Kelly Phase Amps to Battery Amps

[2WheelsMovesTheSoul]

As a KLS-7230S owner, user and abuser, Imma chime in here.

I run mine on QS205 50H V3 3t motor with Statorade and Hubsink fins. I also have a 300a shunt metering power consumption. My pack is 72v nominal and 200a capable.

I routinely pull the controllers limit of 120a which pushes my tubby ass and 150lb bike up to a swift AF 55 mph and have had no overheating issue with the KLS as of yet. It is externally mounted (using Qulbix shield)and gets plenty of air though. As far as it being a small controller, internally, nearly all available space is utilized. Say a KT controller, one of those candy bar shaped ones, there is a huge waste of space inside and there's very little mass for heat to get out. That's why it's so long. The KLS is far heavier in a smaller overall footprint.

However, I do share the same gripe. From 0-10, it is an absolute fuggin slouch. It feels like any basic bitch 500w controller, once you reach say 8-10mph, with the right settings it ramps up like a mofo and begins to perform like a motorcycle. Generally speaking when I ride my bike, I'm only at 0mph when I leave my house and occasionally at traffic lights. When already rolling... even down to 5 mph. The slow start BS virtually disappears.

If you want a smooth, silent and very efficient controller, the KLS is a relatively good choice. I only say good because by now Kelly really should have some sort of LCD, OLED or TFT display. While it may have a Bluetooth app, its by no means a real time HUD. That would be my 2nd biggest gripe.

 

[rg12]

Thats interesting, thanks for sharing that

So I guess I should look into Lyen controllers eh?
All of us in our insane riding group have the 80A greentime with soldered shunt and it beats every controller I have ever seen.
You just lean on the handlebar and whack it and it takes you and your giant balls to 80% speed in like 2 seconds.
Not to mention the awesome burnouts...

P.S, greentime + lipos...thats where its at.

Me doing a burnout with an 18" motorcycle tire:
https://www.youtube.com/watch?v=H9Ezda3LYqI

 

[kebekua]

I tested about a dozen different Kelly controllers, mostly squarewaves, with power ranging from 3kw to 15kw.

I measured peak amps on all of them, which is the figure you need to know to match the proper battery and BMS.

I tested an early KLS 72V500A that had a really nice acceleration ramp, it could give full power at 0rpm. It was so easy to pop wheelies despite my ride's 130KG.

The new kls are not so nice. Same thing for the squarewaves, the older models had a faster throttle response but the difference is not as noticeable as with the KLS.

My KHB 120V 400A peaks at about 220 battery Amps, my KEB 72V 550A peaks at about 300A but hits over 400A for a split second at low speed. The current control is not as fast on square wave Kellys as it is with a sabvoton. With the sabvoton, it peaks at exactly 200A (factory unlocked version) and stays there. I would guess this is probably better for battery life as the BMS can't protect against such short amp burst, but the old KEB's and KHB's are much more fun to ride especially at low speeds!

The sabvoton is locked at 100A until you reach about 25kmh. Its 100A for both 150A and 200A versions. That's the main reason why I'm still using kelly square waves for my builds.

 

[haulincolin]

The original question in this thread has basically been answered: It depends on the equipment and the situation.

But I also want to chime in on how frustrating it is that Kelly rates their controllers by phase amps not by battery amps. I would like to be able to look at the controller name/rating (for instance KLS72601) and say that means I can do 500 amps at 72 volts = 36 kW, or about 48 hp. But that's not true! The high phase current rating only happens at a lower phase voltage. The actual power throughput of the controller is something way lower, based on whatever percent you set your battery amps at. It's probably pretty close to my Sabvoton SV72200 which is rated for 450 phase amps and 200 battery amps. So my max throughput is 200 amps at 72 volts = 14.4 kW, or about 19 hp.

 

[kebekua]

Colin,

The Sabvoton and KEB have very different power output curves.

The KEB has more peak power, the Sabvoton more continuous power. The Sabvoton holds that 200A pretty steady all the way up to top speed while the Kelly holds peak power for a split second then power drops quickly through acceleration. When my current ride is finished I will race it against my friend's, we have the same battery and 3KW QS motors. I got the KEB72V550A and he has the Sabvoton 200A. For sure I beat him on the first 50 meters or so, but I'm almost sure he will pass me after a couple hundred meters.

A real world rating for the my KEB 72 550 would be about 25KW peak and 8500W continuous. The Sabvoton can hold 14KW for several seconds but doesnt have a peak above that.

