Nasty noise from new MSX 84v

[Aneta]

30 minutes ago, Planemo said:

If this is the case, there must be a reason why GW don't use sine? Ebikes have used it for years, and I presume other EUC manufacturers do as well. Is it harder to implement or something? Or is it just laziness?!

Sine BLDC controllers for e-bikes are premium and cost more. However, I remember reading somewhere that trapezoid has more torque at low RPM than pure sine.

[Aneta]

3 minutes ago, Aneta said:

Sine BLDC controllers for e-bikes are premium and cost more. However, I remember reading somewhere that trapezoid has more torque at low RPM than pure sine.

Oops, I probably got it reversed:

https://endless-sphere.com/forums/viewtopic.php?p=981766#p981766

 

[Aneta]

I found sound in @Marty Backe's loose axle nut video quite similar to @Planemo's:

(at 2:08)

The difference in volume is probably explained by Marty's better microphone placed almost under the tire vs. phone mic at greater distance.

Edited December 6, 2019 by Aneta

[Aneta]

Another possibility is that "thanks to design" the heavy battery brick is located right above and probably even touching the control board, possibly transmitting the vibrations to the gyro?

I think this "normal GW noise" should be questioned more, because if you search youtube for hub motor noise , you'll only find either issues with Hall sensors (either broken wire, sensor or even incorrect phase/Hall wire combination, there are 36 combinations total - could GW SNAFU that, too?), or one of fets burnt on the controller, or loose magnets. (Also, lots of results will be geared hub motors, but that's understandable, and we have direct drive motors.) But normally functioning hub motors/controllers do not sound like that, even dirt cheap motors and controllers for 10 bucks from aliexpress!

[Arbolest]

17 minutes ago, Aneta said:

Another possibility is that "thanks to design" the heavy battery brick is located right above and probably even touching the control board, possibly transmitting the vibrations to the gyro?

Luckily, I don't think this is it. The control board actually has a fair amount of clearance between it and the plastic ledge above it that supports the battery.

17 minutes ago, Aneta said:

I think this "normal GW noise" should be questioned more, because if you search youtube for hub motor noise , you'll only find either issues with Hall sensors (either broken wire, sensor or even incorrect phase/Hall wire combination, there are 36 combinations total - could GW SNAFU that, too?), or one of fets burnt on the controller, or loose magnets. (Also, lots of results will be geared hub motors, but that's understandable, and we have direct drive motors.) But normally functioning hub motors/controllers do not sound like that, even dirt cheap motors and controllers for 10 bucks from aliexpress!

I agree! I like questions, and I think it would be nice if Gotway questioned where the noise is originating from and whether it can be addressed. That said, I would just like to point out that the noise is not constant! It disappears at a threshold of around 2-3 mph and everything smooths out. If it was a failed electronic component that was causing it, then I don't believe the problem would just disappear with speed the way it does... 

Again, I feel that the issue originates from something strange in the way Gotway's firmware is designed, not something wrong with the control board or motor electronics. Just remember, Gotway doesn't purchase a separate motor controller, it's something they actually designed and implemented as part of their own control board. Because of that, they probably used some weird home-brew control algorithm for the motor instead of something plug-and-play that requires a separate, isolated set of control hardware to run on.  

Based on nothing more than my own observations (I haven't collected any data or done any rigorous testing to support this assumption) I think that what is going on is very simple: as the rotor turns and the motor needs to change which phases are being energized in order to continue its rotation in the desired direction, Gotway's control algorithm simply switches polarity and/or energized state without any type of ramp. I feel like if they would just implement a short attack ramp when switching phases (possibly reduced relative to rotational speed so available torque isn't lost)  this cogging/growling effect would all go away and we would have our glorious silence at all speeds, even the low ones.

[Tinkererboi]

1 hour ago, Aneta said:

The difference in volume is probably explained by Marty's better microphone placed almost under the tire vs. phone mic at greater distance.

Agree with @Aneta. The satisfying growl of a powerful wheel feels and sounds very different from the shuddering sound heard here. 

