces. In the past 6 months or so I've probably installed 20 versions of the programming app and reloaded or rebuilt monitoring/programming modules nearly as many times. I make no money from this review and my only interest in this product line is that it is a fairly well deployed series of controllers and Vadyms work has made them far better than the Chinese manufacturer ever thought about doing.

A few details and back story...
About a year ago I was looking for a 12 fet controller for a small scooter project and was generally unimpressed with what I found. I bought 2 controllers from Grintech, but they were too basic and had no programming at all. I looked at Lyens products and they were old/dated technology and didn't support FOC and various other things. i bought the BAC2000, but they are rather expensive at $300 a pop for what they classify as a 72 volt max, 2000 watt controller. It is a well developed controller with a few minor details and an impossibly steep learning curve. I was looking for something inexpensive that supported FOC and was programmable. That's when I happened upon Vadym AKA PowerVelocity. Originally I bought a 12 fet from him with an early iteration of the bluetooth programming module and APK. Things worked, but were buggy. I'm a software tester with a background in electrical engineering and playing with things is fun to me so I volunteered to help him out. If I've gotten paid for anything, it's in the form of a couple of newer BT module designs and better prices on controllers than he typically sells them for. Otherwise my efforts are purely voluntary and I have no profit margin or vested interest in seeing this controller line succeed. Later I got an 18 fet controller that I requested could operate at 150 volts max. It too included an early version of the BT programming module. Now I'm into my 3rd 12 fet controller, one 18 fet and now a 24 fet version. These controllers are highly hackable which is exactly what I was wanting.

My first 12 fet controller is now dead. I did something to it and blew up the MCU. That was no ones fault but mine. The second 12 fet is running strong. I'll talk about what I've done to them in a minute. The 18 fet controller is running well too. The 3rd 12 fet and the 24 fet controller have been in my possession for a few hours so I will talk about them later once I give them some love on my bench. The first two 12 fet controllers, I tested extensively on my Currie scooter. It has a 7kw C80100 outrunner on it. That rating IMHO is laughable since it's more realistically a 5kw motor. The 12 fet controllers come with what I think are smallish power and phase wires, but they also originally came with IRF4110 mosfets so that works out to about 2500 watts and the controllers did exactly that...2500 watts. The shunts limit the controller to 40 amps. I soon got bored with mediocre power levels so I pulled all the mosfets, shunts, power and phase wires and upgraded to AOT290 mosfets, bigger shunts and 10 awg wires. Now the controllers do 60 amps and I can say emphatically that the second controller which is still working great can do 4000 watts and not get hot at all thanks to my upgrades. I convinced Vadym to get the controllers redone with AOT290's since they are significantly better mosfets than the old tired IRF4110. From the factory they now come with AOT290's, but the shunts and wires are still smallish. The 12 fet I received just today has AOT290's in it. One of the brilliant features of these controllers is the MCU. It's totally programmable for just about anything you want to set including max battery and phase current. With my upgraded controllers, I no longer run them like they are 2500 watt controllers. I run them like they are 4000 watt controllers...thanks to the various upgrades I've done. The first controller is dead...so I used it as an opportunity to strip it down to the bare board. The below couple of pictures are of the power traces on these controllers. Notice how tall the power traces are. They have 2 layers of copper that add 1mm of thickness to all the power traces on the boards. I've taken apart 10-20 controllers and none of them have that much power trace reinforcement. If anything, they will have extra solder added or a thin copper runner in them. The added copper strips are as wide as the traces and 1mm thick! That's significant extra beefiness for a 12 fet controller. Both of my original 12 fet controllers got used and abused and modded a lot. No wonder one of them died!

This is top and bottom of my original 12 fet controller how I got it from Vadym. There is a decent amount of added solder to those power traces, but that's on top of 1mm of copper on top of the actual copper traces.

top%20of%20the%20board%20-%20original%20wiring%202_zpsh5ukri9r.jpg

Bottom%20of%20board%2014%20awg%20wires_zpscbptvyr6.jpg


I changed out wires, mosfets and few other things to get it to this. The shells were too thin so I added an additional heat spreader inside to help the mosfets get rid of heat better. The third and fourth pics are of the added heat spreader. That piece of 90 degree aluminum was added by me. In my load testing at close to 4000 watts, the added heat spreader has done a great job. BTW...cramming 10 awg wires into through holes designed for 14 awg wires is a very tight squeeze! when I upgrade the wiring on the new 12 fet, I won't be trying to cram all those tiny strands into the through holes, I intend to use a crimp fitting around the wire ends and crush it to fit into the through holes. It took hours of tediousness to get all those tiny strands into the wire holes.

