Magic Pie 5 Vector Suddenly Getting Louder.

[LeftieBiker]

I've had my converted Magnum Metro on the road almost a year, and the motor has always been nearly silent. Lately though, it's making a sound a lot like a geared hubmotor - not sick sounding, just an obvious machine-like whirring sound. I had the bike worked on by a VeloFix bike mechanic recently, and he told me that when he put the rear wheel back on after tuning the spokes, that he had to 'rearrange the rear axle washers'. I'm not sure if that could be it, or not. I loosened the axle nuts a quarter turn, thinking he might have put too much pressure on the bearings, but I don't think it mattered much if any. I have the bike off the road until I can try re-rearranging those washers. The thing is, the bike was fine for a few weeks after it was worked on by the mechanic. Any insights would be welcome.

 

[amberwolf]

Does it do the same thing when riding it as with the bike upside down and wheel in the air? (or otherwise with wheel off ground and unloaded)

If not, I'd suspect the spokes themselves, or the rim/spoke interface, as those are the most likely mechanical things to change behavior between loaded and unloaded. The tire can also change behavior because the pressure at the bottom can slightly change the shape of the rest of the tire via internal pressure, and if it is *very* close to touching the frame or something else stationary when unloaded, it could just barely touch that when loaded.

Does it do the same thing with the motor system turned off and just pedalling, either while riding or offground?

If not, then the problem is somewhere within the motor system, most often cable damage, but if the connectors were unplugged *and* it's an external controller with phase/hall wires in the cable to the motor, a poor or incomplete reconnection of the cable to the controller could cause altered behavior.


Generally, for noises that started after removing and reinstalling a wheel, I'd check for anything mechanical that could cause friction or touching any part of the wheel (tire, spokes, rim, motor casing, etc), and for connections between wheel and rest of the bike.

Regarding the bearings, the ones in a motor don't work like those in a regular bike hub, which are conical, so they generally aren't affected by the axle nut tightness the same way (with some exceptions depending on axle and cover design and fitment).

If loosening the axle nuts actually made a difference, you should investigate why, especially if the motor cabling passes thru that area, *and* you have an external controller (many MPs have internal controllers and wouldn't be affected the same way by cable damage). An external controller's phase/hall wires could be damaged if pinched between frame/hub/spacers, even if there is no visible external damage, by crushing insulation or breaking conductors inside the cable, and that can cause altered motor behavior, which is generally worsened the higher the loading on the motor (so louder at start from a stop, or up a hill, hard acceleration, etc, and quietest with wheel off ground).

 

[LeftieBiker]

Thanks for the detailed reply. After first figuring out how to power the bike with the battery removed to allow for the bike stand, I ran it on the stand just a few minutes ago. It seems to have the same noise - maybe a little less loud now, maybe not. My next step is to drop the wheel just a few inches, and try to put the washers back as they were. Which raises another question: the mechanic added a medium thick washer between the right frame dropout and the chain/8th gear, which he worried was too close to it. I can either ignore that and keep running it with maybe 3mm of clearance, or I can try a washer half as thick as what he added. What would you do? Again, the setup ran fine for nearly a year like that.

 

[amberwolf]

If it worked fine without chain / frame interference, there is no reason to change it, so you could just change it back.

If there had been chain/frame interference, there will be damage to the frame (or paint) where it occured.

 

[LeftieBiker]

I did get the wheel dropped, removed the washer, and raised the wheel back - bungee cords help a lot with this, when alone. Alas, the noise was not just unaffected, I think it's slowly continuing to get louder. After going through this with two hubmotor bikes, it's discouraging to have it happen with a DD. The only good news is that the bike's odd increase in top speed, which appeared just before the noise, seems to be staying as well. Instead of 25-26MPH, it will cheerfully do 26-28MPH.

 

[amberwolf]

Usually an increase in speed, if battery voltage is not higher and the motor winding version isn't different means the phase/hall combination is wrong, or otherwise timing of signals is incorrect causing incorrect driving of the phases by the controller.

This can also cause noise problems, especially ones at specific speeds, and it can also cause problems starting from a stop under load. The excess current drawn by this configuration can eventually damage things.

If you have a single multipin motor/controller connector/cable, this is unlikely as it would require rewiring one end or the other to change the wire order.

But if you have separate connectors for phases and halls, using individual plugs for each phase (and/or each hall wire) then it's possible to put htem back in teh wrong order, whcih can cause this.

 

[LeftieBiker]

I have the single cable, so I don't see how it could be wrong hall timing. I did put a little liquid electrical tape over the end of the kevlar tube that protects the cable where it emerges from the motor, but unless that eats the plastic insulation it's supposed to protect, and then selectively changes the pulses that leak through...nah. Anyway, the bike continues to perform at literally 110%, the noise is still there and may be slowly worsening (or I'm becoming more sensitive to it). If I read the improvised odometer reading correctly, the bike is approaching 1000 miles on the converted drivetrain. The is just about where the original Das Kit motor started to get louder. Does anyone know of a 25+ MPH DD conversion kit that is actually bulletproof and silent...?i

 

[amberwolf]

Kit? No. All the kits I've seen go for cheapness of parts to maximize profit; some are better than others....but they still start from the same basic stuff.

