Nucular 12F + SL-120-50 SPM causing initial dying cat noises

[Rewind3000]

Hi,
Have a look at this video:

I am trying to tune my Nucular 12F to work nicely with a SL120-50 SPM motor but at first when I twist the throttle I get noises like a dying cat.
I am using a Lunacycle Jumbo Shark 52V 17.5Ah, and I think its max discharge is 30A. The hardware, except the motor, is taken directly from a LMX 64 bike.

I did the auto-setup, and setup of the throttle.

I borrowed some PID regulator values from LMX who did PID tuning on the SL120-30 SPM on the LMX 64. It seemed to work better than original values, with less scraping at high speed. But still, I have scraping sounds at initial throttle movements.

I am attaching the config files for anyone to scrutinize. I am very new to this and open to suggestions.

Update: While Senli heard no such noise, Nucular says this is just PWN fluctuations and the sound will be very different when the motor has a load.

 

[DanGT86]

welcome to the club. I think most of us with that LMX SPM motor have this same problem. I was annoyed by that noise as well but in normal riding it never really spends any time there. I got PID tuning suggestions from the Nuc team but never got the bugs completely worked out. Mine runs fantastic in BLDC sensored mode but true FOC has this same noise as well as a stumble at high rpm low load conditions. Like cruising at max speed on flat ground there are some hiccups you can feel and hear. It seems to always want some load to push against.

I have not had a chance to mess with mine in over a year but I was hoping the new firmware would have solved some of my issues.

These motors apparently have some design flaws with the position of the hall sensors and interference coming off the windings. I experimented with external hall sensor mount and a magnet ring on the shaft. This greatly improved the issue with the high RPM low load condition. I don't remember if it got rid of the dead cat noise.

I would suggest you mount it all up and ride it. It might run just fine when loaded.

Here was the external hall sensor experiment. Getting the halls up away from the windings really helped. There were curved magnets in the dark grey ring and the hall bracket was adjustable to experiment with the timing position. It's a lot of machining and trouble for a cheap motor now that the IPM versions of this motor are available. I figured the sensorless algorithms are so good now that by the time I got around to making an internal version of all of this the firmware would have made up for these design flaws of this motor.

 

[Rewind3000]

welcome to the club. I think most of us with that LMX SPM motor have this same problem. I was annoyed by that noise as well but in normal riding it never really spends any time there. I got PID tuning suggestions from the Nuc team but never got the bugs completely worked out. Mine runs fantastic in BLDC sensored mode but true FOC has this same noise as well as a stumble at high rpm low load conditions. Like cruising at max speed on flat ground there are some hiccups you can feel and hear. It seems to always want some load to push against.

I have not had a chance to mess with mine in over a year but I was hoping the new firmware would have solved some of my issues.

These motors apparently have some design flaws with the position of the hall sensors and interference coming off the windings. I experimented with external hall sensor mount and a magnet ring on the shaft. This greatly improved the issue with the high RPM low load condition. I don't remember if it got rid of the dead cat noise.

I would suggest you mount it all up and ride it. It might run just fine when loaded.

Here was the external hall sensor experiment. Getting the halls up away from the windings really helped. There were curved magnets in the dark grey ring and the hall bracket was adjustable to experiment with the timing position. It's a lot of machining and trouble for a cheap motor now that the IPM versions of this motor are available. I figured the sensorless algorithms are so good now that by the time I got around to making an internal version of all of this the firmware would have made up for these design flaws of this motor.

View attachment 358602

Hello,
Thanks for the explanation and for letting me catch up.
I am certain I have overcome the knocking noise at higher RPM, at least during my tests without load. Have a look at my PID Regulator values, and try them out if you dare. I think it requires a restart to take effect.

I also heard IPM was better, but Richard Liu from Senli Motors convinced me that SPM is stronger than IPM.
A shame field weakening is not advised on SPM motors, as I am having pretty cool results with my first attempts.

How do you change to BLDC mode instead of FOC? The Nucular 12F controller manual mentions "BLDC" one time in the first sentence.

Richard Liu seemed sincere when he wrote it was quiet when they tested the motor in their lab. Maybe I will just do like LMX and try another controller.