top%20of%20board%20-%2010%20awg%20wires_zps6owg40pm.jpg

Bottom%20of%20board%2010%20awg%20wires_zpsicuvz3f2.jpg

Heat%20spreader%202_zpsm8emddvg.jpg

Heat%20spreader%203_zpslvemvgwy.jpg


Sometime after getting the original 12 fet controller I got an 18 fet controller. I needed more wattage and more voltage. I am converting an XB-502 moped to run at around 4000 watts and 130 volts. I asked Vadym to make me a controller that would max out at 150 volts. It cost a bit more to replace all the IRF4110's that came in it with IRF4115's and to upgrade the caps and few other minor details, but he delivered me a controller that worked as requested. I've run it at 147 volts and it was fine. I have no intention of running it continuously higher than 130 volts so that I have a decent margin for error. That controller hasn't seen lots of use and none under actual loads, but it has been bench tested many times on a variety of inrunner and outrunner motors. It too got larger phase and battery wires, but is otherwise exactly like he sent it to me. Since we have been playing with programming modules and an android app, that too has been swapped out multiples of times. The controller has worked in all my testing very well. Again, the traces are beefed up at the factory so I have no concerns of them limiting current flow or over heating. If I have a complaint about this controller, it's that the solder on all the power traces wasn't flowed with sufficient heat and so it's lumpy. That's the first picture. With the help of my Hacko solder station, I made quick work of that and reflowed all the solder so that it was not lumpy (second picture). AT the time the manufacturer wasn't making 150 volt controllers so Vadym modded this one for me with larger and higher voltage caps (third picture). These early controllers had thin walled cases so I added a heat spreader inside it's case too.

Cold%20solder%20work_zpsknshdpf0.jpg

Redone%20solder%20work%201_zpsryihywfj.jpg

big%20ass%20caps%201_zpsksvn1e0v.jpg


The final 18 fet controller with 10 awg wires and 5.5mm bullets.

Power%20and%20phase%20ends_zpsvpmyhwbt.jpg
 
So here is my great chance to rant and piss and moan about Vadym. LOL!

OK...not happening because I don't have anything to rant and piss and moan about. Have there been problems? Yes there have, but we've been developing a new differentiation in controllers that don't cost an arm and a leg, supports all the mid/high end controller functions and can be programmed and monitored on your cell phone and actually works. Vadym has been very responsive to my requests for changes and modifications and bug fixes. He has adopted AOT290's over the old tired IRF4110 for all the 100 volt controllers. Anything I've told him to improve or change eventually gets incorporated into the later versions. The original shells on the controllers were too thin so I was adding additional heat spreaders. The latest 12 fet controller now has a nearly 2X thicker shell on it. The BT module went form being super buggy and not working most of the time to a really slick app and working pretty darn well without glitchy behaviors all the time. IMHO, Vadym has been great to work with. There have been issues at times. For example my Alien Power C80100 worked well on my first 12 fet controller, but was super glitchy on the second controller. I think the first one was probably just a bit more forgiving than the second one and so the motor worked better there. I later replaced the halls on the motor and now there are no further problems and the motor works better on the second controller than it ever did on the first one. I think the factory motor halls were glitchy from the get go. He has another customer in another country that is having odd motor behavior. I tested my halls before they ever went into the motor. I think the other guy just swapped them out and has no idea if they were any good before hand or not. I think he still has some occasional motor issues. I honestly think it's his halls and not the controller. These controllers have lots of capabilities and features from the factory and Vadym has just added more and more to them. I've bought 5 controllers from him and I'll probably buy a few more in the future as the need arises.

Shameless plug for PowerVelocity/Vadym and his controllers...
OK OK...there are better controllers out there. BUT for the price and the features you get in these controllers...good luck! The only things I wish they had was variable regen where you can twist a second throttle to increase or decrease regen strength and throttle curves. It's pretty much a linear throttle all the time with some very minor low speed throttle damping. If you set regen braking too high (that's regen when you let off the throttle) and then don't do throttle damping, the EV becomes virtually uncontrollable as it goes from accelerating to stopping and back to accelerating all within seconds. However that's your fault, and not a controller malfunction. That's the worst things about these controllers. The phone app has come a long way and is a really nice app now. You no longer need a CA or watt meter or speedo. Soon you wont need a motor/battery or controller temperature gauge either. The app shows you how much wattage you recover via regen, battery voltage and a battery guage. it shows amps used, Kw used, speed and lots more. You can set real parameters so that the meters show actual data based on your packs real life capabilities. Motor RPM and speed is calculated on a hall signal, motor poles and so on. Both are displayed on screen. You can reconfigure settings in the controller if you want to tweak things all over bluetooth communications. I have to say that I've messed with a few controller apps that run on Android and most are garbage. This app has active development going on and is already a great tool on your EV. I'll probably always have a watt meter on my EV's, but all the rest of the stuff I use to monitor with will be going away on future builds that use a PV controller. They are simply NOT needed anymore. Anyway, before you look at other controllers, talk to me or Vadym and then give Vadym your money. You won't regret it. BTW... I get nothing for doing this review and for saying good things about Vadyms work.
 
