under watt controller
- Thread starter Rafabaca
- Start date
Jeremy Harris
100 MW
texaspyro said:
Sorry, but this is incorrect. It won't let any magic smoke out at all; it's something I've been doing all the time for the past couple of years with only good side effects.
Motors (whatever their nominal rating) will, if unlimited, draw as much current as they can, particularly under low rpm, high torque conditions, they all rely on the controller to limit the current/power to them. The motor rating is a pretty vague nominal figure and the motor cannot control the amount of current it draws in any way, it relies on the controller to stop it from burning out.
All the controllers we use (excepting some RC-type ESCs, which doesn't seem to be the OPs issue) have current limiting that operates a lot of the time in normal use. All that will happen with a 600W controller on a "1000W" motor will that the motor will only deliver around 600W, rather than it's nominally rated 1000W, it will run cooler (a good thing, generally) and the efficiency may be a bit better (another good thing) because the motor I²R losses will be lower than when run at its full potential.
Under-running motors is exactly what I do with my electric boat drives, but I take it to extremes to get good efficiency and reliability. I'm running BLDC motors with a nominal rating of 2,800W at around 120W, using an ordinary ebike controller. The controller is setting the power level, as they always do.
Jeremy
I may be pretty new to all this, but i think what Jeremy is saying is spot on. People often upgrade their controllers and run more amps and more volts on motors for more power, so the controller is the limit. A BLDC motor will certainly eat amps as fast as the battery can provide if a current limiter weren't included, hence a smaller controller = basically a more prudent current limiter.
Also, i don't see how running low wattage or voltage to a BLDC motor would hurt the motor either. It is essentially the same effect as holding a 1000w controller at roughly 60% throttle.
I have the ultimate cheap Chinese controller, a hua yu xing feng ( no info about it on these boards!! ).. the current limiting is spot on according to my turnigy watt meter. I dunno about this whole, questionable current limiting concept.
Also, i don't see how running low wattage or voltage to a BLDC motor would hurt the motor either. It is essentially the same effect as holding a 1000w controller at roughly 60% throttle.
I have the ultimate cheap Chinese controller, a hua yu xing feng ( no info about it on these boards!! ).. the current limiting is spot on according to my turnigy watt meter. I dunno about this whole, questionable current limiting concept.
Jeremy Harris
100 MW
texaspyro said:
I'm afraid there aren't. All BLDC ebike controllers have to operate in current limit mode at start up and low speed, high torque demand, if they didn't, they'd go up in smoke. The motor is close to being a dead short at low speed, high torque, so unless something acts to limit the current things will blow.
Take the common Crystalyte hub motors as an example. The middle range 406 has a winding resistance of around 0.36 ohms. Running on a typical 36 volt system it would draw about 100 amps at start up/stall, unless there was something acting to cut the current back. Even the highest resistance Crystalyte motor has a winding resistance of only 0.57 ohms, not enough to limit start up current to a safe value for teh majority of off-the-shelf ebike controllers.
What happens in reality is that the controller detects the high supply current and reduces the duty cycle of the PWM to limit the current to a safe value, maybe 20 to 40 amps or so. All BLDC ebike controllers do this; I've probably taken apart a couple of dozen and each and every one has had pretty good current detection and limiting built in. The only BLDC controller I've ever seen with no current limiting is a cheap RC ESC. They can get away without current limiting because they are designed to drive propellers, that have a cube law power/rpm curve (so don't present a high torque load at low rpm).
The motor has only two ways of limiting the current it draws; either it spins fast (so increasing the BEMF and reducing the potential difference that's driving current through the motor) or by increasing the winding resistance to the point where the current cannot exceed a safe limit. This latter approach would be so inefficient as to be useless.
Jeremy
This is probably the most welcoming form I've ever been to.
I've been looking for a cheap 48v controller to work with my conhis 1000w. right now im looking at this one
http://www.virtualvillage.com/004616-202.html
I'm a full time student so cheaper is better.
The original controller worked for a few days, then started to jam the wheel. Kentlim has been walking me though a simple diagnosis and repair process but I don't really know how to repair a controller.
If you guys know of any cheap controller that would work with conhis bldcs, please feel free.
