Trying to upgrade the controller on my volt bike mariner

noctrin

1 mW
Joined
May 10, 2022
Messages
15
[UPDATE] I performed the upgrade, check my last comment for more details.

TL;DR:

- KT PAS is miles better, only reason to upgrade.
- speed limit cannot be changed, motor is RPM limited.
- kinda pain in the ass to do since the controller wont fit without modifying the casing.
- all wires are compatible if u buy the right one.


Hi,

I found this board and read a bit while doing some research, but it's a tad overwhelming. Hoping someone who knows more can help me clear up a few things.

Backstory

I purchased a voltbike mariner which has a 500W bafang hub motor, I love the bike, i hate how pedal assist is done and the fact that i cannot remove the speed limit on it.

The bike comes stock with a lishui controller which doesnt seem to be modifiable.
20220422_154400.jpg
The motor is a bafang rm g060 5000 d
20220408_220357.jpg


My Plan

Swap the controller with a KT controller and use the simulated torque assist. I am planning on ordering the controller, USB cable to program it and screen since the one i have is not compatible i believe.

Questions

Can any KT controller be programmed to simulate torque assist?
I plan on buying this one: https://www.aliexpress.com/item/1005003157399358.html?spm=a2g0s.8937460.0.0.367d2e0eUzdPQY

It has (almost) the same connectors as my current one, the only difference:
- The rating which is 25A vs 20A on mine. The ui/settings imply it is pushing 20A and 1000W to the motor, but i assume it's bs since it would be fried by now. It's also not "1000W" powerful.
- It's missing a blue 4 pin connector which i need to figure out what the hell its for (picture)
20220510_211533.jpg

I plan on using this screen (i want a small one)
https://www.aliexpress.com/item/4001142752280.html?spm=a2g0s.8937460.0.0.367d2e0eUzdPQY

The cable seems to match up just fine with the current one:
20220510_212832.jpg

The usb cable to program it seems straightforward (assuming i understood correctly and any KT controller can be programmed)

Anything else i am missing, anything else i should know?

-- Thanks a lot for any information, i really appreciate it!



I atatched some pictures of all the connectors off the current controller.
 

Attachments

  • 20220510_211422.jpg
    20220510_211422.jpg
    732.6 KB · Views: 855
  • 20220510_211452.jpg
    20220510_211452.jpg
    728.8 KB · Views: 855
  • 20220510_211552.jpg
    20220510_211552.jpg
    871.4 KB · Views: 855
  • 20220510_211610.jpg
    20220510_211610.jpg
    903.4 KB · Views: 855
  • 20220510_211829.jpg
    20220510_211829.jpg
    1 MB · Views: 856
noctrin said:
The bike comes stock with a lishui controller which doesnt seem to be modifiable.
There are some threads here about programming Lishui controllers, have you checked them?

Minor warning about attempting to blindly plug in connectors when changing components: Just because the connectors appear to mate, or the wire colors seem to match, doesn't mean they are compatible. Connector pin configurations vary across manufacturers, as do wire colors, which can even vary over time (mfr. may have run out of a certain color wire on the assembly line?).

Verify the pinouts or otherwise follow a tested how-to from someone who has successfully completed the process with similar components.
 
noctrin said:
Swap the controller with a KT controller and use the simulated torque assist. I am planning on ordering the controller, USB cable to program it and screen since the one i have is not compatible i believe.

Can any KT controller be programmed to simulate torque assist?

The usb cable to program it seems straightforward (assuming i understood correctly and any KT controller can be programmed)

Torque simulation is what KT refers to in their implementation of PAS. It doesn't use a torque sensor, but a regular cadence sensor, and no programming cable is required, since all user adjustable parameters are performed via the display.
http://www.szktdz.com/en/news_show.php?article_id=553

If on the other hand, you are talking about the KT Open Source firmware, that was developed by members of this site, that does require a programming cable. The firmware is simply using the KT hardware and customized firmware, and has real torque sensing PAS, which would require adding a torque sensing bottom bracket PAS. If this is what you want, then there is a long thread dedicated to the OSF, and would probably the best place to obtain support.
 
E-HP said:
Torque simulation is what KT refers to in their implementation of PAS. It doesn't use a torque sensor, but a regular cadence sensor, and no programming cable is required, since all user adjustable parameters are performed via the display.
http://www.szktdz.com/en/news_show.php?article_id=553

My plan is to test that one out first and if solid, i will stick to it. Do you have any idea if all KT controllers support that functionality, or only some of them do?

E-HP said:
If on the other hand, you are talking about the KT Open Source firmware, that was developed by members of this site, that does require a programming cable. The firmware is simply using the KT hardware and customized firmware, and has real torque sensing PAS, which would require adding a torque sensing bottom bracket PAS. If this is what you want, then there is a long thread dedicated to the OSF, and would probably the best place to obtain support.

This was going to be my plan B, adding a torque sensor seems fairly trivial, but the custom firmware seems to be a bit tricky to configure the motor and requires some guess-work.

Thank you for the information!
 