I blew up my original 12 fet controller for no fault than my own. Even really stellar hardware can be destroyed. The controller failure is 100% on my shoulders and NOT due to anything that went wrong in the controller. I killed it. Anyway, that meant I needed another 12 fet controller so I asked Vadym to get me one. I haven't powered it up or anything yet...just took it apart to take a few pictures before I give it some bench test love.

The first thing I noticed was that the shell is a good bit thicker. Also, Vadym and I have been going back and forth on the telemetry module which is not in the pictures since mine is a mock-up/prototype. We are pretty sure this latest iteration will be the final one, but I told him mine wouldn't be in any pictures so that's how it goes.

Nice thick 1mm copper power traces buried under a good bit of solder.

12%20fet%20board%20bottom%201.jpg

12%20fet%20board%20bottom%202.jpg


I think the wires are thicker now, but I'm not sure yet. And 12 AOT290 mosfets...brilliant! That black connector is where the telemetry module plugs in. He will be securing them with silicon and possibly a different connector. Mine is prototype work so the final production model may vary a little from this. However he is selling this design right now I believe and it is fully functional with more cool stuff coming soon. The module that plugs into that connector has endless potential for all sorts of things.

***NOTE: these might happen so don't hang Vadym or me for what I'm posting in this note. The module can with some programming monitor your BMS, motor temperature, pack temperature and so much more. This sort of thing will come as the time and desire for it evolves...maybe.

12%20fet%20board%20top.jpg


Unfortunately I don't have a side by side shell comparison, but suffice to say that this shell is close to 2X thicker than the older shells. I don't think I need to add an extra internal heat spreader any more.

12%20fet%20new%20shell.jpg


This is the latest app on android and if you go to Google Play and search for PowerVelocity, you can install it on your phone too. However it only works on PV controllers or on ones that have one of Vadyms clever little modules. You can adapt the module to at least do everything a CA will plus lots more, but the controller programming functions are specific to the XCKJ3232 MCU. IE: if you have some other controller that uses this MCU, you might be able to program it via Vadyms app and module. I know I've tried a controller from elsewhere that had this MCU and it worked, but that doesn't mean that it will always work.

This is your status screen. It may get some new stuff over time or some changes to it's layout. You get a digital speedo and an analog one on the left side of the dial. The right side of the dial is regen on the bottom and and wh used on the top part. you also get digital version of this informations. There's motor RPM, battery voltage, current amp draw, a battery gauge, wh used per mile and more. Vadym has really done some great work here!

Latest%20app%2014092017%201.png


This is the app set up page so you can set all your values for the status page.

Latest%20app%2014092017%204.png


This is the controller programming page. It takes up more than a single screen space so it's several screen shots. Those speed and power settings are kind of cool. They allow you to set up 3 profiles in the controller to limit torque and top speed that you then select via a 3 position switch. This is perfect for handing your EV to a noob. Set up a low power/low speed profile and they can't get into much trouble. Then flip the switch to your high power profile and go back to riding like you want. Soft start allows you to make low throttle positions a little more friendly so that launching off is less harsh. Imagine putting 4000 watts of power on a 60 pound stand-up scooter and you will want a softer start up so that the EV doesn't leave you on your butt.

Latest%20app%2014092017%202.png

Latest%20app%2014092017%203.png
 
I got this controller for a future 15-20kw project. Right now all I have is the 18kw motor and now the controller that will drive it. I tend to do multiple project streams in parallel. When I get blocked on the XB-502 moped or the other moped build I've got in the works, then this controller will see work done to it. Over time it will get the things done to it to make it into the real beast I am looking for. The controller currently has 24 IRF4115 mosfets. Inside the shell is lots of room so I will be adding at least another 18 mosfets and more if I can cram them in. This controller comes from the factory already set up for 150 volts max. The whole controller is about the same size and format as the older style Sabvoton 24 fet controllers. It's not small, but then what high wattage controller is?

The top of the shell where the heat spreader screws down is 1/4" thick. The rest of the shell is 1/8" thick, plus the fins on top of that.

Shell.jpg


The heat spreader is all 1/4" thick aluminum.

Populated%20board.jpg


This is an unpopulated board so you can see where all the electronics fit together.

unpopulated%20board.jpg


The board has all the trace reinforcement you could ever want in all the right places. All the main wires are 8 awg silicon. That's all 1mm thick copper reinforcement on the traces.

Board%20bottom%202.jpg

Mosfet%20Buss%20close-up.jpg


This is my one complaint...if you want to call it that. The bottom of the board where +batt connects in is a bit starved for solder, but on the top of the board there is plenty and that's where the wire is soldered down.

batt%20connection.jpg


Small wire connections.

small%20wire%20connections.jpg

small%20wire%20connection%202.jpg


Top of the heat spreader.

heat%20spreader.jpg


Dc-DC converter. It converts batt voltage to 5 volts.

batt%20%20power%20buss.jpg


+batt copper buss and half the IRF4115 mosfets.

batt%20%20power%20buss%202.jpg


-batt copper buss and the other half of the mosfets.

batt%20-%20power%20buss.jpg


150 volt filter caps throughout.