 

[DanGT86]

I will definitely try your PID values when I get around to working with this setup again. That knocking at high rpm is annoying even if it doesnt hurt anything.

I could see how SPM might offer more torque because the magnets are on a larger radius. With field weakening and advanced tech in the controller world I'm not surprised to see IPM in the majority of new products. I don't believe the field weakening will hurt anything with the SPM. My understanding is that it simply didnt respond as well when trying to extend the RPM range.

I have to snoop around in my menus to find where you switch the control type. I just remember there was a place where you pick the control strategy FOC, sensorless, BLDC etc.

Mine would cut out under full load from a dead stop in FOC mode. In non FOC regular sensored BLDC mode it did not. Again I think it was this annoying motor to blame. Years ago sellers of this motor used to say it was not for use with sine wave controllers.
The long term goal is to tune it to the superior FOC control strategy but if you just want to get your project rolling while slightly less efficient you can run it in non FOC mode and it will do well.

Even using the old fashioned low tech BLDC sensored mode the nucular was still the best controller I have found. I know that is ironic because I am here complaining about it but it is superior to everything else I have tried. I have never used a controller that is as good as this one when it comes to not having a deadband in the throttle. When ramping on the throttle from a roll it responds instantly while being completely controllable. That requires so much tuning to get right in any other platform. I'm sold on the Nucular platform for all my projects. I just wish they were easier to get.

Recently I was messing with a VESC controller and having all kinds of annoying issues with an old cyclone motor. I hooked up the NUC to it and it was riding perfectly in about 5 min. Nailed the auto config on the first try unlike the VESC.

 

[Rewind3000]

It is not so much mine as it is from LMX, from when I bought the controller used from a guy who had too many parts for his LMX 64. LMX did a PID tuning of the LMX 64 bike’s SL120-30 SPM motor, that I put to good use for the 50mm SPM. I will try to adjust the values myself, especially those pertaining to FOC, and see what I can come up with by using any of my senses but my intellect. Alas! with no load.
I have been knocked out by covid for the fifth time this week delaying the build. At one point I was sprawled out on the hallway floor, not able to sit up because of the fever, amongst cables and batteries and throttles, dead determined to get this thing working, if it’s the last thing I do.
I will stick to FOC, since I only found two other alternatives, Square, which was horrible, and DC-DC, which killed the motor, until a most dramatic reboot.

I heard about other reasons a motor can have a dead stop. If you set the Battery>Supply Max Voltage too low, the regen can exceed that and motor is disabled while you go downhill. Simple fix, just set Supply Max Voltage higher.

 

[Rewind3000]

Found an article confirming this phenomena:

Why is my DC motor whining at a lower PWM frequency?

Some DC motors have a whining sound when using PWM to control speed. Learn the pros and cons of adjusting PWM frequency and how it affects motor whining.

www.progressiveautomations.com

And a wiki page describing eletronic hum with a single word - magnetostriction:

Magnetostriction - Wikipedia

en.m.wikipedia.org

Maybe something can be adjusted in here:

 

[DanGT86]

If it is PWM noise that makes sense to me. The pulse width required to dish out such a tiny amount of power at low RPM and no load must just be low enough to get well into the range where we can hear it.

Considering you will be using that motor with gearing it will be so quickly out of this rpm zone you will likely never hear it in use.

Vasily added encoder support not too long ago. If you have a motor with an encoder you can use it instead of hall sensors. I don't think its worth trying to switch it over to an encoder just for this noise.

Its up to you how much the noise thing bothers you but based on my experience I think its safe to say your setup is working safely and properly such that I wouldn't let this prevent you from using that motor and controller.

 

[Rewind3000]

To conclude I attach the answer from Nucular team:

«Yes you can try changing efficiency to ripple.
Other than that there is not much to change, it is a combination of factors where FOC sectors at low pwm randomly switching and PWM have current insertions to measure current at low duty cycle.»

I almost finished last night, but turns out my chain is too short and motor mounts were too flexy. And chain routing needs a rethink.
At least rear wheel looks badass.

Suggestions for motor mount?

 

[DanGT86]

That is a cool project. Did you build the frame yourself?

