Controller and PAS behaviour

[jasonahf]

Question if anyone has run into this...

Bought an ebike kit off ebay for my 12 year old Son. (https://www.ebay.ca/itm/295146889181?var=593373846750)

The controller isn't behaving as expected to though. Main problem is that it doesn't seem to have an ability to limit current.
According to instructions I found for the SW900 display, I should be able to modify parameter P14 to limit current... this doesn't work. Mainly I'm interested in this as the 1000w hub motor is (according to the display) maxing out at 1300w under acceleration. Which is drawing up to 27a. the battery BMS is limited to 20a and keeps cutting power when operating over 1100w. can feather the throttle to keep it lower but it's hard to control it reliably.

Second issue is the behavior of the PAS settings. Has 5 levels, but these seem to only limit speed and not power. So in level 2 for instance, if you start pedaling you get the equivalent of full throttle (1300w) until speed is about 30km/hr (18-19mph) and then drops down to 300-500w to maintain speed.

I could always upgrade the BMS on the 13s5p Battery to 30a... but I'm concerned about what could happen when riding in winter on potentially slippery roads if it's constantly giving full power on acceleration.

Guess I'm wondering if this is normal... is there something that can be changed... or should I look for a better controller? and is so best option to work with the current hub motor.

Any assistance would be appreciated.

 

[amberwolf]

Bought an ebike kit off ebay for my 12 year old Son. (https://www.ebay.ca/itm/295146889181?var=593373846750)

The controller isn't behaving as expected to though. Main problem is that it doesn't seem to have an ability to limit current.
According to instructions I found for the SW900 display, I should be able to modify parameter P14 to limit current... this doesn't work. Mainly I'm interested in this as the 1000w hub motor is (according to the display) maxing out at 1300w under acceleration. Which is drawing up to 27a. the battery BMS is limited to 20a and keeps cutting power when operating over 1100w. can feather the throttle to keep it lower but it's hard to control it reliably.

What happens when a setting is changed in the display depends on what the controller supports; what it does for that particular setting value, or if it responds to that setting at all. That depends on the controller model, and what firmware it has, vs the display's. Unfortunately there is no listing I've seen of which models or versions of anything support what.

It's also possible the controller supports everything you need it to, but the display doesn't know how to tell it to do it (same problem as above, but different solution--new display instead of new controller...but you probably won't know without replacing one or the other)


If you don't mind some DIY modification of the controller, you can physically modify it's shunt to reduce it's current capability. It's not as controllable as a setting, but it works. If you're up for it I can provide some details, if you can post pictures of the inside of the controller (usually the end where the battery wires connect to the board inside).

Second issue is the behavior of the PAS settings. Has 5 levels, but these seem to only limit speed and not power. So in level 2 for instance, if you start pedaling you get the equivalent of full throttle (1300w) until speed is about 30km/hr (18-19mph) and then drops down to 300-500w to maintain speed.

Most of the PAS controllers I've worked with use speed limits on each level. So if you pedal (at all, regardless of how fast or hard), it turns on full output until it reaches the speed limit of that level.

Some controllers (like at least some of the KT series) use what they call "torque simulation" (which it isn't) where they limit current (or power) for each level instead of speed. They're supposed to be much more controllable, more rideable, though I haven't used one like that myself.

Another option, though expensive, is to use a Cycle Analyst v3 from Grin Tech http://ebikes.ca and move the PAS sensor and throttle output to the CA3 PAS input, instead of connecting to the controller. Then connect the CA3's throttle output to the controller. Set the controller up for normal max output; if it will run "headless" without the display, and be at max, is easiest. Then the CA3 can be setup to pass the throttle thru only limiting current to , and use the PAS signal to provide throttle amounts equivalent to how fast you are pedalling, so it just adds power to the pedal power in whatever proportion you prefer. (this is how I use the CA3 on my SB Cruiser trike) You can also set speed limits, and use remote handlebar controls to change presets or settings so the CA3 itself is not visible, if you don't like it's looks.

 

[docw009]

You're right, the SW900 is a good choice for fast throttle equipped bikes, but a poorer choice for pedal assist ridden slower. Mine had the same issues, and got early retirement.

 

[jasonahf]

I'm no stranger to modifying things LOL.

The cycle analyst option is most likely out of my price range... but modifying the shunt... definitely something I can do.

from what I've read, most people are modifying the shunt to decrease the resistance in order to gain amps. Mostly by adding more solder or copper wires in parallel etc.

in order to increase the resistance I'm guessing I would either need to lengthen the shunt wire (not easily done) or I've read a suggestion that thinning out a section of the shunt wire to increase it's resistance might work.

what do you think?

 

[amberwolf]

The simplest way for a drastic (but completely undoable) current cut is to remove one of the shunts, if there are multiples. If there are two, it cuts current in half. If three, by a third. Etc.

A non-undoable current reduction is easy by filing the shunt--you only need to file a notch in it, rather than along it's whole length. The proportion of shunt diameter removed vs that left is the ratio of the current change. Best is to file just a tiny bit (one file stroke) then retest, and repeat until it behaves as desired. PITA but since it's not undoable it's the simplest non-drastic relatively-controllable modification method.

(you can solder over the filed slot to refill it but the resistance of the solder is different than that of the shunt so it isn't really an "undo").

To lengthen a shunt wire, you need to replace it with a new longer shunt wire--it is not regular wire, it is a manganin alloy.

 

[AHicks]

Check out the KT controllers and displays. They can be set to use a defined by you amount power to supply the amount of assist you need, and speed is removed from the picture completely. I generally set PAS 1 for something in the neighborhood of 75-150 watts, which is good for 5-10 mph with excellent low speed control.

MUCH MUCH better low speed control, say 5-12mph than that available from speed based controllers.

 

[jasonahf]

So from what I'm reading... the behavior is typical of a typical cheap Chinese controller.

As for an update, I decided to upgrade the BMS on the 13s5p battery to 30 amp vs the 20amp that originally came with the 13s3p battery.

Problem is... still have the issue with bike cutting power to the motor at peak power/acceleration. The seller is now sending a new controller to see if that resolves the issue.
According to their site the 1000w controller is rated for 25 amps with max of 50 amp.
considering the motor is drawing 1300w peak I suggested they provide a 1500w controller instead of the 1000w.

 

[jasonahf]

update.

cutting out issue was the controller.

new controller fixed the problem