SOLVEDish: ERider Torque Sensor Phaserunner EMI Noise interference problem

Hi

I have an Erider Torque Sensor and a Phaserunner.

I have the problem, that when pedaling, on increasing pedal force the human watts are shown negative and the bike behaves weirdly. The eRider Torque sensor outputs a proper torque signal from 1.5 to 3ish volts, according to pressure on the pedal. Cadence however is a little erratic, drops to zero sometimes. fwd is set to 5v. Cadence is shown positive.

CA3 setting is torque sensor set to custom, 1-wire, 18 poles, 70Nm per Volt, 5V = fwd. Zero Torque routine done.

I contacted grin and Justin wrote to me, "we have seen some cases where it(erider) is actually susceptible to EMI noise of the motor controller and the output voltage drops to 0V for some periods of time."

When I disconnected the motor cable and the human watts then are shown positive only. I have then made sure that the two cables (sensor/motor) do not go in parallel. However they are within 3cm of each other for a moment in the lowest part of the bike. Apart from that the sensor wire goes along the downtube, while the PR and motorcable are on the top-tube. Having the two cables wired apart did not change the problem at all.

Note: I have a 72v Battery (84v full charge) and I use the voltage divider trick to show the correct voltage on the CA3. The CA3 and all other devices get 12V from a DCDC converter, which in theory could have isolated ground, but I bridged ground of course. The DC-DC converter is a Tracopower TEN 20-7212WIR ; 43V-160V to 12V.

Unfortunately Justin hasn't written back to me for about a week now, so I hope somebody here has an idea how to solve this interference problem.

SOLVED-ish...

GrinTech unfortunately didn't have anything helpful to offer for my problem. Here is what I did to solve this, somewhat.

1. I got a relatively pricey line filter chip from Vicor, QPI-3LZ. https://www.mouser.ch/ProductDetail/492-QPI-3LZ, added some caps before and after and put it after my DCDC Converter and before CA3. The important bit is, that the "shield" connector on the chip goes to the metall of my bike frame. My ground does not. Disconnecting the connection to the bike frame renders the device close to useless. This made the throttle-flickering of the CA3 go away completely and most of the negative human-watt values of the erider torque sensor too.

2. Added ferrite rings around cables. Before and after DC-DC, after filter, after Phaserunner on the cable that goes to the WP plug of the CA3, before the eRider torque sensor. This improved everything marginally. But did help a bit and was easy and cheap to implement. https://www.ebay.de/itm/5-x-Würth-EMI-Ferritkern-Klappferrit-Ferritfilter-Ferritring-Entstörfilter-6mm/142492731972?ssPageName=STRK%3AMEBIDX%3AIT&_trksid=p2057872.m2749.l2649 Würth Ferritering, EAN 4051687045293, 74271112 (on mouser they cost 10x times more)

3: Added a 10u and a 100n capacitor between 12V and Ground just before the eRider sensor. This removed almost all of the rest of the problem. Rarely its output still falls to 0V causing negative human watts, but it's mostly only shortly after startup of the bike and I so far have not noticed it while riding.

The final solution is not perfect, complicated and the filter chip is not cheap, but it works good enough. So, if you have EMI Problems, try 2 and 3 first, then 1. In my case 2 and 3 did not work on their own.

I added a couple of picture. I didn't do the prettiest work on this. To avoid the caps to rattle loose from their solder joints I glued them to the chip with Black PA glue "Buehnen C20874S". Then heatshrinked it and let it float relatively lose in the bike. Near the chip you can see one ferrite ring which I harvested from an old power supply. In the last pic you can see the ferrite rings which I bought in different places (white)

Hi Raphael303, I just wanted to thank you for your posts. I had the same issue and was able to fix it by adding some RF chokes on the GMAC/CA/ERider cables, but, most importantly, grounding the negative terminal of the battery to the frame. Since doing the latter, the issue has not come up for me again.

matmat said:

Hi Raphael303, I just wanted to thank you for your posts. I had the same issue and was able to fix it by adding some RF chokes on the GMAC/CA/ERider cables, but, most importantly, grounding the negative terminal of the battery to the frame. Since doing the latter, the issue has not come up for me again.

Click to expand...

How do it ground it! is there a video out there?

zinaambiorix32 said:

How do it ground it! is there a video out there?

Click to expand...

I guess you can wire any ground from any connection plug coming out from your controller to the frame