I'm thinking of just buying a stop gap, then saving up for a (tm) Magic Brussless-Controller
http://www.goldenmotorusa-d.com/index.php?pg=acc
does anyone know about the quality of it?
I've been looking for a cheap 48v controller to work with my conhis 1000w. right now im looking at this one
http://www.virtualvillage.com/004616-202.html
I'm a full time student so cheaper is better.
The original controller worked for a few days, then started to jam the wheel. Kentlim has been walking me though a simple diagnosis and repair process but I don't really know how to repair a controller.
If you guys know of any cheap controller that would work with conhis bldcs, please feel free.
I'm thinking of just buying a stop gap, then saving up for a (tm) Magic Brussless-Controller
http://www.goldenmotorusa-d.com/index.php?pg=acc
does anyone know about the quality of it?
grindz145
1 MW
Depending on how familiar you are with electronics...I would check the hall sensor signals to the controller. It's possible that you're hall effect sensor connections are broken in some way, and if it is on the motor side then changing the controller out won't help. It seems odd that you'd blow up a controller in a matter of days.
Jeremy Harris
100 MW
It sounds as if you either have a shorted FET in the controller, or, perhaps, a bad Hall sensor in the wheel. My money is on a shorted FET.
That controller from Virtual Village is probably a XieChang from the look of it. This one from Keywin Ge :http://cgi.ebay.com/48V-600W-brushless-controller-E-bike-scooter-/300352534034?pt=LH_DefaultDomain_0&hash=item45ee67f612 is definitely a XieChang 12 FET controller that we know to be easily upgradeable to over 1kW, by changing the FETs. You can run it as-is, then when you get some spare cash, just change the FETs and maybe one or two other components and quite significantly increase the power capability. With the right mods, one of these controller will work at up to around 90V and 50 to 60A, so into multi kW territory.
Jeremy
That controller from Virtual Village is probably a XieChang from the look of it. This one from Keywin Ge :http://cgi.ebay.com/48V-600W-brushless-controller-E-bike-scooter-/300352534034?pt=LH_DefaultDomain_0&hash=item45ee67f612 is definitely a XieChang 12 FET controller that we know to be easily upgradeable to over 1kW, by changing the FETs. You can run it as-is, then when you get some spare cash, just change the FETs and maybe one or two other components and quite significantly increase the power capability. With the right mods, one of these controller will work at up to around 90V and 50 to 60A, so into multi kW territory.
Jeremy
DAND214
10 MW
Just because you replace the controller doesn't mean it's fixed. As posted it can be just a loose or broken hall sensor or a loose or shorted phase wire.
Most controllers are strong enough for the 1000w hub. You could run a 250w controller on it but the performance would be a disapointment.
If you need a controller look in the new for sale section and LYEN sells a good controller.
Also buying a controller you will need to figure out the wiring sequence. Some but not most are plug and play but then wires don't match or plugs don't match.
You might doudle check your connections for loose or broken wires before you jump to a new headache. being new with little experence is always tough.
Dan
You might take some pictures and post them and get more help here.
Most controllers are strong enough for the 1000w hub. You could run a 250w controller on it but the performance would be a disapointment.
If you need a controller look in the new for sale section and LYEN sells a good controller.
Also buying a controller you will need to figure out the wiring sequence. Some but not most are plug and play but then wires don't match or plugs don't match.
You might doudle check your connections for loose or broken wires before you jump to a new headache. being new with little experence is always tough.
Dan
You might take some pictures and post them and get more help here.
I'm pretty sure the problem is with the controller, the wheel spins freely when not plugged in. but when the controller is plugged in it feels like a phase short, and regardless of the batteries being plugged in, the wheels' very hard to turn and jerky. I've open the controller up and their's no sign of frayed wires or visible shorts on the circuit bored. I'be also identified that it problem is with the yellow phase wire, but i don't know what to do from there.
I can post pictures if anyone wants to take a look.
I could have a problem with the Hallies too, but I guess I'll deal with the controller first.
thanks for all the info guys
I can post pictures if anyone wants to take a look.
I could have a problem with the Hallies too, but I guess I'll deal with the controller first.
thanks for all the info guys
Jeremy Harris said:
That's what I thought when a friend got a 1kw motor for his bike. No problem, just run it on one of those smaller controllers... snap, crackle, pop. Three of them fried in short order. Motor was fine. Got bigger controller. All was well. It may have been something about the winding resistance or inductance, but it definitely ate cheap controllers like candy.
