Luna BBSHD Ludicrous V2 controller documentation

[marcos]

I'm being told here that each controller is connected over BT to test the connection, so it should be unlikely to have blank BLE modules. Sounds like both of you are having the same issue, if one of you could take the controller off and reprogram it we would all learn something...

I'm attaching the bit of internal documentation related to the BLE programming, you would need to install vesc tool for windows, or vesc tool for linux from here https://vesc-project.com/node/17 and connect over usb.

[choochoo]

I'm being told here that each controller is connected over BT to test the connection, so it should be unlikely to have blank BLE modules. Sounds like both of you are having the same issue, if one of you could take the controller off and reprogram it we would all learn something...

I'm attaching the bit of internal documentation related to the BLE programming, you would need to install vesc tool for windows, or vesc tool for linux from here https://vesc-project.com/node/17 and connect over usb.

I have the controller connected via USB and have the Windows VESC tool open. I followed the five directions then power cycled the controller.

I initially left the USB plugged in and didn't see it listed in Bluetooth. Once I removed the usb and power cycled my Android found it!!

So it looks like a batch of controllers escaped without being properly programmed.

It would be very helpful if post #1 had valid links to instructions, files, user manual, etc. I initially used the early user manual etc. only to later find that post #88 that has better info.

[marcos]

Thanks a lot for trying the programming and reporting back, I'm glad it solved your problem. Yeah usb connection provides its own power so that didn't count as a power cycle.

I'm still not fully convinced that the bluetooth was not programmed, in fact it was more likely to be a disconnected antenna right before being potted as it happened to me during development. If this happens to you again please let me know, there is a non-zero chance of having the BT module memory getting corrupted with the EMI inside the housing.

Um, the 1st post and the post #96 are pointing to the same user manual, and all the links to files are into the manual. I think one centralized doc is a tidier approach

[Dlew]

i'm not totally clear; is it working now...are you able to see the controller now without the USB connection?

[choochoo]

Yes that solved the problem. My Android found it instantly after that. I changed settings to 72V and the system works!

[xcnick]

Hello Marcos. The motor stutters and stalls when starting on a hill, even in a low gear. I just set up another V2 and this one is really bad and shutters on flat ground. They both have the same settings. Is there a setting like linear throttle I should play with to get it to pull a heavy load from a dead stop?

[xcnick]

AZ guy, do you think it matters? It sits in a close tollerance ring, seams to work, it just might be clocked a thousands off with only the bolts.

[AZguy]

If I didn't have one on hand I might run without it until I could get one in... but I don't know, I haven't had mine apart in a long time... those pins locate things pretty precisely and it would be odd if a chinese supplier where cost is near the top of the list of concerns would put a part, no matter how inexpensive, in if it didn't provide good product value (reliability, etc.)

OTOH I remember being very impressed when I saw how the plastic gear (primary reduction) cover had threaded holes to allow the screws that hold the cover on to be used as a cover puller since it would be hard to get that off without... they had to thread those holes which costs money and they don't provide any value from reducing assembly labor costs or adding reliability, just for disassembly

[marcos]

oh that pin, no clue where to get those but I reckon they are important. The air gap between rotor and stator is only 0.5mm so it needs to be precisely centered

[xcnick]

I did the swap and now they both work properly. Bad connection is my thought. the long white connector goes into a different colored female in each motor. The problem setup has a green connector and I can't get the white one from the controller to lock into place. It doesn't have a lip for it to clip to. I think with the installation I just did, it didn't wiggle loose and why it works now. I started a new post so I could give a picture of both.

[xcnick]

doh! I looked at the old controller and no connector stuck to it. I may have made this mistake years ago. It looks like if it sits in there just right it still makes contact with the board. I will go through my bone yard of controllers in the next couple of days. Will I have to solder it back to the board if and when I find it?

[marcos]

If its making contact and working I'd leave it that way. The point of the white plastic is to function as a latch, but if the whole thing came off I reckon it could slip out again.

[xcnick]

How many mistakes can one guy make? I happen to have a motor core in storage, it might solve the alignment pin and connector boneheads.