 

[flippy]

personally i have no issues with the kelly's low takeoff the first 5ft.
running a 4kW 10" hub motor at 80~90V at 300 phase amps (and 150A on the battery) means it takes off without spinning out (but it's really borderline) and the real power comes on when you have the traction and some foward momentum. its saves tires and increases control during takeoff. it also prevents the front tire from really getting some air.

i dont know what causes the kelly's behaviour but it is something i dont mind and helps less experienced people to drive it without crashing instantly.

in my setup it still chews up tires like free kale at a vegan convention....

 

[izeman]

i dont know what causes the kelly's behaviour but it is something i dont mind and helps less experienced people to drive it without crashing instantly.

THIS is something i'd like to know as well. it's obvious that that's the way it is with most sine wave controllers, but noone was able to tell me why. maybe it's because of the processing taking place inside the controller to calculate the sine wave form and apply it at the right time? who knows?!

 

[flippy]

my guess it looks like it is artifically limiting current at very low rpm's. turning on phases from a stop is basically a dead short. so keeping the current and switching time under control until the motor voltage is above something more managable before dumping the full current into the motor is my guess.
on the face of it it does behave like this is the case. not something i find a issue. it prevents the mosfets from burning up and all those bad things.

also: it is REALLY hard to make a sine wave on a slow turning motor.

 

[haulincolin]

Kebekua- That's very interesting information. Thanks sharing your experience about how the numbers equate to real world performance!

 

[Sunder]

Really glad I stumbled across this thread. I was about to place an order for a Kelly controller to be paired with a QS273. I thought I'd only want the smallest one, the QS site recommends a battery current of under 200A, and the smallest Kelly for 144v is 200A. Seemed like the perfect pairing...

But then Kelly says:

Usually the performance will be optimized if the power of the controller is 2.0-3.0 times of the motor's. Please take this point as reference when place an order online."

http://kellycontroller.com/faqs.php

While QS Motor says:

· Peak Battery Current: 210A (96V)
· Suggest Peak Phase Current: 530-660A

http://www.cnqsmotor.com/en/article_read/QS%20High%20Power%2017inch%202000W%20-%208000W%2012kw%2014kw%20273%20DC%20Brushless%20Electric%20Motorcycle%20Hub%20Motor%20for%20motorcycle/784.html

So they're pretty much bang on in agreement. Phase current seems to be advertised around 3x. So that probably answers @rg12's original question, (with a fair bit of marketing flexibility to adjust those figures). But leaves me with a small problem:

Does this mean the largest 144v Kelly controller, handling only 400A phase amps, won't make the most out of this motor?

 

[flippy]

if you increase the voltage above the rated 96v you need to drop the amps to stay within the power limits of the motor.
i would use the controller in your voltage range you want. a 96V controller can do 120V with a charged battery. the 144V controllers can do 180V(!) fully charged. no way that motor will survive that load at 200 battery amps and full phase amps. that is 35kW+ for reference.
the metric you gave is for the 8kW hub. so not the lowest.

you need to make the proper calculations of what winding motor you want to match the voltage you are going to use and the desired top speed. only if you know that you can shop for the proper sized controller. personally i recommend maxing out the voltage and match the other metrics to that. with a higher voltage you can have lower winding on the motor. having a higher voltage means you can lower the current and you dont have to use such huge cables to carry said current.

 

[Sunder]

Thanks Flippy, but I don't think you understand the physics behind it.

It was a bit of a rhetorical question, since it's a "how much is enough" type question. The answer, to paraphrase Creedence Clearwater Revival "they only answer 'More! More! More!'".

The 8kw rated power at 96v has little bearing on the instantaneous phase currents of the motors. All 8kw means is that at 96v, you can put 83A through it, and it will reach a non-destructive equilibrium temperature in a 25*C room. You could put a little more than that in for hours, or way more than that in for seconds.

The peak phase currents are probably more to do with the fact that for a moment in time, every electromagnet in a BLDC is effectively a short circuit. In that case, increasing voltage probably does mean de-rating the peak current, but whether it's linear or not, I'm not sure. Because as discussed on page one of this document, when the speed is slow, back EMF causes the apparent voltage to drop.

The thing is, I am building a commuter bike, not a race bike. High loads for long periods aren't going to happen. But pulling away from traffic (and making other motorcyclists go "WTF?") is my goal. If I can get this thing to go 0-100 in 3 seconds or less, which I think I can do, then it's not going to see 35kw for more than 3 seconds at absolute most. More likely 2 seconds, since most speed zones around here are 60-80km/h.

The question is also rhetorical, because I want a sine wave controller, and there's no bigger version of it available, so whether the motor can cope with peak phase currents of 660A as listed on the specification sheet, it is de-rated substantially to below the 400A the controller can supply, I'm stuck with that limit. (Of course, if I have to derate by more than 40%, then yes, I could kill this motor fast).

