No announcement yet.

Kelly blames Cyclone, Clylone blames Kelly

  • Filter
  • Time
  • Show
Clear All
new posts

    Kelly blames Cyclone, Clylone blames Kelly

    My controller produces a hall sensor error just after the motor starts vibrating and then shuts down always after about 30 minutes of riding. Each day i ride my bike again it works fine again for about 30 minutes. If i ride it hard it will only last 15 minutes.

    the common problem is clearly heat which is affecting the hall sensors.

    i contacted Cyclone and they say my Kelly controller is tripping on high temperature. When i contact Kelly, Kelly say that the because i get hall sensor error it's a problem with the hall sensors and not the controller because if if was high controller temperature then i would get a high controller temperature error.

    how do work out what the problem is??

    Kelly has a plausible reason. After all, it's the controller doing the sensing, it "knows". I'd go with them.

    Could bring cold spray, ride until hot, spray controller, and see if it works, as a test.
    Fabrication is fun! Build something today. Show someone. Let them help. Inspire and share. Spread the desire.


      So is your Kelly controller a model KBS-X that is displaying an error code of 4,2 flashes? How hot is the motor getting? Too hot to touch? A motor hall sensor is rated up to 150 *C or 302 *F maximum temperature.
      That is pretty hot. Are your riding habits within the cyclones parameters? Do you have a way of seeing your amp or wattage readings of your motor under load?
      See my completed Magic Pie V5 rear hub motor E-Bike build HERE.


        yeah that's the one. KBS48121X with error code of 4,2 flashes. i have set my kelly to a maximum of 120 amps at 58 volts so i am putting a peak of 6.9 kw through it and my cyclone is a 3kw.

        that's interesting to know that the hall sensors will fail or temporarily stop working when hot or 150 degrees. my motor is getting hot but i can still easily touch it, it's not that hot.


          Wow, over double the motor's rated capacity possible. :-0 (well, 3234 watts continuous) That explains it. I surely wouldn't want to depend on the halls as an over temp safety. Some of the newer cyclones have a temperature limit switch in series with the hall wiring... you sure you don't have one buried somewhere? If not, would be a great time to install one... or even better a motor temperature thermistor sensor and display to keep an eye on the motor's internal temperature, especially when living on the edge! ;-)

          And maxing the controller's voltage...may be time to dial it back a bit...
          See my completed Magic Pie V5 rear hub motor E-Bike build HERE.


            I think if considering the "advertised" design parameters of your 3KW Cyclone, versus those of your controller--it's clear that the Cyclone is the source of your woes... But Cyclone--shouldn't blame Kelly, and Kelly is also wrong to blame Cyclone.

            Instead, both of them should be focusing blame on the user... That's you!

            Now--don't take offense--I'm all for pushing the boundaries of our gear. That's a lot of what we do here, so you're in very good company! :D

            But yeah... The 3KW Cyclone is designed for a rated output of 3000W; with a max of 4000W. But the thing is, that wattage is based on a draw of 40A at the maximum-rated 72 volts.

            Voltage and Amperage are inversely porportional--to achieve any given wattage, if one is decreased, the other must increase.

            Knowing this, what it looks like you've attempted to do, is to force more amperage through the motor--believing that this should be just fine, because you're still within the "wattage-rating" of the motor...

            Am I on track? Would that be a fair guess? Problem is, you just can't do a flip-flop. Why? It's all in the wiring, I'm afraid. To increase the amperage--to gain more wattage without increasing the voltage--you'd need to also increase the size of your conductors. But the motor was manufactured one particular way, and you can't (realistically) change its windings. These physical limits are why motors also come with an Amperage Rating. Your motor's Amperage Rating is for 40A.

            Your 3KW Cyclone, isn't operating in a 72V system, it's operating in a 48V system (fully-charged would be 58V). So, a normal-continuous load of up to 40A would be no problem for your Cyclone. A maximum peak of 50A here or there would be fine as well. But that's not going to yield you the same 4,000 Watts peak. No, due to the lower system voltage you've selected, your 3KW Cyclone--will only produce (according to its design parameters), 2,240 Watts continuous, and 2,900 Watts peak.