If i were you I would try the following first before trying anything more invasive

Grab the tyre in one hand and pedal in the other, wiggle and try looking for any play at all. There should be zero. 

Check for loose shell pillar screws.

[Marty Backe]

8 hours ago, Aneta said:

I found sound in @Marty Backe's loose axle nut video quite similar to @Planemo's:

(at 2:08)

The difference in volume is probably explained by Marty's better microphone placed almost under the tire vs. phone mic at greater distance.

In my case (and others I've seen) there is an associated, very noticeable, vibration emanating from the wheel. If his wheel feels smooth then it's just the motor noise. If there's a shaking and vibration, then there's a problem.

[Planemo]

I believe the cogging is being amplified by something. All wheels will cog, its the nature of having magnets and a charged coil passing each other. It can be reduced by increasing the number of magnets/coils or better firmware but either way, GW are trailing behind somewhere because other wheels around 2Kw are not nearly as bad.

That said, I think I may have found part of the 'amplification' tonight.

I put the msx onto the ground (the wheel was off) harder than I would usually and I heard the same 'tone'. My ears are accustomed to tone having been into and making music for much of my life. I nailed it down to the pedal play between the axles/shims. They clunk. Fowards/backwards/up/down you name it. The stock msx and Nikola pedals were the same in this regard - sloppy.

I need to do further testing but it appears (due to bone dry pedal axles and poor clearances) that the cogging is being amplified via the pedal system.

I dont want to keep dragging the Z10 into this but its all I have to measure against. Whilst the Z also came to me with bone dry axles, the tolerances are hugely better than the MSX. Both axle and shim fit are 'snug' so there isnt a lot of play at all. A bit like a slick gearchange on a sports car. No play, accurate and precise. They fold up, they fold down. No other movement anywhere. I am however discounting that the Z could be better on cogging due to lower voltage/motor power.

I will be filing the Nikola pedals in the next few days to get the angle I want. At the same time I will grease the axles and see if I can make some shims to reduce the axial movement (not much I can do about the radial) and take another vid on the same phone at the same distance/location. We'll see.

One thing I do know, having stuck my nose around the guts of the msx recently, the build quality between it and the Z10 is staggeringly different. One is akin to a Rolls Royce, the other a Suzuki.

If it wasnt for the msx's range, performance and ride quality I wouldnt take a piss on it, the bastard lovechild of an Armitage Shanks toilet cistern and a Samsonite briefcase.

[Aneta]

15 hours ago, Arbolest said:

Again, I feel that the issue originates from something strange in the way Gotway's firmware is designed, not something wrong with the control board or motor electronics. Just remember, Gotway doesn't purchase a separate motor controller, it's something they actually designed and implemented as part of their own control board. Because of that, they probably used some weird home-brew control algorithm for the motor instead of something plug-and-play that requires a separate, isolated set of control hardware to run on.  

Based on nothing more than my own observations (I haven't collected any data or done any rigorous testing to support this assumption) I think that what is going on is very simple: as the rotor turns and the motor needs to change which phases are being energized in order to continue its rotation in the desired direction, Gotway's control algorithm simply switches polarity and/or energized state without any type of ramp. I feel like if they would just implement a short attack ramp when switching phases (possibly reduced relative to rotational speed so available torque isn't lost)  this cogging/growling effect would all go away and we would have our glorious silence at all speeds, even the low ones.

A simple way to test this theory would be to install GW controller on a different wheel known to be quiet, e.g. KS 16X. No need to install inside - 16X has a huge flat "deck" on top, just secure GW controller to it using some tape, find a way to route the cables (will probably need to make cable adapters/extensions), and "trojan" it instead of 16X's controller. Will 16X then sound like "Gotway raw power"?

Anyone brave & curious?