Have you considered using the swing arm pivot as the location of a jackshaft so you have the same length chain through the entire suspension travel?

Sheet metal plates held on by the threaded bottom bracket cups are usually pretty stable motor mounts.

 

[Rewind3000]

That is a cool project. Did you build the frame yourself?

Have you considered using the swing arm pivot as the location of a jackshaft so you have the same length chain through the entire suspension travel?

Sheet metal plates held on by the threaded bottom bracket cups are usually pretty stable motor mounts.

Yes. Motorcycle light stents are not the trusty motor mounts I thought they were.

The frame is custom and just ended up like this after the it got a new random rear triangle. Then I had it welded to get a linear leverage ratio.

I did some simulations in Linkage X3 (bikechecker.com) and the absolute shortest chain stretch is when placing the motor as far away as here:

With rollers for top and bottom of the chain, ofc.
This requires a longer chain than the LMX 64. I am ordering a stump of D.I.D. 219 HTZ and two chain links. 219H chains are so short, but thankfully I already have the LMX chain to add on to.

 

[DanGT86]

Neat!

I have spent way to many hours thinking about my own build and playing with linkage software. It is annoying how long the chain ends up.

After years of regular mountainbiking I am really concerned about the 219 chain flinging around. The 219 is so heavy compared to regular bike chain. I am absolutely terrified of a chain derailment on my solid mounted non freewheeling 219 design. I finally settled on a plan that uses the swing arm pivot as the common shaft between the motor chain and drive chain so I can tension the 219 rear chain once without having to worry about the suspension slackening the chain. The thought of regen is nice as well.

Also, the way I ride I really don't want the motor down low in front of the bottom bracket. That part of the bike is basically a 3rd tire for me when it comes to big rocks.

I think this size motor and 4-6kw mid-drive is my ideal ebike. Your thread here is making me really want to focus on getting mine done.

 

[Rewind3000]

Neat!

I have spent way to many hours thinking about my own build and playing with linkage software. It is annoying how long the chain ends up.

After years of regular mountainbiking I am really concerned about the 219 chain flinging around. The 219 is so heavy compared to regular bike chain. I am absolutely terrified of a chain derailment on my solid mounted non freewheeling 219 design. I finally settled on a plan that uses the swing arm pivot as the common shaft between the motor chain and drive chain so I can tension the 219 rear chain once without having to worry about the suspension slackening the chain. The thought of regen is nice as well.

Also, the way I ride I really don't want the motor down low in front of the bottom bracket. That part of the bike is basically a 3rd tire for me when it comes to big rocks.

I think this size motor and 4-6kw mid-drive is my ideal ebike. Your thread here is making me really want to focus on getting mine done.

View attachment 358829

Where did you find that high pivot sprocket? I have spent all day at work thinking about how to build the high idler sprocket for the 219 chain. You may have noticed that no sprockets with bearings are sold for 219 - only 35 chains and up. I could also go for a deflection roller in my design, but I am afraid of putting anything made of plastic in this place.

I will go ahead and put the motor in front of the BB, but high enough to clear the ground as my bike is already super low. But I will probably contact the welder again.

The seller of the LMX 64 warned me about lightning rod solutions, and the added complexity. So without knowing any better, I will warn you as well. I may just be jealous since I can’t do this on my bike.
I am confident I can make a plastic chain guard somehow to protect the bike - and myself.

 

[DanGT86]

The sprockets in the pic are just waterjet cut stand ins to test the alignment of everything. The chain is crazy close to the swing arm and the cranks. I cheated by using 83mm cranks with spacers on a 73mm BB. That moves the crank out just far enough to clear that pivot sprocket.

I have a full machine shop at my disposal so my solutions can be a bit custom.

there is a company called Buller that makes 2 stroke racing clutches. They have a variety of 219 sprockets with needle roller bearings in them. I believe the bearing sizes are common nominal sizes. They are a bit expensive but I feel like they are higher quality sprockets than the cheap out of round $8 ones I have ended up with for 219.