filter%20caps.jpg

filter%20caps%202.jpg


I will be making daughter boards that mount more mosfets. I have 2 spare heat spreaders that I can mess with. Inside the shell is a large gap along one side. There's enough room there to mount 18 mosfets to a heat spreader down the length of the shell. Also the existing heat spreader inside the controller is quite large. It wouldn't be hard to mount another 24 mosfets to it as well. Ideally, I can turn this 24 fet controller into a 66 fet controller or conservatively a 16kw controller using IRF4115's. The phase and battery wires terminate at the shell in 6mm screws. The exisiting terminals aren't large enough for all the extra wattage I intend for this controller so I will need to beef them up a fair bit. I intend to solder the phase and battery wires directly to the back side of the screw terminals and have separate shunts on each daughter board. This will mess with current sensing. I bet I can get Vadym to make me some custom settings so the module can handle more than one current test point and then add them together for total current. The idea is to NOT increase the physical foot print of the controller and yet have close to 20kw. Vadym has some requests for higher wattage controllers and I of course want to make one so this is going to kill multiple birds with one stone. The app will need to have a 66 fet controller option in it. After I do the development work and max out what can be done with this form factor, then he can duplicate it or pass those high power users on to me for the custom work.
 
OK. lets face it, this is a Chinese MCU and not the best MCU there is. However it does work pretty well and is good to about 50,000 eRPM. I have the datasheet for an older MCU in the same product line. The older MCU doesn't show a crystal or R/C circuit or clock pin for the MCU. It may be that the 3232 has this externally, but my suspicion is that it has an internal programmable clock.

MCU%20close-up.png


There are several things I really like about this chip.
1. Fully programmable for battery and phase amps. If you want to to use a 12 fet controller to make yourself a 600 fet franken controller, well go for it.
2. Decent set of functionality for setting up user profiles. IE: 3 sets of power and speed levels.
3. Field weakening...admittedly it's only 30%, but in this case too much of a good thing is actually bad. My ASI BAC2000 can do 160% and yes that's insane!
4. Adjustable regen and throttle braking. It can't do adjustable regen via a second throttle, but you can set regen to whatever strength you want and leave it there.
5. Regen voltage limiting. I have a Kelly controller that can set the strength of regen, but won't cap the max voltage of regen. As a result, from time to time it does regen too high and that disables the controller.
6. The LVC really works. On the 12 fet in my Currie, it is set to 48 volts on a 16S pack. Right at 48 volts on my watt meter, the controller is shutting off.
7. Reverse via a switch and then a setting in the MCU.
8. Sinusoidal phases

Things that are a little bit meh...
1. It is a sinusoidal controller, but they are actually stepped so not a pure sine wave as some better MCU's
2. No variable regen via a second throttle
3. There are times when 50k eRPM is too slow. For example a 14 pole motor at 130 volts and 50 kv is 91,000 eRPM. This MCU will lose sync with the motor.
4. No motor discovery. You have to plug it all in correctly to get it to run. This is minor, most controllers in this class or less don't do motor discovery.
5. No throttle curves.
6. This isn't really the MCU, but the controllers have wiring positions for a status LED, but then there is no LED.
7. This is an old gripe that in the current MCU's is fixed. They now support sensorless motor control. Lose a hall or halls and just unplug the hall cable and keep going.
 
I have pics of both and putting them side lets you see the shell differences.

Old shell. Thinner walls and minimal fins on the bottom.

Heat%20spreader%203_zpslvemvgwy.jpg


New shell. The walls don't look it, but they are a good bit thicker. The shell is slightly wider so there is about 1/4" more space inside than before. With the square internal corners, adding a secondary heat spreader, if it's needed, will be straight forward.

12%20fet%20new%20shell.jpg
 
Nice review. Great information and good progress. I have one of these controllers in the "to do" box, but I guess it's already out of date. :)

You mention these are stepped sine, as opposed to pure. How does this compare with ASI or Sabvoton? I would think at some level they are all stepped, is there a difference in the granularity?

Are these true FOC, or one of the other Sinewave algorithms?
 
Alan B said:
Nice review. Great information and good progress. I have one of these controllers in the "to do" box, but I guess it's already out of date. :)

You mention these are stepped sine, as opposed to pure. How does this compare with ASI or Sabvoton? I would think at some level they are all stepped, is there a difference in the granularity?

Are these true FOC, or one of the other Sinewave algorithms?

The only place you might notice a difference is on a hub motor, but it's sufficiently sinusoidal that the difference between true sine and this is not very great. If this was your first sinusoidal controller, you wouldn't know it all. If you have a true sinusoidal to do a side by side comparison, the difference would be small to essentially irrelevant.

FOC boils down to field weakening and how you detect phases. and these MCU's do those things so while it may not have some of the more blue sky, off the charts FOC capabilities of say the VESC or ASI MCU's, it's good enough for what most folks want it for.