[Marty Backe]

1 hour ago, Aneta said:

A simple way to test this theory would be to install GW controller on a different wheel known to be quiet, e.g. KS 16X. No need to install inside - 16X has a huge flat "deck" on top, just secure GW controller to it using some tape, find a way to route the cables (will probably need to make cable adapters/extensions), and "trojan" it instead of 16X's controller. Will 16X then sound like "Gotway raw power"?

Anyone brave & curious?

Not me. I'd rather be riding the wheel 

[Aneta]

1 hour ago, Aneta said:

No need to install inside - 16X has a huge flat "deck" on top, just secure GW controller to it using some tape

Oops, forgot that GW controllers are mounted sideways.

[Aneta]

18 minutes ago, Marty Backe said:

Not me. I'd rather be riding the wheel 

It's been raining in SC since Thanksgiving, with flash floods and all, and the weekend is gonna be miserable, too. C'mon, Marty, from our Hero become our SUPERHERO by establishing once and for all, whether the "Gotway raw power growl" can entirely be attributed to their controllers, or not. Just do it!^TM

[Aneta]

Besides earning the title of EUC Superhero Of All Time and the prestigious virtual Superman's cape:

and answering the Ultimate Question of EUC life, the EUC Universe, and the EUC Everything:

- this inexpensive (you just need to make a couple of adapters between XT90 and MT60 for battery power and JST (?) connectors for Hall wires) and not very time consuming experiment over the rainy weekend may revolutionize the EUC world as we know it by:

- establishing that it's possible to run the wheel of one manufacturer by using the motherboad of another mfr

- which also opens the possibility of troubleshooting dead wheels by using a known-to-be-good controller from another wheel (since many riders have more than one wheel) - why the wheel is not working? is it burnt controller? burnt motor? bad battery?

- open a possibility of insane mods where a second controller is installed in the empty space (often found in GW wheels on the opposite to controller side), which lets the rider to choose a different riding characteristic by switching around wires from one controller to another - e.g. have a 16X auxiliary controller in MSX for different feel/handling for offroad rides, for example. Controllers are relatively cheap (<$200), so having a nice torque-y sine wave controller in "Armitage Shanks toilet cistern" would be cool and useful!

If I had a second wheel, I'd do it, but I don't. I hope you or someone else will find this experiment interesting!

 

[Aneta]

33 minutes ago, Aneta said:

have a 16X auxiliary controller in MSX

I again forgot that 16X controller needs to be mounted horizontally. Nicola has a large horizontal compartment for extended battery, right? - can be used for secondary controller.

[Marty Backe]

3 hours ago, Aneta said:

Besides earning the title of EUC Superhero Of All Time and the prestigious virtual Superman's cape:

and answering the Ultimate Question of EUC life, the EUC Universe, and the EUC Everything:

- this inexpensive (you just need to make a couple of adapters between XT90 and MT60 for battery power and JST (?) connectors for Hall wires) and not very time consuming experiment over the rainy weekend may revolutionize the EUC world as we know it by:

- establishing that it's possible to run the wheel of one manufacturer by using the motherboad of another mfr

- which also opens the possibility of troubleshooting dead wheels by using a known-to-be-good controller from another wheel (since many riders have more than one wheel) - why the wheel is not working? is it burnt controller? burnt motor? bad battery?

- open a possibility of insane mods where a second controller is installed in the empty space (often found in GW wheels on the opposite to controller side), which lets the rider to choose a different riding characteristic by switching around wires from one controller to another - e.g. have a 16X auxiliary controller in MSX for different feel/handling for offroad rides, for example. Controllers are relatively cheap (<$200), so having a nice torque-y sine wave controller in "Armitage Shanks toilet cistern" would be cool and useful!

If I had a second wheel, I'd do it, but I don't. I hope you or someone else will find this experiment interesting!

 

This just isn't where I'm at in life right now  Much more of a user rather than a tinkerer.

Edited December 7, 2019 by Marty Backe

[mrelwood]

3 hours ago, Aneta said:

I again forgot that 16X controller needs to be mounted horizontally.