99t4 said:
Verify the pinouts or otherwise follow a tested how-to from someone who has successfully completed the process with similar components.

I found that thread, apparently the controller needs to have an "-f-" in the model number to be programable. Mine sadly, does not.

Thank you for the info on the pinouts, i assumed it would be standardized. I will make sure to double check them.
 
Just in case someone finds this --

I performed the upgrade:

1. It was a pain in the arse, the kt controllers are bigger than lishui and it required some creative carving with my dremel to make the casing smaller. Just dont mess with the side the mosfets are on since its also a heatsink. But.. you can fit the 22A in there.

2. The speed limit was not enforced by the controller, but rather the motor. With the new controller, after fumbling through the configs and managing to properly do it. It cut out at the exact same speed, ~28.7 km/h. This is about 300 RPM and with this info and some google, seems these hub motors are in fact RPM limited. So, if you have a bike with 20" wheels, your top speed is 29 km/h no matter what you do it short of increasing the diameter.

3. I couldnt find a controller that had a rear-light plug with brake indicator. I ditched the brake indicator and wired it to the front light.

4. The torque simulation PAS is just miles better than the speed one, this alone was worth the upgrade.
4a) For the PAS sensor config (C1) it's 5-7 with 5 being the highest sensitivity. Id keep it on 5. Anything else will not work at all.

5. Old controller was 20A, new one is 22A. While not mind blowing, i can actually feel the slight bump in power and its a bit more stable. The motor did not get hotter.

6. I got a sinewave controller, pretty sure the lishui it came with was also sinewave as i feel absolutely no difference.

7. I got the smaller KT-LCD4 which mount so nicely where the buttons on the old used to be, leaving room for my phone in the middle. I found the big screen completely useless and a waste of space.


Overall:

Difficulty: 7/10 -- not for the faint of heart, you need to re-route some cables. You need to cut the controller casing. Make sure you get a bafang motor extension 3ft is enough. It took me about 2-3 hrs.
All the wires matched up properly except the lights.

Worth it: Depends. If you're happy with the pedal assist, i would not bother changing it.

IMO, the simulated torque is just miles better. You get a little bit of extra power. You get the cruise control function. You also have the option of getting the small lcd4 and opening up some realestate.

The other reason i swapped it was to increase the speed limit to about 35, but that's impossible without changing the motor to something thats willing to do more than 300 RPM. But.. for about 80 canadian rupees, the bike feels way smoother on PAS.
 
noctrin said:
2. The speed limit was not enforced by the controller, but rather the motor. With the new controller, after fumbling through the configs and managing to properly do it. It cut out at the exact same speed, ~28.7 km/h. This is about 300 RPM and with this info and some google, seems these hub motors are in fact RPM limited. So, if you have a bike with 20" wheels, your top speed is 29 km/h no matter what you do it short of increasing the diameter.

Your motor will not go faster no matter what current is applied, this is true since the speed of your system right now is not current limited.

All common motors used for ebikes are rpm limited by the motor winding configuration and voltage of the system. It’s not enforced, it’s just physics. If you want it to go faster above the kV*Volt rpm you can either increase voltage of the system, buy a motor with a higher kV or get a controller with field weakening function.
 
larsb said:
noctrin said:
2. The speed limit was not enforced by the controller, but rather the motor. With the new controller, after fumbling through the configs and managing to properly do it. It cut out at the exact same speed, ~28.7 km/h. This is about 300 RPM and with this info and some google, seems these hub motors are in fact RPM limited. So, if you have a bike with 20" wheels, your top speed is 29 km/h no matter what you do it short of increasing the diameter.

Your motor will not go faster no matter what current is applied, this is true since the speed of your system right now is not current limited.

All common motors used for ebikes are rpm limited by the motor winding configuration and voltage of the system. It’s not enforced, it’s just physics. If you want it to go faster above the kV*Volt rpm you can either increase voltage of the system, buy a motor with a higher kV or get a controller with field weakening function.

Amazing, I have never heard of this. Thank you for the detailed explanation!

Would you happen to know of any controller that is capable of doing this?

Thanks!
 
noctrin said:
Amazing, I have never heard of this. Thank you for the detailed explanation!

Would you happen to know of any controller that is capable of doing this?

Thanks!

Its not really a limit, it's just the speed of the motor for the voltage being applied (motor Kv rating). Higher voltage, higher speed; lower voltage, lower speed. Your controller is basically varying the voltage applied to the motor, based on the throttle input, to make it go faster or slower. But the max voltage the controller can provide the motor is just shy of the full battery voltage supplying the controller; and the max speed (what you refer to as limit) occurs when the voltage from the controller to the motor hits the wall where the voltage is equal to the back EMF of the motor at that speed. So, you can't just change out the controller for more speed, you have to up the input voltage (from the battery, increasing the voltage at max throttle).

The exception to just increasing voltage for higher speed is if you have a controller than can apply field weakening. That feature pumps more current into the controller to reduce the back EMF (moves the wall), thereby allowing the motor to attain a higher speed.
 
E-HP said:
noctrin said:
Amazing, I have never heard of this. Thank you for the detailed explanation!