If you don't care about sensorless operation which is firmware in the MCU, then that's the only thing that is external to Vadyms work. You can ask Vadym for a new telemetry module and that will otherwise get you back up to current status. Worst case, the app will still program your controller with the older modules. Right now there have been 4 iterations of the module. The first module had blue tooth issues so the BT radio was replaced with a better one, the software in the module has seen loads of iterations as new functionality or bug fixes have been implemented. The brain (nano) has remained the same since the beginning, but is running very different software now. Recently have been a couple of iterations of a daughter board that interfaces with the controller for the actual monitoring of various things that the nano can't do. The current telemetry board will probably hang around a good while. It has lots of as yet untapped potential. If you get a current module it will include the latest telemetry module which does everything a CA will do and with additional programming can do lots more. It also has some currently unpopulated places on it for additional components for more "stuff"...if it's needed. We've talked about developing all sorts of things for the telemetry board to handle, but things get prioritized and Vadym usually has his finger on the pulse where it matters most.
 
I thought I would mention some motor issues I've had or that I know about. It's my opinion that less than wonderful halls are the problem in all of these situations

1. I have an Alien Power C80100. It worked OK on my original 12 fet controller, but was quirky at best on my second 12 fet. In the end, I added a second set of halls in the motor and now it works perfectly.
2. I have an Alien Power 12090. It never worked right with any PV controller until I added new halls. Now it works great on any controller I have tried.
3. I have an AstroFlight 3220 at 113kv. My original hall placement was wrong and that created issues for any controller. Recently I redid the halls and now it works great.
4. I know of someone who has a bafang motor like Lunacycles sells. He's had problems. I told him to replace his halls, but he didn't test them before installation so who knows if the new halls are good or not.

Here's a list of known good configurations that I have tried. You will notice a theme here...good halls make a big difference and marginal halls will make for quirky results.

1. Alien Power outrunners with good halls.
2. LightningRods/HLD small block and big block. This has been tested by me and Vadym independently.
3. AstroFlight 3220 with properly placed/good halls
4. Pretty much any hub motor. I have 3 hubs of various manufacture. Vadym has tried quite a few too with good results.
5. BOMA inrunners...assuming the halls are good.
6. Turnigy CA80-160kv outrunner with good halls.

Halls are the single biggest thing that has caused me troubles with various motors on PV or any controller. If you are wondering what good halls are, go read my hall testing thread. Don't assume that because an LED lights up that your halls are good. In this thread I talk about a motor and controller tester tool. It has LED's that turn on and off to indicate functionality. I've tried all my motors on it at various times and found that the halls always light up the LED's, BUT then I try to run a motor and it doesn't run right. Any motor that gets halls in it from me, first gets the halls thoroughly vetted before they are installed. Name brand or cheap Chinese knock-offs seem to have significant failure rates. I don't trust any hall until I see it working independent of a motor or magnets and then behaves predictably in the presence of magnets. It's my opinion that people assume far too much when it comes to halls. They go to digi-key and over-pay for halls and assume that since they come from a reputable place that it means they got good halls. I don't believe it until I see it for myself. I buy a fair number of AOT290's. They are a great mosfet, but in testing them, the specs may be one thing, but in any particular batch, there will be some that are out of spec or marginal in some way. I use the best of them in controllers and the rest in less important roles. Halls are no different and in fact worse. "Test and see that I am good" (Psalms 34:8). And everyone thought that an ancient book was irrelevant when it comes to EV's. OK...99.999999999% irrelevant.

Link to hall testing thread:
https://endless-sphere.com/forums/viewtopic.php?f=2&t=83802
 
Could you share a link to the thread about the Hall Sensors, or tell us more here?
EDIT: i have seen the link :lol:

When swapping the halls on my MXUS 3k DD motor to honeywell SS411A, i noticed much better efficiency (not so quick overheating at abuse).
 
madin88 said:
Could you share a link to the thread about the Hall Sensors, or tell us more here?
EDIT: i have seen the link :lol:

When swapping the halls on my MXUS 3k DD motor to honeywell SS411A, i noticed much better efficiency (not so quick overheating at abuse).

You probably had a hall or 2 that were not reading clearly and the controller wasn't able to run the motor as effectively as it should. Maybe the controller had timing off a bit as a result? Also, some controllers will run with a bad hall, but with reduced performance or will switch to trapezoidal if they are sinusoidal. Trapezoidal is a bit less efficient than sinusoidal so that may have been why you saw quicker over heating. Generally speaking the only place that sinusoidal vs trapezoidal is noticeable is on hub motors. Run just about any inrunner or outrunner and you probably won't be able to tell it is running trapezoidal or sinusoidal. On a hub motor, there's a small sound difference between the two.
 
Alan B said:
An FOC controller should not use the halls above a certain speed, it computes the rotor angle from the phase currents with much greater precision than the hall sensors resolve. Halls may be used for starting and at very low speeds, where it is not actually running in FOC mode.

OK...thanks for the info, but what does that have to do with a controller review?
 