 

[flippy]

i understand the physics just fine.

but the trick is that if you increase the battery voltage you can change the kv of the motor wich means you can get the same effect with less amps. its a trade off between current and voltage. if you hit the current limit of the controller you can change the kv and increase the battery voltage to compensate for this. all those 3 factors need to be in balance. so you start with the voltage limit (i would stick to the highest the controller supports) and calculate the needed kv to get he top speed you want and then you get the proper specs for the motor you need. then you can play with the phase and battery currents to get the acceleration you want.

you can use a KLS96601-8080H for example. it has 600A max and 200+ on the battery. but you need to get the kv for a 28S battery so your kv needs to be calculated from a 84V baseline otherwise you will lose top speed when the battery drains. with that controller you will slighty exeed the rating of the motor if you max it out.

other choice would be the KLS14401-8080I. you could go up to 180V or a 42S battery. it has "only" 400A on the phases but you can compensate by using a slower motor that can take advantage of the higher voltage faster so the current/voltage ratio is basically the same between both controllers.

in practice the high voltage controller will run away from the lower voltage one as soon as you pick up speed as it can keep pumping more power into the motor due to the higher voltage that is available.

in your case i would go for the higher voltage and less current solution as the thinner cables are cheaper, simpler to work with and allows for longer sustained high loads.battery construction might be a tad more challenging but can be done if you split the battery in half and use 2 bms in series to get up to the 42S.

however you choose i expect this combo to destroy all exept a few hardcore racing bikes that expect a lame ass electric scooter to hit 0-100 in sub-5 second times.

 

[Sunder]

No, you're definitely off kilter there. I think I'll just let this one go. There have been a few threads, both started by me and by a few other people where someone asks one question and you answer another: nothing factually false, just not relevant.

 

[flippy]

your question is about how much phase amps you need to max out this motor. but phase amps capacity is dependant on the actual kv/winding of the motor you order wich also has a relation to the top speed you can get combined with the battery voltage.

all these are factors in getting to the proper rating of the controller you need. and at these high power levels you dont need 3 times the controller rating.

doubling or tripling the current rating is fine on smaller sub 3kW motors, but beyond that its simply not needed and very cost prohibitive.

so yes, i do understand it but you need to see if you are asking the right questions. once you have the calculated requirements for the battery, controller or motor you can ask more detailed questions but those questions would still not be all encompassing for everyone. there is nog golden rule for these things, it has to be looked at by a case by case basis.

if you get a small KLS for a ebike then yes, triple up the rating as they get hot. but once you go past the 15~20kW peak ratings you need to look much more carefully at the stuff you want to achieve and what the cost will be.

 

[Sunder]

This is exactly what I'm talking about. Nothing incorrect about what you've said. It's simply not relevant.

My understanding is that peak phase amps is most affected by winding losses. Winding losses are calculated by (I^2) * R. Notice that the formula doesn't contain "V"? That should mean in theory, that voltage does not impact winding losses, and hence peak phase amps. Of course, there's theory, and then there's the other half dozen laws of physics that really like to screw up your models.

Nothing to do with Kv (except very indirectly that high KV motors use fewer winds of lower resistance wiring) , nothing to do with 8kw maximum continuous power rating. Nothing to do with any of the stuff you were talking about.

So, if you know how to de-rate maximum phase amps (or even if de-rating was necessary) for higher voltages. I'd appreciate an answer. But I am not interested in the knowledge you do have, if it's not relevant.

 

[flippy]

winding resistance is a factor yes. but its only a part of the whole equasion. you need low resistance in order to get the current flow but in order to get the current flow you also need volts to push said current to overcome the resistance. so the lower the resistance and the higher the voltage the more amps can flow. in general it does not matter if you have the same motor with different windings, the rated power remains the same, there is just a change in phase amps it can handle.

once you start using the motor you need this:

you cant have current flow if you dont have the volts to overcome the restance.

as you see the winding has a massive inpact on the phase current it can absorb. you can put 400A into a T8 winding but it simply wont actually do 400A until you reach a certain voltage that allows such a mount of current flow but you will burn the motor out long before that.

so you can have a 10000A controller but if the winding resistance is too high and voltage in the motor is too low then the current simply cant flow to what you have set and wont get the peformance you desire. this is why the chosen winding is important because it directly impacts the resistance of the motor and it's current capacity.
and yes, you can have a very low turn motor with yuge amps but you would also need a LOT of voltage to get any kind of speed from it once the phase current starts to equalize with the battery amps. at that point you will accelerate slower and slower until you reach saturation or the power limit of the motor.

in short: you can have 2 motors with different windings that preform the same in power output but one will do that with less amps and more volts. and if you hit a current limit with your controller choice you need to change the winding and voltage in order to get the same performance as you did with the lower turn motor.
you can have 600A of "performance" from a 400A controller if you choose the right motor winding and battery voltage.