Sprockets & Misc. Bully Items - Buller Built Bully Clutches

My motor was a 42kv unit from a classified ad that LMX was running a few years back here on ES. I wish I would have bought some extra sprockets for it at the time. Back then the LMX site was a little less official and they responded quickly to messages here on the forum. Now the site is fancier and harder to navigate. I would love one of the original LMX bikes when they were more like a motorcycle and less like a mountainbike. That looked like a super fun platform to try all kinds of different motors on.

As for lightning rods. He has been developing parts here on the forum for years. I like that he has always provided options for people who want to go way outside of reasonable power limits. Most companies offering kits tend to be very conservative with power. I have not done any business with them so I can't attest to their customer service. I do think the lightning rods stuff looks extra complex and heavy. Lots of extra bearings and the belt primary probably doesn't deal well with mud. Gearing down 6kw just to overdrive it at the cranks is something I don't want to do. I may end up buying a motor and sprockets from lightning rods one of these days. Better than trying to get one from aliexpress. I think the new ipm version runs pretty well.

 

[Rewind3000]

Of course! Why didn’t I just water jet my sprockets! If ever you make some, consider me!

I tried to buy some parts from LMX this summer they eventually ignored me.

I was smart enough to buy the Hope Evo crankset that I recently rebuilt with a super boost spindle, adding 5mm to each side.

The LMX 161 is still available?

 

[DanGT86]

Yeah, its a bummer that LMX may have drifted away from the relatable customer service. It has to be really hard to transition from a one at a time hand builder in a shed to a large company. I remember back when the original LMX was developed on the forum. Lots of people followed along and there was lots of back and forth with suggestions during the development. This was pre sur-ron and there was nothing else like it. There was a time when you could buy the frame as a kit and put your own parts on it.

I'm not actually sure if the LMX 161 is still available. It was far out of my price range. I really liked the simplicity of the original LMX. Its less intimidating for me to weld to a simple steel frame to try experiments. I'm usually pretty bored with nice finished stuff.

This is the LMX I wish I had bought.

 

[Rewind3000]

Yeah, its a bummer that LMX may have drifted away from the relatable customer service. It has to be really hard to transition from a one at a time hand builder in a shed to a large company. I remember back when the original LMX was developed on the forum. Lots of people followed along and there was lots of back and forth with suggestions during the development. This was pre sur-ron and there was nothing else like it. There was a time when you could buy the frame as a kit and put your own parts on it.

I'm not actually sure if the LMX 161 is still available. It was far out of my price range. I really liked the simplicity of the original LMX. Its less intimidating for me to weld to a simple steel frame to try experiments. I'm usually pretty bored with nice finished stuff.

This is the LMX I wish I had bought.

Surron, isn’t that for asian guys? I prefer starting with a bicycle, as I don’t need a motorcycle, just an active feeling e-bike. Then focus is more on fit and geometry. I doubt a Surron can be used by fully grown Scandinavian male.

What kind of battery will outout the amps you need for 8kW, on this Transition bike frame?

Again, I prefer starting with a bicycle frame, designed for speed. Not just - something with a motor. Then you end up with something overpowered and uncontrollable with road bike qualities.

 

[DanGT86]

The transition will probably end up with more like 4kw. My Nucular controller is a 6fet and I wouldn't want to try to package much more than 10ah at 75v (20s). Going for high power light ebike rather than motorcycle with that one.

I'm currently still using hobbyking Lipo batteries. Id like to eventually get a nice 21700 pack built. I have learned my lesson about buying the expensive parts first. Now days I make sure I get a running functioning project before spending the big money on batteries. The lipo batteries are cheap, power dense, and modular. They are versatile until they burn your house down.

I have a specialized Big Hit DIY ebike with a mxus hub motor that will do 7kw. Its fun to launch it on full power in open spaces but its not really that practical for much else. When I am off road I use almost exclusively the 3500 watt setting. Its nice to have the full 7kw for big hill climbs or blasting through grass. In my experience 3500 watts is more than enough for anything resembling a mountain bike trail.

I'm sure there are tons of varying opinions on power. I feel like above 6-7kw is just a totally different type of riding than a bicycle frame. At high power I want a large seat I can move around on like a motorcycle. I would expect to lose traction often and put my foot down a lot in tight quarters and mud. Once the motor is enough to frequently break traction and or power wheelie from cruising speed I think its beyond pedals and a bike seat.