Thanks for posting this review. I'm just trying to understand this sinewave controller. It seems to be of the Non-FOC Sinewave design type:

http://www.motioncontroltips.com/faq-whats-the-difference-between-field-oriented-control-and-sinusoidal-commutation/ Sinewave vs FOC Control


This is a sinewave controller that is often quoted as having FOC (by users), but the characteristics reported don't quite seem to match FOC type controller characteristics. There are many sinewave controller algorithms, not all use Space-Vector control, aka Field Oriented Control. Nothing wrong with that, just not quite the same characteristics and features (or cost).

FOC is not mentioned in the PowerVelocity descriptions of these controllers (as I recall). Users mention it from time to time but this is generally not responded to, at least not that I have seen. You didn't answer it when I asked earlier. So this seems to be a point of confusion.
 
Alan B said:
Thanks for posting this review. I'm just trying to understand this sinewave controller. It seems to be of the Non-FOC Sinewave design type.

This is a sinewave controller that is often quoted as having FOC (by users), but the characteristics reported don't quite seem to match FOC type controller characteristics. There are many sinewave controller algorithms, not all use Space-Vector control, aka Field Oriented Control. Nothing wrong with that, just not quite the same characteristics and features (or cost).

FOC is not mentioned in the PowerVelocity descriptions of these controllers (as I recall). Users mention it from time to time but this is generally not responded to, at least not that I have seen. You didn't answer it when I asked earlier. So this seems to be a point of confusion.


I said this earlier.
"FOC boils down to field weakening and how you detect phases. and these MCU's do those things so while it may not have some of the more blue sky, off the charts FOC capabilities of say the VESC or ASI MCU's, it's good enough for what most folks want it for."

I've also said that this MCU is not in the same league as other MCU's so if you are looking for something that truly implements FOC, then look elsewhere, but the important details are there. I think I was pretty specific and honest about the 3232 MCU.
 
Awesome post and great info.. thanks a lot for posting this up.
 
Since you have some high pole/high rpm motors lying around.. have you tried overclocking the MCU?

I have an eZee motor in a 20" wheel here and an old EB3.. at 39mph / 72v, i'm at 52,000rpm and the controller will choke..

I need another 10,000 erpm..
 
neptronix said:
Since you have some high pole/high rpm motors lying around.. have you tried overclocking the MCU?

I have an eZee motor in a 20" wheel here and an old EB3.. at 39mph / 72v, i'm at 52,000rpm and the controller will choke..

I need another 10,000 erpm..

The 12090 I have is probably the highest eRPM motor I have tried. It's 14 poles and 50 kv at 72 volts that works out to 50,400 eRPM. It runs on these controllers at 72 volts or less, but loses sync at higher voltages. I've run it up to WOT at 66 volts with no problems under no load. I don't know what the Kv or poles of your motor is, but if you stay at 50k eRPM, you should be OK. I looked at the datasheet that I have and I didn't see anywhere that it took an external clock signal. My guess is that it has an internal clock. IE: overclocking doesn't appear to be possible. BUT I have to admit that finding the actual 3232 datasheet has been fruitless. I have the 3132 data sheet and I'm assuming that the 3232 is very similar.

If you need more eRPM than that, then ASI, or Sabvoton or Kelly all have high eRPM controllers. I'm not sure why this is, but hub motors do better at higher eRPM's than inrunners and outrunners do on these controllers. IE: You may get away with a good bit more than 50,000 eRPM.

I have a 4kw QSmotor hub at 13kv, 16 poles and at 82 volts that works out to 17,056 eRPM. I can't imagine why your hub is running at 50,000 eRPM or higher. Maybe you have the version with with planetary gears? I imagine they have fairly high Kv's.

These controllers have seen several revisions in the MCU coding. Since I have 2 brand new controllers that are waiting for me to test them out, in a few days, I'll connect one up to my 12090 at 82 volts and see what happens. That works out to 73,800 eRPM. If the motor runs up to WOT under no load, then there's no reason why it won't work for you too.
 
Cool, thanks.
The eZee is basically a slightly narrower and lighter MAC, but still has 16 poles with a 5:1 internal reduction.
39.6mph on 20" = 662 rpm x 5 = 3310rpm x 16 = 52,960 eRPM.

I'm thinking i might have to go the way of the phaserunner... but damn, if i don't like the simplicity of these infineon clones and love that noise of a trap controller for some reason :)
 
neptronix said:
Cool, thanks.
The eZee is basically a slightly narrower and lighter MAC, but still has 16 poles with a 5:1 internal reduction.
39.6mph on 20" = 662 rpm x 5 = 3310rpm x 16 = 52,960 eRPM.

I'm thinking i might have to go the way of the phaserunner... but damn, if i don't like the simplicity of these infineon clones and love that noise of a trap controller for some reason :)

Trapezoidal on a hub does make a cool sound that you lose a good bit of on a sinusoidal controller.