[xcnick]

I ran a hill climb at 72 volts which only made it heat up to about 65 centigrade. At 52 it did get hotter over the same task.

[marcos]

Cool, it looks like an easy ride, about 2kW with a brief 3kW peak, at about 1/3 of the available torque

[xcnick]

I am learning it doesn't take as much power as I thought. I am more rock crawling than blasting up. I hardly crack the throttle, but it has to be precise or I loose traction. The ramp up was working well for me. Id amps only went to about 10, so it was a fairly efficient climb. This analysis program is an eye opener.

[xcnick]

The number of cells does not always follow the change from 72 to 52 volts and back again.

[marcos]

Oops, let me check. That number only affects the battery SOC gauge, doesn't affect anything else so I might have missed that

[marcos]

I can't reproduce the issue... Is this happening when you tap the voltage presets (WOLF 52V / DIRE WOLF 52 / POWER WOLF 72V) or when you read/write the settings to the controller?

[marcos]

Oh sorry, I couldn't reproduce it because I didn't notice I was running development firmware. It looks like the stable firmware you have has that bug and its already fixed for the next release

[Harrisy]

Connected successfully to the VESC app for the first time. Followed the firmware update instructions, and now I'm getting error code 30H and can't connect to the VESC via Bluetooth due to it not recognizing the firmware. I'm not sure what I did wrong. Do I need a USB reflash?

[marcos]

If a controller power cycle and killing the phone app doesn't fix that, I would flash over USB

[StvMan]

Marcos,

PAS not working

In previous firmware and/or VESC Tools, I was allowed to set my %PAS and Crank Rpm to initiate PAS.

Under the old Firmware/VESC Tool I had set those to 0% and a high RPM respectively.

Could it be that those parameters are still set in the background…. thus disabling my PAS?

best,
Steve

[marcos]

Um, the settings are the same... but if you reflashed the firmware all those old settings were wiped out and returned to the default values of a working controller.
The new app skin only exposes the PAS % and hides the crank rpm settings. Those settings are still accesible over USB if you would really like to play with them.

The latest firmware has a more rugged algorithm for PAS proccesing, but I haven't heard of any bug in that so far. Maybe you are in level 1 and doesn't provide much assist but you have 100% on the throtle, even in level 1

[StvMan]

Thanks….I will take another look.

Steve

[xcnick]

Me again. How does the V2 match up to the new bbshd. Luna facebooked they got a new shipment. Specifically the bare motor Luna sells, maybe they are still selling the old ones bare.

I am driving myself nuts and five hundred bucks seems cheep all of sudden. Something inside my hd stripped. I acts like a bad clutch in the rear wheel in that it slips under power. From what I understand a bad hd clutch is when you can't pedal. I took everything apart and it looked fine... to me anyways. I put in a new hd clutch just in case, changed the rear wheel just in case it was the clutch there, and I still have the problem.

thanks
nick

[xcnick]

That is probably it. I didn't take the pinion out, I just looked at the teeth. I also didn't see that clutch for sale at Luna. Time to tear it down again. But the question still stands Marco. Am I nuts for just wanting one on the shelf anyways?

[AZguy]

The "replacement" for that clutch is to get the whole gear - the "bearing" is the clutch

Sometimes improper grease (too much, too little, wrong stuff, etc.) is enough to get that clutch not to work - if you clean it out thoroughly (brake cleaner or something) and repack appropriately it may work... until you get the better replacement ;-}

[StvMan]

XCNick, Like AZguy said … I had the same issue as you and it was the one-way clutch bearing in the plastic gear that was causing the issue… replaced it and the slipping (sounded like a metal on metal chattering noise) went away.

[marcos]

I'm not sure if I'm reading correctly, but it looks like you stripped the nylon gear? Inside the nylon gear is the one-way clutch.

I recommend upgrading to the Peek gear, some go further and use the noisier steel gear. I use the Peek gear in the bbshd dyno without any issue.

https://lunacycle.com/bbshd-silent-peek-gear-nylon-gear-replacement/

More install info in the link at the end of the page.