Jeremy Harris
100 MW
texaspyro said:
You're experience with controllers (and batteries) seems to be a bit unusual!
If you look inside any ebike controller you'll see that it has current detection and limiting. It has to have in order to work, as I mentioned above. My guess is that your friend may just have bought an unsuitable controller that had some other issues, maybe it wasn't 60/120 compatible (some hubs are 60), maybe the wiring wasn't compatible, or maybe there was some other problem. What I do know for sure is that there is no problem with running a known to be good 600W controller with a known to be good 1000W motor. There will always be problems with mis-described parts though, so knowing for sure that any controller/motor will deliver or run at its rated power comes down to knowing the real limits of the components.
Jeremy
John in CR
100 TW
The results that Texaspyro's friend had is likely due to the motor running in current limiting virtually all of the time. Big motors with small controllers can blow controllers quite readily depending upon the type of mismatch, even without trying to push the controller beyond it's limits through modifications.
I blew a Knuckles 72V20A controller 2 years ago on the very first ride with a 2kw scooter motor, and that was a very conservative ride less than a mile on flat ground. I also managed to kill a Kelly ebike controller with the 1200W motor I've been running for the past 2 years. Those were both controllers that ran fine with another lesser hubmotor.
The only controller I've had that wasn't consistently hot during operation with my speed wind hubbies is the 24 FET Lyen controller I run now, and I had to turn it down to 75A battery 190A phase limits before it cooled off. Hot operation includes the stock factory controllers that are 15fet Infineons limited to 20A operation, though none of those blew until shunt mods pushed them past 50A battery side.
I'd have to say that it really depends on the motor. For example, my motor was the death of a Methods 100V100A controller. Luke and I hammered on it hard for almost a week, though with me aboard is was flat ground running. Then the first time I took it up a hill at partial throttle it blew in less than 1/4 mile. Luke, who weighs 30lbs less had already been up and over the same 1.5 mile hill. That controller was always hot. On the other hand Arlo1 runs his 100A Methods controller at much higher current limits and power with an X5304 without issue.
John
I blew a Knuckles 72V20A controller 2 years ago on the very first ride with a 2kw scooter motor, and that was a very conservative ride less than a mile on flat ground. I also managed to kill a Kelly ebike controller with the 1200W motor I've been running for the past 2 years. Those were both controllers that ran fine with another lesser hubmotor.
The only controller I've had that wasn't consistently hot during operation with my speed wind hubbies is the 24 FET Lyen controller I run now, and I had to turn it down to 75A battery 190A phase limits before it cooled off. Hot operation includes the stock factory controllers that are 15fet Infineons limited to 20A operation, though none of those blew until shunt mods pushed them past 50A battery side.
I'd have to say that it really depends on the motor. For example, my motor was the death of a Methods 100V100A controller. Luke and I hammered on it hard for almost a week, though with me aboard is was flat ground running. Then the first time I took it up a hill at partial throttle it blew in less than 1/4 mile. Luke, who weighs 30lbs less had already been up and over the same 1.5 mile hill. That controller was always hot. On the other hand Arlo1 runs his 100A Methods controller at much higher current limits and power with an X5304 without issue.
John
Jeremy Harris
100 MW
John,
I agree that a duff (i.e. incapable of running at its rated power) controller will get hot and blow, but it will do that on pretty much any motor its fitted to. The key is modding the controller to the point where the FET junctions can't get rid of the heat fast enough, either by poor thermal design, poor construction techniques or by increasing the power to a value greater than the manufacturer intended.
The actual set values of current on controllers when they leave the factory are quite conservative usually. For example, the recommended values for the XieChang range are:
6 FET:
24V - 10A current limit
36V - 12 to 15A current limit
48V - 15 to 18A current limit
9 FET:
48V - 20 to 25A current limit
12 FET:
48V - 28 to 35A
Judging all controllers on the basis of modded ones, perhaps originally modded by individuals who may not have understood the real limiting factors (which is almost always FET junction temperature) isn't really fair. The size of the Chinese domestic market for electric bikes means that there are now literally millions of these controllers working OK every day in their standard state. My guess is that if they failed regularly the manufacturer wouldn't stay around, as there seems to be a healthy level of competition in the market, even inside China.