            Those numbers may be smaller than you'd like to see, but that's actually a huge amount of power. Even with the "factory" Cyclone Controller, if frequently operating at those kinds of high power-outputs, you ought to be regularly snapping drive-components--no matter what your build--as you'd still be peaking at about 4 horsepower (746 Watts per Horsepower).

            If it turns out you don't break a lot of stuff, I'd have to also consider: Maybe your Kelly controller might not be programmed quite perfectly (in terms of settings) to match your 3KW Cyclone (not that I'd have a clue what the correct settings might be).

            I had wondered to myself, "What the heck is all that energy going to?"... You've shared none of your project details, so for all I knew it might have all made perfect sense... Heck I could be totally wrong about all of this--and it still just might!

            But then I started thinking... With my own 4KW Cyclone Mid-Drive (also 48V system--58V fully-charged), and its 60A factory YYK brand "Cyclone Bluetooth Controller" I've sheared off sprockets and broken heavy-duty chains at less than half the amperage you're pulling with your 3KW motor! So--yeah. Curiosity had the better of me...

            Now, I'm pretty sure I know where all that energy is going. All that extra amperage, is just turning into heat. That's why you're not mentioning breaking parts constantly--you're not getting extra power from your motor--you are just cooking it with your snazzy powerful Kelly controller.

            But listen--I'm not working for Cyclone. I'm not criticizing. I'm just sort-of amazed! That little Cyclone of yours can sure withstand a heck of a lot more than I ever imagined! Wow. That's truly astonishing!

            I do recommend reviewing your Kelly controller's settings, since it's apparent that you're currently not unleashing your Cyclone's full power. If, or when you've finally cooked your 3KW Cyclone for good, I'd suggest you try the 4KW version. It's not much more investment at all, and it has a higher efficiency-rating. Plus, it has also got a lot more mass.

            Anyways, best of luck!

            Last edited by tklop; 06-05-2019, 12:58 AM.


              yeah i know i'm putting far to much power through the motor. i'm putting through it 6kw peak in a 3kw motor so it's definitely going to overheat. the question i originally had is regarding the hall sensor error i was getting. i can only assume that the hall sensors stop working when they get to hot, i don't know? that's what it seems like it's doing.

              however the motor isn't getting that hot. it only gets warm to touch before it i gets the error. i would expect it to be much hotter. maybe they are cheap crap hall sensors i don't know.

              regarding where is all the power going. running on high power certainly gives the bike phenomenal acceleration. i've heard that people have sheared / bent sprockets with the cyclone motor. I don't buy that. if i put down too much power the chain will slip over the cogs. i never put down full power on a high gear only on the low gears (large cogs). i can climb any hill i can get traction on. i have bent and bucked my wheel by doing this and also have snapped quite a few chains also.but not sheered sprockets. i only weigh 70 kgs though so i could imagine fatties who weight 120 kg's stressing the drive chain a lot more.

              i don't think i am cooking the motor though. it just cuts out too quickly before it actually gets really hot. maybe the extra power isn't worth it. i.e it could be getting to magnetic saturation and the extra current doesn't make a difference except for heat everything. i'll dial back the power and see if i notice much of a reduced acceleration.


              • tklop
                tklop commented
                Editing a comment
                Oh, I am sure you get nice acceleration. So do I--running the stock-controller, and heavily de-tuned.

                Respectfully, there's no way you're getting anywhere near the rated output of your motor. Not even close. You're also not getting anywhere nearly as much power out of it as you would get with the stock 40A Cyclone controller. As sold, the 3KW Cyclone, in a 48V system, is plenty strong enough to bust all kinds of things. If you were actually getting output in porportion to your ridiculous amperage, you would be breaking stuff even faster, blowing up IGH's, shearing sprockets off, etc. Daily probably.

                The Cyclone with its stock controller will provide a lot more power than what you describe, with no "extra" amperage needed.