Some motors are certainly better built than others, though (and same for controllers etc., but I suspect your problem is mechanical, and almost certainly caused by the maintenance since it didn't exist before that and started as soon as that was done).

For example: I'm still abusing an old Stromer bike's "500w" Ultramotor relaced into a 20" wheel on the right side of SB Cruiser, at a few times it's rated power for a few seconds at every startup, with an external controller (it runs at about it's rated power for cruising). Mechanically it's been bulletproof and silent for the years of use and thousands of miles it's seen for me (after who knows how long it was used on the bike it came from, under what conditions).

The bigger and much heavier MXUS 3k 450x on the left side has fared less well, and the one I used on the right side (replaced by the Ultramotor) even less well (broken axles on both, etc).

The QSMotor QS205 would probably be "bulletproof" in either of our applications, but it's even bigger and heavier than the 450x. I have one I need to put a new axle in (previous owner broke it using several times it's rated power in offroad use, with insufficient axle mounting, so it's still "bulletproof", just not "howitzer-proof". ), to test it out once it's laced into a wheel.

As for silent...as long as it's a sinewave controller and the motor is sensored and works well with it, it'll be as silent as you can get...with the possible exception of an FOC controller that is properly tuned for that specific motor (depends on the specific controller; some aren't silent because they deliberately generate "noise" to keep track of rotor position).


Various people riding under various conditions with various jobs their systems have to do have had varying results with pretty much every kit and motor out there, from what I've read here on ES and elsewhere...so there's no one kit I can think of that will work for everyone equally or equally reliably. Even stuff by Grin, who probably does more to vet their source products than any other company I can think of, has problems with some things.


And...anything can get broken or damaged during maintenance or other times it is removed or reinstalled, or any other changes happen, especially if it's by someone that doesn't know the system, or changes things about how it was originally setup or installed, if those changes/etc have the potential to damage something.

 

[LeftieBiker]

"Some motors are certainly better built than others, though (and same for controllers etc., but I suspect your problem is mechanical, and almost certainly caused by the maintenance since it didn't exist before that and started as soon as that was done)."

I remembered something he said, after reading that. One of the reasons I had used him again was that a spoke has loosened, and the nipple was lost in the rim. I didn't want to have to pull the wheel, and thought it unlikely that I could fish the nipple out otherwise. He said that he was trying to get the spokes really tight, so they'd stay put. So should I loosen each spoke 1/5 of a turn? And if I do get it quiet again, should I then threadlock the nipples?

 

[amberwolf]

I wouldn't loosen spokes...but if the spokes had to be tightened because they keep loosening, you should very carefully examine the rim, in direct sunlight, around the nipple holes, with a magnifying glass (or other method of significant magnification) for deformations or cracks.

It is not uncommon for such loosening to be caused by too-thick spokes (for the rim used) to require such high tension that they deform or crack the rim over time, and as that happens, the spokes are then detensioned. Further tensioning is required, which damages the rim further, in the same or different places, and continues the process, and eventually rim or spoke failures can occur.


None of these would change the *motor* sound, however, unless such high spoke tension was used on an extremely thin motor shell that the shell itself deformed and caused changes in the mechanical / magnetic operation of the motor itself (very unlikely). They might cause pinging, knocking, rattling, creaking, or other periodic sounds at the rim or from the spokes (once per revolution for each problem area), usually only when loaded (riding on the ground, rather than wheel in the air), but I've never heard a whirring or motor-like sound, or seen any changes in motor behavior (speed increase, etc).

 

[LeftieBiker]

Sorry for the late reply - I'm working on my house. Anyway, the motor is staying at about the same, tolerable volume, so aside from inspecting the rim when I get the chance, I'm going to just keep riding it. Thanks for all your help.

 

[LeftieBiker]

More issues with the bike. The one that most concerns me is that now it is often reluctant to start moving under throttle. It jerks weakly for a second, then after a stronger jerk or two, it rides away normally. Aside from a failing battery, with attendant voltage sag at lower states of charge, it behaves normally after that.

 

[amberwolf]

The most common problem that causes that, aside from motor or controller damage, is a controller setting that limits power / current more under greater voltage sag. I don't know if the GM controller has this, or if it is a user-accessible setting (they're not always even visible in settings, just something the controller software is sometimes designed to do to help a user get the most out of their battery).

If it happens more at lower SOC, that's probably the cause.

Sometimes electrolytic capacitors fail over time, especially under excessive heat conditions (such as for controllers embedded inside housings that contain other heat-generating devices, like some of the GMs with internal controllers). These often exhibit swelling, peeling heatshrink/labelling, but sometimes there is no visible sign and they simply have less capacitance and higher ESR so they do their job less well and the devices they are in become more susceptible to problems during high current / power usage.

