Cycle Analyst ebrake/PAS issue

Get all your technical information about electric bikes here.
Post Reply
20-20
10 µW
10 µW
Posts: 6
Joined: Dec 16, 2017 7:35 pm

Cycle Analyst ebrake/PAS issue

Post by 20-20 » Dec 16, 2017 8:23 pm

Motor doesn't power up after PAS gear shifts when using gear sensor or tapping ebrake lever.

When I hit the brakes, motor power is switched off but doesn't reappear for approx 10-20 seconds after releasing the brake. But it can be an intermittent problem - sometimes motor power is reconnected immediately after releasing the brake lever but very infrequently (maybe 1 in 10 at best). I've also had a Risunmotor gear sensor wired into the EBRAKE input of CA3 with similar results i.e. it always switches off motor power on each gear change, but in most instances fails to switch it back on again following the shift.

Once the motor fails to restart after each gear shift (or tap of the EBK), neither PAS or throttle will kick back in for approx 10-20 secs.

I've tried this with the gear sensor and EBK wired in together, and with both wired in separately. I get the same results with each combination.

If I shift gear by back-pedaling to cut the motor, the motor powers on and off as it should. Looks like this problem is specific to the CA's EBRAKE connection.

TIA for any advice.

Cyclone 3000w, 52v, 20Ah battery, CA v3.03, 40 amp controller, torque PAS.

User avatar
amberwolf
100 GW
100 GW
Posts: 24231
Joined: Aug 17, 2009 6:43 am
Location: Phoenix, AZ, USA, Earth, Sol, Local Bubble, Orion Arm, Milky Way, Local Group
Contact:

Re: Cycle Analyst ebrake/PAS issue

Post by amberwolf » Dec 17, 2017 1:51 am

First, what do you do to "reset" the power when it won't come back on?


Seen here at the lower left corner of the CA screen is an icon you can use to troubleshoot:
http://www.ebikes.ca/product-info/cycle ... ay-screens
A throttle slider icon shows visually where your input throttle is at, switching to a flashing mode if throttle auto-cruise is latched, and then to a brake lever if ebrakes are engaged.
So when you let go of the brake (or the gearsensor disengages), does the ebrake symbol also go away at the same time?

Or is it delayed like the motor power?

If it's also delayed, perhaps something is wrong at the wiring itself, and sometimes the ebrake signal is staying engaged due to a connector / wiring issue.

Alternately, if the ebrake symbol goes away as it should, what does the CA's "limiter" screen show when power is working vs when it is not?

(capital letters indicate a limit engaged)


There's also some time delays you can change within the CA settings (ramping times) but I expect that if it works normally sometimes, these are already set correctly.

User avatar
teklektik
1.21 GW
1.21 GW
Posts: 3566
Joined: Mar 26, 2011 1:15 pm
Location: CT, USA

Re: Cycle Analyst ebrake/PAS issue

Post by teklektik » Dec 17, 2017 11:32 am

20-20 wrote:
Dec 16, 2017 8:23 pm
Motor doesn't power up after PAS gear shifts when using gear sensor or tapping ebrake lever.

When I hit the brakes, motor power is switched off but doesn't reappear for approx 10-20 seconds after releasing the brake. But it can be an intermittent problem...

If I shift gear by back-pedaling to cut the motor, the motor powers on and off as it should. Looks like this problem is specific to the CA's EBRAKE connection.
To set the stage:
  • What version of firmware are you using?
  • Are the ebrakes wired to the controller as well as the CA?
  • If you are using 3.1b20 - what is your EBk->MinTime setting?
The Main Screen ebrake glyph is pretty much a direct indication of the ebrake input AND the EBk->MinTime setting. I would start by setting MinTime to 0 so the glyph should track the ebrake input alone. Another alternative is to watch the ebrake glyph on the "SETUP EBRAKE" preview screen. That one is the EBk input pure and simple. Although this is a better indication of the ebrake state, it's in Setup and so you can't detect any resultant motor inactivity since the motor never runs in Setup.

After clearing MinTime, test to see if the ebrake glyph tracks the ebrakes or the power loss. Also check and see if the Thr OUT field on the diagnostic screen tracks the power loss. The results will help localize the issue.

Also, if clearing MinTime makes the problem go away, then we may have a firmware bug there...
Visit Grin Technologies at www.ebikes.ca
Build Thread: 2WD Yuba Mundo V4

20-20
10 µW
10 µW
Posts: 6
Joined: Dec 16, 2017 7:35 pm

Re: Cycle Analyst ebrake/PAS issue

Post by 20-20 » Dec 17, 2017 2:41 pm

teklektik wrote:
Dec 17, 2017 11:32 am

To set the stage:
  • What version of firmware are you using?
  • Are the ebrakes wired to the controller as well as the CA?
  • If you are using 3.1b20 - what is your EBk->MinTime setting?
I'm on version 3.03, so no "SETUP EBRAKE" option.
The brakes are wired to the CA only.