I have always liked that LMX made mountain bike like products that had respectable power. They are the only company I know of that bridge that gap in the market between 500watt oem Ebikes and motorcycles. Sadly I am too cheap to ever buy one. My goal with the transition is to end up with an LMX like mountain bike.

 

[Rewind3000]

The transition will probably end up with more like 4kw. My Nucular controller is a 6fet and I wouldn't want to try to package much more than 10ah at 75v (20s). Going for high power light ebike rather than motorcycle with that one.

I'm currently still using hobbyking Lipo batteries. Id like to eventually get a nice 21700 pack built. I have learned my lesson about buying the expensive parts first. Now days I make sure I get a running functioning project before spending the big money on batteries. The lipo batteries are cheap, power dense, and modular. They are versatile until they burn your house down.

I have a specialized Big Hit DIY ebike with a mxus hub motor that will do 7kw. Its fun to launch it on full power in open spaces but its not really that practical for much else. When I am off road I use almost exclusively the 3500 watt setting. Its nice to have the full 7kw for big hill climbs or blasting through grass. In my experience 3500 watts is more than enough for anything resembling a mountain bike trail.

I'm sure there are tons of varying opinions on power. I feel like above 6-7kw is just a totally different type of riding than a bicycle frame. At high power I want a large seat I can move around on like a motorcycle. I would expect to lose traction often and put my foot down a lot in tight quarters and mud. Once the motor is enough to frequently break traction and or power wheelie from cruising speed I think its beyond pedals and a bike seat.

I already went back to 26er rear wheel, I can just switch to banana seat as well. I was too young to ever get to enjoy one as a kid.
It is amazing where this project is taking me.

But should we do a collaboration to develop the 219 idler wheel, since we both seem to need one?

 

[DanGT86]

All the cool people ride 26". You are on your own with the banana seat.

I am the worst person to collaborate with. I have hard time completing projects. I also have very little free time and a broken foot that is currently keeping me out of the shop. But I highly suggest you check out that link to the Buller website. Their 219 sprockets are already heat treated to the proper hardness and have tooth profiles that work great on high end race karts. It would be cheaper than prototyping sprockets and having them heat treated.

I have heard 219 chain was used in timing chains in motorcycle and car engines. If you have a motorcycle salvage yard near you that might be a source.

Jackshaft/idler is going to require 2 sprockets on one hub. I was going to try pre-hard material and see if it holds up. Sending parts off to get heat treated for proper hardness is expensive.

What do you envision needing in an idler? Just a straight sprocket with a bearing?

I will let you know when I get around to machining mine but please don't wait for me if you ever want to finish anything.

Where are you located?

 

[Rewind3000]

All the cool people ride 26". You are on your own with the banana seat.

I am the worst person to collaborate with. I have hard time completing projects. I also have very little free time and a broken foot that is currently keeping me out of the shop. But I highly suggest you check out that link to the Buller website. Their 219 sprockets are already heat treated to the proper hardness and have tooth profiles that work great on high end race karts. It would be cheaper than prototyping sprockets and having them heat treated.

I have heard 219 chain was used in timing chains in motorcycle and car engines. If you have a motorcycle salvage yard near you that might be a source.

Jackshaft/idler is going to require 2 sprockets on one hub. I was going to try pre-hard material and see if it holds up. Sending parts off to get heat treated for proper hardness is expensive.

What do you envision needing in an idler? Just a straight sprocket with a bearing?

I will let you know when I get around to machining mine but please don't wait for me if you ever want to finish anything.

Where are you located?

Considering the Lightingrod champion of complexity, I will go for the easiest and most common solution - a "reversed deflection slide", or like motorcycle people call it, a chain slide:

I did try a chain guide with cog for bicycle chains and it was a noisy and draggy experience.

Of all the bicycle chain guides I tried, Straitline worked best and was the most quiet. Then I found out about narrow wide chainrings and didn't need it anyway.


The latest simulation kind of looks like the Joker.


You transferred your fear of flying chains to me and I will cover everything up with a big plate, that will also be handy to place random slides and rollers. The welder will help me secure it in place.