Cycle Analyst CA3 + TDCM torque sensor issues

[bendik.vignes@gmail.com]

Hi all!
I have experienced some strange behaviour using the Cycle Analyst V3 with a TDCM torque sensor. I hope someone can help!

Here are some general details:
-Three wheel cargo bike. Dual (front) motor drive: Golden Motor Blt800, chain drive)
- 2x Grinfineon 35A controllers
-external CA3 shunt
-cycle analyst CA3-DPS
-TDCM torque bottom bracket

Note: The bike behaves fine when operated with the thumb throttle

Some relevant cycle analyst settings:
-Max current: 55A
-Max power: 2200W
-A Gain: 60
-W Gain: 10
-Throttle in, Min input: 1.05V (0.20V above actual min throttle input of 0.85V)
-pas mode: TrqPas
-pas poles: 12
-start/stop thrsh: 7/11


There seem to be some strange behaviour related to the torque sensor settings:
-Whenever asststart is a positive number (including +0W) there is no throttle out voltage sent to the controllers, so no motor power (and limit flag for power is a capital W). When asststart is any negative number (including -0W) a throttle out voltage is sent to the controllers when pedaling.
- With asstfactr of more than 0.5 results in high throttle out voltage and the bike just flyes away. Asstfactr is now set to 0.15
-when trqscale is set to approx 50Nm/V, both live Nm readings and Hwt looks fine, but throttle out voltage is apparently unaffected of pedal torque or pedal rpm
-when trqscale is set very high (several hundreds Nm/V) there is some proportional throttle out voltage: 1.6V when very light pedaling, to approx 2V for high pedal torque and pedal rpm.
-now the CA is set to asstfactr = 0.15 and trqscale =650Nm/V and the bike behaves "almost ok". But still too much "start power", and too much assist power when pedaling light
-limit flag for power is constantly W during PAS operation (even if motor power is far from 2200W)

Could there be an issue with the power limit feedback function of the CA?

I hope someone can help

 

[robbie]

The settings on the CA are pretty strange with 650Nm/V, thats way way higher than it should be. Usually TDCM sensors will saturate at like 3.5V (or so) when an adult is standing up in the pedals, especially with smaller chainrings.

Out of curiousity, what is the mechanical configuration of the pedals? Is this a recumbent? The installation geometry of the TDCM is super important, and is based on a regular diamond style bicycle configuration. Pedal force down translates to chain tension, which is what the sensor actually measures against a spring internally. The sensor has a slotted bearing cavity that must be aligned parallel with the direction of the chain tension, so that pedal force is not measured, and chain tension is isolated.

Is your pedal force inline with the belt tension force? Did you over tension your belt?

What is the resting voltage of the sensor?

 

[bendik.vignes@gmail.com]

Hi and thanks for your reply!

I think the the torque sensor is installed ok and works ok. Torque sensor voltage at no pedal force is 2.48V and max pedal force is approx 3.78. With a torqscale setting of approx 30Nm/V (is it should be) gives expected live torque readings: from 0 to approx 70-80Nm

No, it is not a recumbent bike. It is a dutch cargo bike called Johnny Loco. The pedaling an force is similar to a normal bike.

 

[bendik.vignes@gmail.com]

I got an email from Grin Tech: the ysuggest a software update on my cycle analyst, so I will try that!
Anyway, thanks for your suggestions!