The bottom line is that the motor can't set or control the amount of power drawn from the battery. Only the controller can do that.
A quick look at some motor specifications will show low power motors with low resistance and high power ones with high resistance, as well as the other way around. The power rating of the motor is driven primarily by its ability to lose heat, not the resistance or inductance of its windings. Here on my bench I have a couple of examples that illustrate this fairly well. An old Ezee motor, rated at 400 watts, with a winding resistance of about 0.1 ohm and an even older Crystalyte 406 rated at 800 watts with a winding resistance of 0.36 ohms. Spot the oddity? Yes, that's right, the higher powered motor actually has more than three times the winding resistance of the lower powered one, not the other way around. Without the controller to limit the current, the little Ezee would try and draw more than three times the current of the big Crystalyte.
The 5000 series motors (all rated at around a couple of kW I think) vary from about 0.06 ohms for the 5302 up to around 0.33 ohms for the 5306, so not that different from the little Ezee and 406 comparison I made above.
Jeremy
I agree that a duff (i.e. incapable of running at its rated power) controller will get hot and blow, but it will do that on pretty much any motor its fitted to. The key is modding the controller to the point where the FET junctions can't get rid of the heat fast enough, either by poor thermal design, poor construction techniques or by increasing the power to a value greater than the manufacturer intended.
The actual set values of current on controllers when they leave the factory are quite conservative usually. For example, the recommended values for the XieChang range are:
6 FET:
24V - 10A current limit
36V - 12 to 15A current limit
48V - 15 to 18A current limit
9 FET:
48V - 20 to 25A current limit
12 FET:
48V - 28 to 35A
Judging all controllers on the basis of modded ones, perhaps originally modded by individuals who may not have understood the real limiting factors (which is almost always FET junction temperature) isn't really fair. The size of the Chinese domestic market for electric bikes means that there are now literally millions of these controllers working OK every day in their standard state. My guess is that if they failed regularly the manufacturer wouldn't stay around, as there seems to be a healthy level of competition in the market, even inside China.
The bottom line is that the motor can't set or control the amount of power drawn from the battery. Only the controller can do that.
A quick look at some motor specifications will show low power motors with low resistance and high power ones with high resistance, as well as the other way around. The power rating of the motor is driven primarily by its ability to lose heat, not the resistance or inductance of its windings. Here on my bench I have a couple of examples that illustrate this fairly well. An old Ezee motor, rated at 400 watts, with a winding resistance of about 0.1 ohm and an even older Crystalyte 406 rated at 800 watts with a winding resistance of 0.36 ohms. Spot the oddity? Yes, that's right, the higher powered motor actually has more than three times the winding resistance of the lower powered one, not the other way around. Without the controller to limit the current, the little Ezee would try and draw more than three times the current of the big Crystalyte.
The 5000 series motors (all rated at around a couple of kW I think) vary from about 0.06 ohms for the 5302 up to around 0.33 ohms for the 5306, so not that different from the little Ezee and 406 comparison I made above.
Jeremy
John in CR
100 TW
Maybe the issue is controllers from a couple of years ago didn't have a good handle on phase current limiting, and that caused that first failure I had with the original Keywin/Knuckles pre Infineon controller. I think Kelly still recommends not trying to run a big motor with a small controller. The one I got was a fairly early model of Kelly's ebike controllers, so maybe they missed a related issue that has since been corrected. My other controller failures were all quite predictable though, and caused by shunt mods I did that pushed them past their designed performance level combined with a hard to drive motor.
I do have a question though. If controllers have full control over themselves, does that mean I can go ahead and buy one of the 12kw Colossus motors, and drive it with a controller with proven durable settings? My 24fet has proven to handle 99V100A without issue, so I should be able to run the Colossus at 10kw in the meantime while I wait for you guys to solve the extreme power controller issues? If I can't reliably do that, what part am I missing?
John
I do have a question though. If controllers have full control over themselves, does that mean I can go ahead and buy one of the 12kw Colossus motors, and drive it with a controller with proven durable settings? My 24fet has proven to handle 99V100A without issue, so I should be able to run the Colossus at 10kw in the meantime while I wait for you guys to solve the extreme power controller issues? If I can't reliably do that, what part am I missing?