Look at the Kelly controllers. They have a really large controller selection. I've never personally tried a phase runner, but they are a 6 fet controller based on the AOT290 so the wattage they are capable of is limited to a maximum of 2000 watts. I have no idea if you can get the full capabilities out of the mosfets in that controller or not. It's just an opinion, but barely enough controller to meet my needs is a significantly limiting factor. I don't own any 6 fet controllers for this reason. It's just not enough wattage for anything I'm likely to ever want to build. It's better to have it and not need it than to need it and not have it. Anyway, the phaserunner in a 12 or 18 fet version would be brilliant. I'd buy one just to try it out. I essentially did that with the ASI and Sabvoton controllers I have. AND, that's how I got into these controllers. The compelling difference was pretty simple. They got me basic FOC (Feild weakening), sinusoidal control, are very hack-able and they cost a good bit less than anything that does what they do. Anyway, Kelly will have something that does what you need. I have 2 Kelly 10kw controllers which is overkill for you, but they also have lower wattage controllers. You will want to get the high speed option. Any of the KBS series are trapezoidal, but well built controllers. I have a KBS72221 that I have been using for almost 3 years now. Kelly is conservative with their numbers. They say their controllers work at 72 volts, but I know for a fact that the 72 volt controllers work just fine at 82 volts or 90 volts max. If you do get a Kelly, then buy direct from them. They are good folks and will not do you a dirty deal. I have found that their trapezoidal controllers tend to be more lenient when it comes to marginal halls or noisy motors, while their sinusoidal controllers are much more picky about that sort of thing.

This is straight from the KBS manual. You will want the high speed option.
"Up to 40,000 electric RPM standard. Optional high speed 70,000 ERPM."
"Controller supply voltage range: PWR, 18V to 90V for controllers rated equal or lower than 72V."
 
PhaseRunner Notes (BAC800)

compact FOC motor controller
waterproof
90V max (36-72V Nominal Battery voltage)
96A peak phase current (firmware limited)
40A continuous phase current (nominal, depends on heatsink)

info from PhaseRunner thread on ES

one user reports max battery current of 43A reached with MXUS 4T 45mm
measurements indicate 50A phase current clamped to bike tube, and with a better heatsink up to 70A


The excellent PowerVelocity controllers use Sinusoidal Commutation rather than the mathematics of FOC. Mild Field Weakening doesn't require FOC. Different algorithms have different costs, benefits and capabilities.

For more information on Trapezoidal, Sinusoidal and Field Oriented (Space-Vector) Motor Controls see papers such as this one: http://www.copleycontrols.com/Motion/pdf/Field-Oriented-Control.pdf
 
Alan B said:
PhaseRunner Notes (BAC800)

compact FOC motor controller
waterproof
90V max (36-72V Nominal Battery voltage)
96A peak phase current (firmware limited)
40A continuous phase current (nominal, depends on heatsink)

info from PhaseRunner thread on ES

one user reports max battery current of 43A reached with MXUS 4T 45mm
measurements indicate 50A phase current clamped to bike tube, and with a better heatsink up to 70A


The excellent PowerVelocity controllers use Sinusoidal Commutation rather than the mathematics of FOC. Mild Field Weakening doesn't require FOC. Different algorithms have different costs, benefits and capabilities.

For more information on Trapezoidal, Sinusoidal and Field Oriented (Space-Vector) Motor Controls see papers such as this one: http://www.copleycontrols.com/Motion/pdf/Field-Oriented-Control.pdf

I love specs for stuff. Almost always there is some exaggeration going on or at least incomplete information that makes product X look so much better than it really is. Name a motor manufacturer that doesn't hype their wattage numbers. Name a controller seller that doesn't hype their power numbers. It's industry wide and rife. The phaserunner specs are no different. The ONLY time you will get that 40 amps is at 48 volts or less AND that's maximum amps, not continuous as I explain in the next paragraph.

Lets look at a little math. The Phase runner uses 6 AOT290's which is a great mosfet. The phaserunner is a great controller. As we all know, at any point in time a maximum of two motor phases and therefore 2 sets of mosfets are on and the other phase/mosfet set is off. It's not really quite that simple as there are H bridges in play that control current flow in the motor windings. But simply speaking any controller will have mosfets divisible by 3 and at any point in time no more than 2/3 of all the mosfets are powered. So back to the AOT290. It is an excellent mosfet and in fact in the 100 volt range, my favorite. I wish more companies used them in their controllers. They are the only 500 watt mosfet in the TO-220 package that doesn't compromise other specs to get there. At 25C you can run them at 500 watts and at 100C 250 watts. So lets assume that you can keep those mosfets at an ideal 25C or 77F, then that means they can run at 500 watts MAX each. Then the next question to ask is what voltage are you running them at. The greater the voltage, then less amperage. The lesser the voltage, then more amps. BUT always equaling 500 watts. This is simple math...volts x amps = watts. So run that AOT290 at 77F and 500 watts and 42 amps and you get 11.9 volts. Who is going to run at 12 volts? Obviously no one and there are really 4 mosfets, not 1 running at a time in a phaserunner. Lets assume that this mystery person wasn't exaggerating and that they are running at reasonable EV voltages such as 48 volts. Then that means 500 watts at 48 volts = 10.4 amps per mosfet. Since there are 6 of them that means 4 are actually on at any moment in time which comes out to 41.6 amps. That's close enough to 42 amps which is a maximum rating for four of these mosfets. So then, lets bump to 16S/66 volts...that's 30 amps max or 20S/82 volts...that's 24 amps max. The AOT290 IS an excellent mosfet, but those numbers are maximums and based on actual specs for this mosfet from it's datasheet. Sure you might get away with running them harder for a little while and certainly controlling heat will help, BUT they ARE 500 watt mosfets under ideal conditions. All I can say is good luck running them over their specs. They will last a little while and then die. Running anything...mosfets, car engine, motors...will greatly shorten it's life span and leaves no margin for over load conditions. By all means, run those mosfets at their maximum rating and see how long they last.