The ebrake glyph responds with the brake lever i.e. it's only displayed when the lever is activated. It's not on view when the motor power is still off several seconds after the lever has been released.

When I had the gear sensor connected (also directly into the CA's EBRAKE input, not the controller), the ebrake glyph didn't display at all when shifting. Is that expected behaviour?

I've not tried using the diagnostics screen. I'll do that tomorrow and report back with any flags.

A big thank you to amberwolf and teklektik for your help and advice.

User avatar
teklektik
1.21 GW
1.21 GW
Posts: 3566
Joined: Mar 26, 2011 1:15 pm
Location: CT, USA

Re: Cycle Analyst ebrake/PAS issue

Post by teklektik » Dec 17, 2017 3:24 pm

20-20 wrote:
Dec 17, 2017 2:41 pm
The ebrake glyph responds with the brake lever i.e. it's only displayed when the lever is activated. It's not on view when the motor power is still off several seconds after the lever has been released.
So from this we know the ebrake signal is good up to the CA and the CA is interpreting it correctly.
20-20 wrote:
Dec 17, 2017 2:41 pm
When I had the gear sensor connected (also directly into the CA's EBRAKE input, not the controller), the ebrake glyph didn't display at all when shifting. Is that expected behaviour?
No. The gear shift detector should affect the CA the same as applying the ebrakes. The plan here should be to leave that sensor disconnected for these tests to reduce the number of variables - let's just focus on misbehavior due to ebrakes...
20-20 wrote:
Dec 17, 2017 2:41 pm
I've not tried using the diagnostics screen. I'll do that tomorrow and report back with any flags
Good. This will clear things up.
If the OUT field tracks the ebrakes but the motor stays OFF anyway, then the controller is unhappy and ignoring throttle control from the CA. If the OUT field tracks the motor power, then the CA is actually suppressing the controller throttle and the CA configuration will need review.
Visit Grin Technologies at www.ebikes.ca
Build Thread: 2WD Yuba Mundo V4

20-20
10 µW
10 µW
Posts: 6
Joined: Dec 16, 2017 7:35 pm

Re: Cycle Analyst ebrake/PAS issue

Post by 20-20 » Dec 17, 2017 8:03 pm

amberwolf wrote:
Dec 17, 2017 1:51 am
First, what do you do to "reset" the power when it won't come back on?
I start pedaling and playing with the throttle. Initially nothing happens - it's like I've lost all power. Eventually the motor kicks back in again, with both throttle and PAS responding as they should.

20-20
10 µW
10 µW
Posts: 6
Joined: Dec 16, 2017 7:35 pm

Re: Cycle Analyst ebrake/PAS issue

Post by 20-20 » Dec 18, 2017 9:01 am

I may have solved it...

Testing while monitoring the diagnostics screen: when I blip the throttle lever to kill the motor momentarily for the gear change, I can see the voltage OUT (top right of screen) drop then immediately climb back up again, but the motor doesn't restart for several seconds, regardless of voltage OUT. None of the flags went upper case.

My fix was to raise ThrO>Brake Out from zero to 0.90 volts. That's below my ThrO>Min In of 1.17 volts and no signs of motor creep. Not sure if technically that's the right approach but it's working - a quick tap of the ebrake results in a momentary loss of motor power. Exactly what I was hoping for :D

If that's an acceptable fix, I'll wire the gear sensor back in and hopefully that will work as intended too.

While riding the bike with the diagnostics screen displayed, I noticed a couple of flags on view, but not in connection with the ebrake issue.
  • 'A' was displayed much of the time but I guess that's standard when riding in Torque PAS mode (3-postion switch)?
  • 'W' appeared whenever I hit the throttle

User avatar
teklektik
1.21 GW
1.21 GW
Posts: 3566
Joined: Mar 26, 2011 1:15 pm
Location: CT, USA

Re: Cycle Analyst ebrake/PAS issue

Post by teklektik » Dec 18, 2017 10:35 am

20-20 wrote:
Dec 18, 2017 9:01 am
Testing while monitoring the diagnostics screen: when I blip the throttle lever to kill the motor momentarily for the gear change, I can see the voltage OUT (top right of screen) drop then immediately climb back up again, but the motor doesn't restart for several seconds, regardless of voltage OUT. None of the flags went upper case.
Ahha! Good sleuthing!
20-20 wrote:
Dec 18, 2017 9:01 am
My fix was to raise ThrO>Brake Out from zero to 0.90 volts. That's below my ThrO>Min In of 1.17 volts and no signs of motor creep. Not sure if technically that's the right approach but it's working - a quick tap of the ebrake results in a momentary loss of motor power. Exactly what I was hoping for :D

If that's an acceptable fix,...
That sounds perfect.
We don't know anything about your controller, but my guess is that it has low voltage throttle fault detection. Setting ThrO>Brake Out=0 is probably tripping the fault which kills the motor because the controller thinks your throttle has failed. It resets when things are back in range for a bit. Since the CA doesn't actually drive the controller ebrakes and merely turns off the throttle itself, any voltage ThrO->Min Out or lower is just fine. This setting is used with Grin controllers to control regen when brakes are applied (0V=maxRegen) and most controllers don't have a low voltage fault so ThrO>Brake Out=0 is a good default. But not so much for your controller where your adjustment seems necessary and appropriate... :D