John
dogman dan
1 PW
Back to the original poster's dilemma, Might look at this thread in the for sale section. http://endless-sphere.com/forums/viewtopic.php?f=31&t=21558
I don't know what it's called in wattage, but I see 800-1200 watts on my wattmeter using this controller all the time.
I don't know what it's called in wattage, but I see 800-1200 watts on my wattmeter using this controller all the time.
Jeremy Harris
100 MW
The only problem with the extreme motors, like the big Colossus outrunner, is that it really needs a controller with a higher PWM frequency. It's an extreme case (unlike the OPs original question which was based on off-the-shelf ebike parts) and so needs an extreme solution.
The 12kW Colossus, unlike any hub motor (or other ebike motor, come to that) I know of, has a very low winding inductance and a very low winding resistance. This means that, at the 15kHz PWM frequency that most ebike controllers seem to run at, the phase current can exceed the FET ratings, because none of the ebike controllers I'm familiar with use fast phase current sensing, they all guess at the phase current from the rpm, throttle demand and battery current. The answer is either a faster, more robust, controller with phase current sensing, or to add inductors in series with the phase wires. The latter approach is the one commonly adopted in fixed applications and big EVs, but isn't really that viable for an ebike, because of the size and weight of the inductors needed.
The bottom line is that the vast majority of ebike controllers were designed to drive low rpm, high inductance, relatively high resistance hub motors, so they aren't well suited to driving high rpm, low inductance, very low resistance RC outrunners without some mods. Even when modded there will be some extreme motors they just can't cope with, a bit like the situation with brushed motor controllers when the Lynch motor first came out - slow PWM controllers like the Curtis struggled to drive the low inductance Lynch motor without an external inductor in series with the motor. The Curtis controllers had been fine for years driving a very wide range of series wound motors, because they all had a pretty high inductance and fairly high winding resistance.
The Kelly issue was, I think, to do with the low PWM frequency and low commutation speed limit that some of their early controllers had, and has, I think, now been sorted. There were also some quality issues with early Kelly controllers, I think.
It is fairly easy to build a controller that will run at a higher PWM frequency and have fast enough current limiting that it can cut back on a cycle-by-cycle basis. It's a bit OTT for an ebike controller, but may address the demands of the new crop of high power, low inductance, low resistance BLDC motors that seem to be coming out now.
Jeremy
The 12kW Colossus, unlike any hub motor (or other ebike motor, come to that) I know of, has a very low winding inductance and a very low winding resistance. This means that, at the 15kHz PWM frequency that most ebike controllers seem to run at, the phase current can exceed the FET ratings, because none of the ebike controllers I'm familiar with use fast phase current sensing, they all guess at the phase current from the rpm, throttle demand and battery current. The answer is either a faster, more robust, controller with phase current sensing, or to add inductors in series with the phase wires. The latter approach is the one commonly adopted in fixed applications and big EVs, but isn't really that viable for an ebike, because of the size and weight of the inductors needed.
The bottom line is that the vast majority of ebike controllers were designed to drive low rpm, high inductance, relatively high resistance hub motors, so they aren't well suited to driving high rpm, low inductance, very low resistance RC outrunners without some mods. Even when modded there will be some extreme motors they just can't cope with, a bit like the situation with brushed motor controllers when the Lynch motor first came out - slow PWM controllers like the Curtis struggled to drive the low inductance Lynch motor without an external inductor in series with the motor. The Curtis controllers had been fine for years driving a very wide range of series wound motors, because they all had a pretty high inductance and fairly high winding resistance.
The Kelly issue was, I think, to do with the low PWM frequency and low commutation speed limit that some of their early controllers had, and has, I think, now been sorted. There were also some quality issues with early Kelly controllers, I think.
It is fairly easy to build a controller that will run at a higher PWM frequency and have fast enough current limiting that it can cut back on a cycle-by-cycle basis. It's a bit OTT for an ebike controller, but may address the demands of the new crop of high power, low inductance, low resistance BLDC motors that seem to be coming out now.
Jeremy
grindz145
1 MW
So right Jeremy, The Collosus is a completely different animal. For hub motors a short PWM will be tolerable for just about any controller.
Similar threads