Would you happen to know of any controller that is capable of doing this?

Thanks!

Its not really a limit, it's just the speed of the motor for the voltage being applied (motor Kv rating). Higher voltage, higher speed; lower voltage, lower speed. Your controller is basically varying the voltage applied to the motor, based on the throttle input, to make it go faster or slower. But the max voltage the controller can provide the motor is just shy of the full battery voltage supplying the controller; and the max speed (what you refer to as limit) occurs when the voltage from the controller to the motor hits the wall where the voltage is equal to the back EMF of the motor at that speed. So, you can't just change out the controller for more speed, you have to up the input voltage (from the battery, increasing the voltage at max throttle).

The exception to just increasing voltage for higher speed is if you have a controller than can apply field weakening. That feature pumps more current into the controller to reduce the back EMF (moves the wall), thereby allowing the motor to attain a higher speed.

My apologies, i was asking if they (or you) happen to know of any controllers that can apply field weakening. Moreover curious if you'd happen to know how much this actually can help -- ie: if it cuts out at ~28km/hr would that realistically be able to push it to 35.

I found this -- https://ebikes.ca/product-info/grin-products/frankenrunner.html

They're also local which is great, seems i need the controller, cycle analyst and probably a torque spindle. so about ~600, not sure if it would be worth it. If I could use the controller only, that would make it extremely appealing.

Random question if anyone knows -- i checked the load on the stock controller with an ammeter, it was pulling about 17A under load, given the 48v battery and 500W rating, it doesn't add up. The motor is not heating up or anything, but i have no idea where the 500W rating is coming from.

I really appreciate all the great information, i'm learning a lot. Thank you guys.
 
noctrin said:
My apologies, i was asking if they (or you) happen to know of any controllers that can apply field weakening. Moreover curious if you'd happen to know how much this actually can help -- ie: if it cuts out at ~28km/hr would that realistically be able to push it to 35.

I found this -- https://ebikes.ca/product-info/grin-products/frankenrunner.html

They're also local which is great, seems i need the controller, cycle analyst and probably a torque spindle. so about ~600, not sure if it would be worth it. If I could use the controller only, that would make it extremely appealing.

That controller, the Phaserunner, and Baserunner controllers from Grin all support field weakening. Others manufacturers do as well. The search tool at the top of the page works well.
 
Some more updates -

Picked up the baserunner Z9;

It has 3 ways of being used, i planned on using a 3rd party display and using my current harness which is well embeded in the body. This was by far the cleanest but, unsupported option.

I got most things working (well, with only light testing) except the display. I had 2 on hand - an 800S/C965 (not really sure which is the proper name or if there are 2 varations of the same one with slight differences, more on this later) and a kt display.

The kt is totally clueless about everything and i didnt expect it to work, the lishui one reads the battery level but shows a com error. I assume it measures battery using the input wires.
Both will turn the controller on and off (not much of a surprise, simple mechanism).

Their docs on this are very light: km5s protocol display and i was pretty sure the lishui one uses that, but i believe the official name is UART-5S (?). I looked in their baserunner software and there are flags for display protocol, but without documentation i have no idea what the values are and what they map to (if i get desperate enough i might start probing there).

Difficulty wise, it was annoying. I had to cut the plug off the controller, i used the kt controller plug as a donor and soldered the wires -- they have a breakout board, but it was not in stock. I mapped the pinouts by looking at the board markings on the kt conntroller (since i knew it worked) and matched them to the schematics. The motor spins properly, the thorttle works, and brakes also work. Display wiring seems correct, i tried reversing the tx/rx since some online schematics implied it, but that wasnt it.

As far as i can tell, this needs a bafang display or kingmeter, i found some diagrams from eggrider that imply bafang uses a different pinout and they had lishui/kt as the "other" (both mine use the 'other' -- so, it seems the lishui one might have the same "hardware" but they come in 2 flavors and i have the wrong one). I can't find a single kingmeter or bafang display locally, so it looks like i'll be waiting a few weeks to get one from China. I am hoping the guys from Grin will get back to me and share a bit more info to make sure i get the right one.

The cadence sensor i didnt try yet, the big gotcha is their plug supplies 12v and most will need 5v. Conviently, the control cable has 5v on it for the throttle and i plan on pulling it from there. I plan on double checking voltages with OEM controller and verifying the 5v, worst case i have a few ways to change that 12v to whatever i need. Leaving this for last..

The config options and parameters are amazing, the build quality is phenomenal. Great little device, quite impressive. Size wise, absolutely no issues fitting it in, it's unbeliveably tiny. The metal housing it goes in will act as a heatsink and i plan on mounting the controller to it with some thermal compund.

It's MUCH smaller than the 22/25 kt and can supply almost twice the current.

I like that i can also quite easily integrate a proper torque sensor should i choose to down the road.

I hope i can sort out the display, if not, i might have to suck it up and use their CA3. But, i only need it for battery and changing PAS, so.. not too eager to go down that road if i can avoid it.
 
Did you ever get that working? I really like the baserunner, but not so much the cav3
 
Back
Top