20-20 wrote:
Dec 18, 2017 9:01 am
While riding the bike with the diagnostics screen displayed, I noticed a couple of flags on view, but not in connection with the ebrake issue.
  • 'A' was displayed much of the time but I guess that's standard when riding in Torque PAS mode (3-postion switch)?
  • 'W' appeared whenever I hit the throttle
I think you have those backwards (W=torque PAS and A=Current throttle).
This is normal and expected. Torque PAS works by multiplying Human Watts into Assist Watts and so uses the Power PID controller (the W). Current Throttle uses the Current PID controller (the A). In both cases these controllers work by assuming 100% power or 100% current and then 'limit' it down to the PAS or throttle level with the result that the W or A flags always show limiting.

Anyhow, looks like you solved this one. The CA has screens and indicators to help you figure out what's up, but it's a bit of a complicated beast and knowing what to examine is a learning experience...
Visit Grin Technologies at www.ebikes.ca
Build Thread: 2WD Yuba Mundo V4

20-20
10 µW
10 µW
Posts: 6
Joined: Dec 16, 2017 7:35 pm

Re: Cycle Analyst ebrake/PAS issue

Post by 20-20 » Dec 18, 2017 1:19 pm

I've added the gear sensor to the CA Ebrake input. Based on a short road test it's working great: brief motor cutoffs with silky smooth shifts!

There's still one odd thing though... on the bike stand (no load on rear wheel), at larger throttle openings I still get long motor cutouts when blipping the brake lever. Maybe there's something about the greater voltage difference between Throttle Out volts and ThrO Brake Out? Observing the diagnostics screen, everything's OK up to 2.5 volts Throttle Out. Above that, blipping the brake results in motor cutouts of several seconds.

Wondering if there's something in CA's settings I can tweak to achieve consistent motor cuts throughout the entire operating range?

Controller is a 40amp YuYang King (with bluetooth). All inputs go via CA. Pass Thru throttle and TorqPAS.

User avatar
teklektik
1.21 GW
1.21 GW
Posts: 3566
Joined: Mar 26, 2011 1:15 pm
Location: CT, USA

Re: Cycle Analyst ebrake/PAS issue

Post by teklektik » Dec 18, 2017 2:11 pm

Good news on the gear sensor. You might like to post up which unit you are using for your readers.

You may be breaking new ground with that controller - not sure what's going on with the 'on the stand' tests. There are ramping adjustments in the CA, but without understanding why the controller is behaving oddly, it's all just experimentation. I would look to the BT controller configuration for things to twiddle.

Since all is well on the road, I'd ignore the 'stand' tests. This is not dissimilar from running a bike on the stand with CA Current Throttle - the wheel speed just keeps increasing as the CA keeps cranking the controller throttle up in an attempt to to hit the current level defined by the operator throttle (but there's no load so no large current). This always freaks people out, but it's normal and all works well on the ground. :D
Visit Grin Technologies at www.ebikes.ca
Build Thread: 2WD Yuba Mundo V4

20-20
10 µW
10 µW
Posts: 6
Joined: Dec 16, 2017 7:35 pm

Re: Cycle Analyst ebrake/PAS issue

Post by 20-20 » Dec 18, 2017 3:33 pm

teklektik wrote:
Dec 18, 2017 2:11 pm
Good news on the gear sensor. You might like to post up which unit you are using for your readers.
RisunMotor variable gear sensor, model MS-SS-01.
Got mine for $21 USD via AliExpress: https://www.aliexpress.com/item/Our-Exc ... 0.0.IiaVj8

Looks very much like the one sold by gearsensor.com
Whether the internals are as good, only time will tell.

It's marketed as plug and play for Bafang mid-drives. I chopped off the Bafang Higo connector and fitted a 4 pin JST-SM connector to match up with CA's Ebrake plug. Wiring was as follows:
1. White 5v
2. Red GND
4. Black SIGNAL
teklektik wrote:
Dec 18, 2017 2:11 pm
You may be breaking new ground with that controller - not sure what's going on with the 'on the stand' tests. There are ramping adjustments in the CA, but without understanding why the controller is behaving oddly, it's all just experimentation. I would look to the BT controller configuration for things to twiddle.

Since all is well on the road, I'd ignore the 'stand' tests. This is not dissimilar from running a bike on the stand with CA Current Throttle - the wheel speed just keeps increasing as the CA keeps cranking the controller throttle up in an attempt to to hit the current level defined by the operator throttle (but there's no load so no large current). This always freaks people out, but it's normal and all works well on the ground. :D
Agreed - it's how it performs on the road that counts. If I discover any further tweaks that improve gearshift performance I'll report back.

Thanks for your help.

Post Reply