 

[Sunder]

Alright, just read two good pages on this issue:

https://www.electric-skateboard.builders/t/how-a-bldc-motor-works-and-why-a-hub-motor-gets-hotter/6753/32

and

https://www.linengineering.com/resources/white-papers/amps-phase-vs-amps-peak/

This stuff is finally starting to make sense, especially how Back-EMF contributes to motor heating.

In an ideal world, at unloaded speed, Back EMF = Input power, so zero current is required.

In an ideal world, a locked motor would have zero back EMF, therefore, the input power is infinite. In that case, the limit is the total power wasted to heat in a short circuit. Which is I^2R. Every watt you don't get rid of every second, you get one joule of heat, and eventually the temps rise until the enamel wire burns and shorts, hence burning out the motor.

Where this matters is in between the locked motor and the unloaded motor. As you move the motor from the first state to the second, back EMF rises, limiting the current that the motor will "accept". This current limit is usually higher than the motor can cope with, in terms of winding losses, so one of the roles of the controller is to limit that current.

So the short answer to my own question is: "No need to de-rate for increased voltage. The Kelly controller will limit the acceleration potential of the QS 273, but will also keep it safe".

 

[flippy]

or just change the winding so you can max out the controller and suck the last bit of power you can from the motor (without blowing it up).

for example: my 205 motor has a 4T winding so it can take a bit over 300A. and with a KLS7230S controller i can max it out, keep the battery current around 150A at 85V and still not blow up anything. the final speed is also pretty much on par with the legal top speed so i wont get any problems with the law while maxing out the power i can pull over the entire voltage range.

 

[haulincolin]

My QS 205 V3 50H 5T runs awesome with an unlocked Sabvoton 72150 (200 battery A, 450 phase A) even though QS calls it a 3 kW motor with maximum phase amps of 244. I can watch my power rise to over 14 kW briefly on my Cycle Analyst. Even after going up a big hill maintaining 7-8 kW, the motor is only slightly warm, and I'm not using ferrofluid or any additional cooling techniques.

What this tells me is that, in short bursts, I can easily push 4-5 times the kW through these QS motors as the advertised nominal amount. So if I were buying a 273 V3 50H that's rated at 8 kW, to make the most of the capability to do short high power bursts, I would want a controller that can push 40 kW. Does that seem reasonable?

 

[Sunder]

I wouldn't guarantee that.

It may be that even though the QS 205s might be able well more than their rating, it could be a "marketing trick" to try to encourage people to buy the bigger and more profitable QS 273s. As such, the QS 273s might not be as under-rated as the 205s.

If you really wanted that power, I'd try very incrementally increasing power, and watching heat. Not just immediately after a full power run, but for minute or so after the hard acceleration. It might take some time for the heat to conduct around to the temp sensor.

All that said, I am intending to push it as hard as I can go, but beyond the Kelly controllers, things start getting pretty expensive, so I have to accept that as a limit.

 

[flippy]

main difference between the 205 and 273 is the dropout width and is designed for (mountain)bikes and stuff with a smaller dropout. it makes up the lack of width for the wider magnets with a taller motor and more surface area. so they cant be directly compaired, they each have their place.

in your setup i would simply maximize voltage so you can keep the windings low and just crank up the amps as far as she will suffer.

 

[Sunder]

The 273 has a 273mm diameter stator, where as the 205 has a 205mm stator, strangely enough.

The 273 caps out at 80H where as the 205 caps out at 50H, so can be wider and taller. There's no "making up" with more surface area however you think that works...

On a width for width comparison, the qs273 packs a lot more copper.

Nobody was directly comparing them. That again is an irrelevancy you made up. Haulincolin was just asking whether the same under-rating which was applied to the 205 was likely to be applied to the 273, and without empirical testing none of us know for sure.

 

[flippy]

dude, now you are just being rude. please stop making it personal. this is a forum, not facebook.

i will try and make it more clear what i said:

the 273 HAS MORE STATOR SURFACE AREA. that makes it able to absord more heat then the 205 motors.

the same derating applies as the same amount of heat gets generated in the windings but the 273 will recover a lot faster from high loads and it can take higher loads longer as the stator takes longer to saturate.
in that regard the 273 (4kw) is clearly superior to the 205 (4kw) and should be de default choice if you have the space for that size motor.
and you can still offset the higher price with a lower tier motor if you really pressed for budget and do some statorade after installing.
but considering the 205 is not in a 8kw version you cant compare them on that power level. on the 4kW range i believe the lower winding motors are more reliable if you are constantly trowing high amps into the motor. the thinner strands of high turn motors are more delicate.