                By unleashing the motor's potential power, you'll find that you'll need to "turn it down"--operating it with "soft" settings. These "slow start" type settings will give you very much the performance you currently describe--but you'll be seeing that performace at between 10A and 30A (if you were to monitor things). You don't believe me, but it's true. Your current performance (as you describe it) is approximate to the Cyclone's "soft" settings.

                But that's great news! You'll be able to enjoy your current performance level, at a fraction of the energy-consumption--with a properly configured controller (or the stock controller). And then--of course--if you're willing to tempt fate--you could still "turn it up". Quite a lot, in fact. And if you did, you'd far exceed any performance you've yet experienced by using the wrong controller and controller settings. And you'd start breaking things.

                You sound like you're still trying to convince yourself that you're doing something clever--but you're really just wasting electricity. You could get WAY more power--with WAY more range--if you used the correct settings, and/or the correct controller.

                If you had a stock controller, you'd already know exactly what I mean.

                I do hope that you can figure it out.

                As great as you think your Cyclone is right now--you'll love it ten times more--I can assure you!

                Best of luck--and be safe!

                Last edited by tklop; 08-14-2019, 02:31 PM.

              I would like to believe you as that sounds great. however i have already tried and tested the stock controller. the kelly controller produced significant more acceleration than the stock controller. so yes i am convinced by trying out both. of course you need to compare like for like gears.

              i understood that BLDCs controllers and motors are fairly standardised and only sine wave controllers need any worth while programming. i would be interested to know more. send through links for more info.

              regarding sheering of sprockets. i am using a high tensile steel cassette and an ebike chain. it's not possible to bend the cogs sideways as the chain force is in the same direction as the direction of the direction of the rotation of the sprockets. Do you mean shearing of the teeth? The cogs are incredibly strong in the direction of rotation. By far is the weakest point are the individual chain links. i only put down full power on large rear cogs. with a 30 tooth chain ring the loads is spread across 15 or so teeth. the teeth will not fail when spread over 15 when compared to individual chain links. that's why i only snap chains. what damage have you actually done to your cassette? i would be interested to know what. in my experience too much power always
              = chains slipping. it would be good to know how you keep your chain in contact with the cassette long enough to do any damage....

              ANYWAY. i only wanted to know why my hall sensors where cutting out. i already know that 6kw is obviously too much power. i just wanted to know that that's normal for hall sensors to 'shutdown' under too much heat and then work again once cooled.


                i was double checking kelly controllers compatibility with Cyclone motors. as far as i can tell they are meant to be plug and play. Luna were selling them as an upgraded option for the cyclone motors. the main parameter in the set up you need to get right is the hall sensor angle 60 or 120 degrees. They have removed the 'auto learn' function from the latest operating software for some reason.

                it doesn't look like there needs to be careful programming implemented. plus i am getting more acceleration than the stock controller. but like you say i definitely don't want to be expelling loads of extra energy for no increased performance. you're quite right, that would be a total waste and needless.



                  One thing which you stated a couple times is that heat is related to power level. Really, heat is much more closely related to current. If you run double the power at same current (comparing 2 hypothetical usage scenarios with 2x voltage on the 'fast' one) you won't make much more heat. This is oversimplified, but basically true.

                  Higher voltage allows more power at same current, for better power efficiency. So you can up the power in efficient, or inefficient, ways. Raising current more than voltage, as you more or less did, is the way to make more heat with more power.

                  I think that's a core of TKlops comments you may not have understood, so am pointing it out specifically as a helpful guiding concept for your quest to have higher power with acceptable heat losses.
                  Fabrication is fun! Build something today. Show someone. Let them help. Inspire and share. Spread the desire.


                    Not to pile, but I am seeing this exact thing with my Phaserunner / BBSHD combo. Once I surpassed 50A current with 52v(58.8v) pack I started getting negative returns on both torque and speed. Not to mention the heat, I have additional heatsinking. To counter for this I am rebuilding my packs to 72v(84v) and decreasing the max current. It will run cooler, and have a higher top end.