So anyway, this is a thread about the PowerVelocity controllers, not phaserunners and others.

Edit on this post...

THIS IS WHAT YOUR WATT METER SHOULD NOT EXCEED and has nothing to do with what is happening on the non-linear side of the controller or what the phases are seeing.

Lets look at what you should realistically see as maximums for the PowerVelocity controllers under ideal conditions (25C or 77F). This stuff isn't rocket science and applies to any controller from any manufacturer. I'm not referring to phase amps or what happens on the non-linear side of the controller. This is straight up what your watt meter will read that sits between your controller and battery pack. I'll let you divvy up those wattage numbers by your actual operating voltage to get what your real maximum amperage should be. Notice how I keep saying "maximum". I want to be clear that these are your upper limits at which point things start blowing up and the magic smoke in your components comes out. Always plan to leave a margin for error. I can rightfully say that the PV 12 fet running AOT290's is a 4000 watt MAX controller because that's the REAL specs of the mosfet. BUT I will never tell someone to run it at 4000 watts. Instead I will tell them, to run it at 3500 watts so there is some margin for error.

12 fet (IRF4110): 370 watts x 8 mosfets = 2960 watts maximum
12 fet (AOT290): 500 watts x 8 mosfets = 4000 watts maximum
18 fet (IRF4110): 370 watts x 12 mosfets = 4440 watts maximum
18 fet (AOT290): 500 watts x 12 mosfets = 6000 watts maximum
24 fet (IRF4115): 375 watts x 18 mosfets = 6750 watts maximum ***

*** the 24 fet controller is rated to 150 volts max so it doesn't come with the AOT290.

NOTE: Whatever controller you have from whatever manufacturer is making it can hype the numbers all the want, but the cold hard facts are that your controllers real specs will be exactly like the list I just made. I have a Sabvoton 24 fet with IRF4115 mosfets...that's a 6750 watt MAX controller. I have a 24 fet PV controller. It too is a 6750 watts max controller. Live it, know it, believe it. The mosfet specs ARE the mosfets specs and no one can exceed those specs for long without blowing up their controller.
 
Sorry, but that math is not accurate. The FET device rated dissipation is a rating of the heat path out of the die, and has little to do with the FET power handling capability in switched mode, which is the way they are operated in these motor controllers. Doubling the number of FETs does NOT double the power handling capability of a controller due to mismatch and current imbalance. The more FETs you add, the less benefit you get from each additional FET due to the uneven current division.

The numbers I mentioned for the PhaseRunner are actual user measurements from real tests, not theoretical calculations or pie in the sky specs. In my own experiments I have run 25 amps at 80+ volts and the PhaseRunner did not even get warm. It was not close to any limits. Most controllers don't drive their FETs well enough to reach full device capability, and they use the inherent body diode which generates more heat than the FET itself, limiting the system performance (as well as other limiting design flaws). The PhaseRunner doesn't have these faults, so it has more capability than a 6 FET controller generally has (and has a higher cost due in part to the components [and engineering] required to make this happen). It is in a different class.

The PowerVelocity Sinusiodal controller provides some of the features at a lower price point and with less tuning complexity. They are different products at different price points. I don't really view them as directly competing. If I need small size, waterproof, FOC, torque throttle, real time variable regen/braking, or extreme field weakening then the ASI/Grin PhaseRunner is appropriate to consider (or the larger ASI units if more power is needed). If I need something that is lower in cost and still sinewave quiet then the PowerVelocity is a good one to consider, if real-time variable ebraking is not needed. I'm not sure but I think the PowerVelocity doesn't have torque throttle either. When I wanted high power, FOC sinewave control, torque throttle and smooth full range electric braking I ended up with the Sabvoton. Different devices for different requirements.

It would be better to stick to actual experience rather than theoretical bashing of other products in this review thread. I responded due to incorrect theory and presented some actual data about the PhaseRunner since it was being discussed. The PhaseRunner data (and more) can be found in it's own (long) thread.
 
Status
Not open for further replies.
Back
Top