If it happens exactly the same regardless, and also happens even with the wheel offground but with some load on it (like a gloved hand on the tire, or a brake lightly held), it could be a hall signal issue. Sensor failed, or it's signal wire disconnected between sensor and controller. Often this type of problem occurs slightly differently depending on the exact wheel rotational position.

It's unlikely to be a FET or phase problem, as that usually causes other operational problems, but a poor phase connection sometimes causes this.

 

[calab]

Might be interference with the controller, or controller getting older and degrading over time.
Did you over heat the motor or have phase wire strands touching each other at the axle exit?
Sensorless controllers can be a bit jerky from a stop.
Cheap controllers act up even more.
I notice characteristics of my setup change over time with different noises and vibrations and its the exact same cheap generic controller and Leaf dd hub motor. Might be a bearing is a bit old, dirty and off, and having one torque arm adds a new noise or vibration vs 2 ta's.

 

[LeftieBiker]

The problem doesn't happen most of the time, but it does happen pretty reliably the first time I use the throttle to get moving at the beginning of a ride. After that it's more like once in a while.

The most common problem that causes that, aside from motor or controller damage, is a controller setting that limits power / current more under greater voltage sag. I don't know if the GM controller has this, or if it is a user-accessible setting (they're not always even visible in settings, just something the controller software is sometimes designed to do to help a user get the most out of their battery).

If it happens more at lower SOC, that's probably the cause.

What would this parameter look like? I've been playing with "Acceleration" which is essentially how soft the start is. I can make the problem worse with that dialed back but not better. I had lowered battery current limit from 25 to 22 amps, to spare the battery; I meant to set that back today, but forgot, so I'll try that next.

 

[amberwolf]

The problem doesn't happen most of the time, but it does happen pretty reliably the first time I use the throttle to get moving at the beginning of a ride. After that it's more like once in a while.

A battery issue is less likely at the start of a ride when the battery voltage is higher and systme temperatures are lower, but if it is a poor connection (anywhere in a high current path especially) affected by temperature, that could make sense. Vibration can also affect such a connection.

If it's a capacitor, they usually get worse at higher temps, or really really cold ones.

What would this parameter look like?

I don't know what the setting would be called if ti has one, there are so many ways to translate or mistranslate technical terms, and so many things one might call it even in English.

But what the setting *does* is only limit current when the battery level drops below a certain voltage. So it is most likely a two part setting, the first part being a specific voltage that is above but near the LVC level, and the second part being a current that is near but below the regular current limit.

It's possible (probable, really) that one or boht of them is hard coded and thus not available to the end user.

If the software is well-written, these limits should automatically adjust themselves up or down based on the actual LVC and current limits, and if they aren't user-accessible they probably do.

I've been playing with "Acceleration" which is essentially how soft the start is. I can make the problem worse with that dialed back but not better. I had lowered battery current limit from 25 to 22 amps, to spare the battery; I meant to set that back today, but forgot, so I'll try that next.

Seems to behave the opposite of what I woudl expect. Such a problem would normally be worse the higher the current is, if it is caused by anything in any part of the current path where the current draw causes a voltage drop in or across something that then causes the controller to react.

 

[Chalo]

If your controller is "dual mode" (with or without Hall sensors) and there's a problem with your Hall sensors, cable, or plug, well then it might be defaulting to sensorless operation.

 

[LeftieBiker]

It seems the MP5 is dual mode. My bike will usually start by itself when the throttle is applied, though.

I didn't express myself clearly about the battery: I don't think it's involved with the juddering - it just happens to be failing while the other issue is going on...

 

[LeftieBiker]

If your controller is "dual mode" (with or without Hall sensors) and there's a problem with your Hall sensors, cable, or plug, well then it might be defaulting to sensorless operation.

We have a winner, folks! The Hall sensors seem to be going or have gone already, as I now recognize the symptoms of a dual mode controller in sensorless mode.

 

[LeftieBiker]

Funny: I got an email showing a reply here that seems to be...not here. Anyway, it suggested that the motor overheated the Hall sensors and toasted them. AFAIK, though, the motor has never gotten Hot. What's up with the bike now? I'm running dual batteries to try to compensate for the degraded and weak cells. I had a new battery manager on the bike, and it was working very well, but it died yesterday after my housemate unplugged the chargers from the batteries without shutting them off first. I'm surprised that this was enough to kill it, so I won't be replacing it - at least not with the same brand. For now, I'm plugging in one battery at a time, via the modular XT90 cabling I installed when I installed the MP5 kit. This at least lets me use Regen again.

 

[amberwolf]

Do the sensors actually test as failed?

A controller could go into sensorless mode for any case where expected signals aren't happening, which can include noise in the signals, connection issues, no signals, wrong signals, failure inside controller of pullup resistors or signal inputs or power supply to halls, or failure of controller to filter noise in general (bad capacitors, etc).