You also forgot that a controller requires cooling. Most if not all wheels have the Mosfet cooling plate exposed to the wheel well to offer cooling by air. The second controller should as well. A bit more work involved in shaping a hole in the shell and sealing it water tight.

[Unventor]

On 12/5/2019 at 7:00 PM, Planemo said:

Wow I am gobsmacked and shocked but relieved all the same! Really appreciate getting others thoughts, what a forum

As I say, you wouldnt have heard anything from the Z so I found it hard to believe that such a stark difference was normal!

We are now properly into winter here now Marty, -1 deg C yesterday so as much as I want to get on this new MSX, I prob wont do much riding for the next 3 months

(cue all the hardcore Scandinavian riders!)

You can ride all seasons, it just a very different ride style and speed in winter time. Yes your are like to experience a slide or a crash. 

I had my winter touch down 1 week ago. I didn't expect it to be icy when going home from work. But doing a sharp lean L-turn while braking, I hit an icy patch on pedestrian cross marks on the road, wheel started to slide and I over corrected it And it resulted me tipping over landing on my right elbow. It was as low speed but my reflex jacket got torn All-in-all good thing it happened. It put you in respect of road conditions and reset my ride style into winter mode. (I will get back to that).

I bought a V10F to ride in wet and winter days. To get a faster feel of its dynamics I found riding at high tire pressure help to feel the behavior of the wheel. But I had not lowered it to my normal ride preference at the time (I have do that now) . I also had the wheel in comfort mode, It made it act a bit like a spring effect to both cushion bumps but also to make it mode dynamic in ride sensation (that is now set to normal firm mode). Once we get proper snow fall I might let out a little more air of the tire.

So ride style....

You want to ride with all your weight over the axel, no leaning forward, backward or side turn leans.....if you start to lean the risk of wheel starting to slide is so much higher. This here takes practice. just like relearning to ride, it is hard at first. 'Main reason is you need to  ride very different and not over react when you are caught by surprise, which is hard as you are fighting your instinct and reflex reactions. But it can be done....as long you are not at too high a speed.

Next thing if you feel you hit a icy patch, do not try to turn or change speed, just let it roll and you deal with changes you need to do after you are free of the icy patch. If it is only icy conditions then you likely need to add studs to your tire. I have not done this myself yet. 

BTW, I am not hardcore, only a daily commuter trying to replace my car with EUC. I have 1800ish m to work, but I can't walk due elevation changes makes my knees to act up. On a normal summer day I get to work in 5-7 min depending on traffic. Yesterday because of icy conditions it took me 15 min. I rode last winter on a day when it was black mirror ice all over. took me 28min. but I never fell or slid. Stepping off the wheel almost took me down, as I then realized how slippery it actually was that day. 

As a side not, your summer riding skills improve too, as you can stay more cool headed in tricky situations. Or that is how it worked out for me. It could be a combinations of the added mileages too. Something happened after my 2500km total ride (V8+ KS18L) and the 200km+ on the KS16X. 

[mrelwood]

 

1 hour ago, Unventor said:

Next thing if you feel you hit a icy patch, do not try to turn or change speed, just let it roll and you deal with changes you need to do after you are free of the icy patch.

This is indeed crucial. Whenever it gets really slippery, I always try to mimic a big, heavy, free rolling ball: If the road banks right, turn right like the ball would. If I’m on a downhill, I let the wheel accelerate just like the ball would. Extremely slight turning or braking is of course sometimes required, but the ball trajectory is the center line that I only very gently may veer away from.

1 hour ago, Unventor said:

If it is only icy conditions then you likely need to add studs to your tire. I have not done this myself yet.

I should get my new studs today! I won’t install them right away though, since we have a warmer week coming up. The snow already melted for now.

[Planemo]

Ok I fitted the Nikola pedals, greasing all the moving parts in the process and its definately better but not enough to warrant doing another vid. I have however managed to get a quick 20 min/4 mile ride with lots of stops/starts/low speed stuff and I have to admit, once outside I never heard it! As people have said, it does seem just a normal quirk of the wheel. I still think it should be better though..!