Its easy to strip the stock nylon gear so I think you'll be happier with 1 ruggedized drive than with 2 drives with the pretty weak nylon gear. Newer BBSHDs come with a less accurate temperature sensor that is placed in a much better location, so not much difference in the end. I should make a tutorial on how to improve the stock temp sensor to reduce the sensing lag.

[marcos]

Not much difference between 52V and 72V max RPM because RPM is software-limited such that if things go south the motor-generated voltage won't fry the controller (52V rpm is already significantly extended with field weakening) . However having much more volts available allow to drive much more phase current on the whole rpm range which makes a huge difference, it really comes to life.

[xcnick]

"It really comes to life." Is that a technical term? LOL, but that is the best way to describe it. It is most apparent when I go back to 52. I keep looking down at the motor as if there is something wrong.

[AZguy]

All makes sense - appreciate the explanation

[AZguy]

xcnick There are more than a few different kinds but the type most likely here is going to have little ramps on the inside of the outer race that catch the needle bearings going one way and let them be loose going the other... here's a picture that's close enough although I doubt there are any springs in the BBS clutches, just the needles and the ramps are enough to do the job and there are plenty of these clutches without springs, just the rollers and ramps

[xcnick]

Interesting. The diameter of the shaft must be critical to make this work. It could be a worn shaft, not a bad clutch. Speaking of clutches, I always break the clutch in the bicycle wheel hub. After a couple of sprag replacements the whole thing is useless. I have just poured JB weld, 9,000 into the whole thing to lock it up. It is now a fixie. Coasting, the chain spins like on a motorcycle. I am thinking this clutch we are talking about in the HD it is about to get a work out. The test is on a HD with a V1 Ludi controller. Single speed, I don't know if this would work with a derailleur. But I have high hopes this fixes the weakest part of my snow bike concept.

[AZguy]

That makes sense - the freewheel in the hubs just aren't built for that kind of torque

JB weld has many uses =]

I'm guessing you removed the pedals and crank arms - put some pegs on the beast?

I doubt the shaft is the problem, the forces are highest on the rollers and ramps and if worn or damaged I'd be betting on the ramps... but you never know and like I said the lubrication is really important

[xcnick]

Peddles work fine. It is hard to notice with a chain guard. I went on line to see if there exist a fixie hub in 120mm. Nope.

[xcnick]

RPM questions. What RPM does the V1 stop at? What RPM is the start of field weakening? I am guessing 3000. When I use a transmission I am thinking the most efficient acceleration is to shift at 3000 RPM and then go to 5000 RPM when in high gear and you need to go fifty miles and hour. I wish the display showed RPM.

[marcos]

Yeah those I know. The important bit that we don't know is the exact part# of the temperature sensor so we sort of reverse engineered the temp curve with focus on the 100°C region.

[AZguy]

That I didn't get

Figured you knew the poles, etc. - sort of hard to get the motor control right without

Hey while I'm thinking of it, what development tools are you using for the code? Is it an ST processor?

[marcos]

Yeah, I guess you could start from here: https://vesc-project.com/node/309
I mostly use kdevelop or eclipse, and I don't really use a debugger anymore so with a single USB cable I can do pretty much anything.
For hardcore stuff I use a CANbus<->USB dongle as its more rugged against the noise of a >15 kW drive, sometimes I'm concerned that my laptop USB port could get damaged if I screw up

[marcos]

RPM is limited to 6500rpm, field weakening start is dynamic, depends on the input voltage. If you are running at low voltage (or battery is sagging a lot) FW will start much earlier. In the logs you cab see exactly when FW happens by checking the id current.

Efficiency is best in mid-high rpm and low torque, and tanks at full throttle.

Conceptually it would look something like this (redder is more efficiency)



So basically, if you are after efficiency reduce the torque (phase Amps), its the most effective approach.

[xcnick]

What is the risk of moving from the default 5000 max RPM to 6000 or 6500?

Interesting graph, I need to look again the id from my rides.

[xcnick]

the slider in vesc. It is at 5000, but it will slide up to 6500

[marcos]

Ah, no harm if its set to 6500rpm. In fact 6500rpm should be the default in the ludicrous profile.

[xcnick]

stupid is as stupid does, can't wait to try it.