Help troubleshooting Negative Hwatts from Erider BB CA3

[merdi55]

I've found a couple posts with vaguely similar issues, but nothing specifically describing what I'm seeing. If someone can find a discussion of these symptoms somewhere, please redirect me. I wanted to try to see if I could figure this out without having to contact Grin directly, they've already been so much help, I will reach out to them pronto if no one here has ideas for me.

I'm working on a new build from Grin tech. I had a simple cheap motor setup about 10 years ago, but I'm essentially new to all of this.

Gmac 10t
CA3 WP
Baserunner L10
E rider Torque Sensing BB
UPP 52V 28Ah Li ion
tripwire installed and regen braking enabled.

I have read the manuals, zeroed the amps, and double checked all the PAS setup and PAS config screens that grin tech set before shipping. Everything looked correct to me there.

Human watts seems to be largely accurate until motor runs. It reads between 160-230 when I'm giving it some mustard.

If I turn the PAS on, then as the motor starts to engage, then the Human Watts reading goes haywire, jumping to 300, down to 160, flipping negative, and back, as the assistance tries to adjust to garbage input. On the diagnostics page, the W limit flag comes on as soon as I start pedaling, before I can even feel the motor kick in.

I get a similar effect if I use the thumb throttle, although the Human watts will flip as far as -800, back to positive, all over the place, but as long as the motor is in use, it tends towards negative territory, such that over her maiden 30 mile journey, CA showed a negative total and negative average HWatts.

However, on diagnostic page the W limiter does not activate when I use the throttle whether I'm pedaling or not (even when PAS is still turned on).

If I ride throttle without pedaling, the human watts reads 0.

Thanks for the help!
Ben

Info from setup screens to facilitate assistance:

Throttle mode: Thru

Power limits:
Plim 9900W
Wgain 15.0 (I could not find documentation on this number?)

PAS Setup:
homepage with no torque on pedals: V1.51 1.1 n/m responsive to stepping on pedals
Custom Trq 18 poles, 1 wire, 5V = FWD (Left pedal indicator activates for forward pedaling)
Trq scale = +70.0 Nm/v
Zero trq 1.49 (1.51) [changes appropriately when I step on pedal]

PAS Config (some of these I altered slightly before I noticed the negative human watts from the 30 mile ride, I just gave up on PAS for the ride and used throttle. I thought maybe PAS was just not being responsive, so my changes aren't the only problem if they are part of the problem at all)
PAS Mode: Torque
Start level: 15 Hwatts
Scale 5.0 W/HW (with digital PAS Level adjuster, I get same issue at all PAS scales)
Assist Avg: 6 poles (was 9 before I changed it, negative HW at both 6 and 9)
Start Thresh .6s/pole
stop thresh: .45s/pole

 

[amberwolf]

AFAICR, the Wgain is just like the Again only for watts. It's probably documented in Teklektik's UUG, linked in the CA3 beta thread and the CA3 product info page on grin tech's site.

However, it shouldn't be affecting the Hwatts reading.


I don't know that this is the case here, but in general random / spurious readings on things tend to be caused by faulty wiring, very often bad grounds, or ground loops, causing a difference in ground levels from one place to another when higher currents are flowing, so they look fine when nothing is happening, but when the system is in use under load the faulty ground "lifts" and causes incorrect readings from the device(s) it's for.

Almost all electrical-system problems tend to be caused by connections and wiring, so I generally recommend starting there.

If your system is the newer HIGO-style wiring, it's less likely than with the older JST stuff, but it's still possible. A bit harder to troubleshoot, though, as it doesn't leave you any access points to check the wiring. But you can try disconnecting and reconnecting everything, at all connectors, just to ensure they are all fully seated.

If yours is the older JST style connectors, there's a number of possible problems to look for, including wires broken at the back of a contact, a contact backing out of a connector shell when plugged in (but appearing normal when unplugged), a bad crimp, etc.



Another possibility is induced currents. If the PAS wiring happens to run parallel to either battery or motor wiring, there is a small chance of current spikes in the battery/motor wires inducing signals into the PAS wiring.

 

[merdi55]

Thanks! Those are great thoughts.

I had already looked at Teklektik's guide, I found mention of Wgain, but didn't make sense of it, but I agree, I don't think it's the issue. The main thing I remember is him saying that normally it isn't used for anything anyways, and I think I have a pretty normal setup.

No avail.

I unplugged/replugged, and no wires seem arranged in a way that could produce induction. I will take voltmeter to everything I can imagine tomorrow and dig into every connector I have access to. I think they are mostly newer Higo connectors... IDK, it's the newest CA.

I'll see if I can learn anything. I had high hopes about that induction idea, since it seemed like the only issue took place when there was signal from the PAS and juice from battery to motor. However, another ride after disconnecting/reconnecting seemed like PAS data may not be good even when no power to motor.
Thanks again.

 

[amberwolf]

However, another ride after disconnecting/reconnecting seemed like PAS data may not be good even when no power to motor.

That's actually a good sign, because if you had good PAS data before disconnect/reconnect, it may indicate the connector itself or the wire immediately around it is the source of the problem.

If you have a second set of eyes/hands available, you could try operating the bike off-ground while engaging the brake on the pedal-driven wheel slightly (just enough to make some load on it to create an Hwatts reading), and gently move around, pull on, twist, etc, the PAS/sensor plug(s). Those are the main ones that should make a difference to what you're seeing.

 

[merdi55]

thanks so much for replying Amberwolf.

I haven't had a chance to do any real diagnostic work yet, adventure-times press. But I had more symptoms pointing to faulty wiring somewhere (it seems to me).

I got caught in a small rain storm before I was done fishing, so everything got wet, and for the first 7 miles or so riding home, the torque sensor seemed to be giving accurate readings, then reverted back to negatives when the throttle was engaged.

So, soon, I'll do the work and see what I can figure.

 

[momus3]

These are the type of electrical gremlins that can drive you a little batty. Patience will be your friend, and sometimes it's a good idea to take pics and notes. Like amberwolf said, many times it's simply a bad connection somewhere. Always ck the most easy to get to things first, as well as the most obvious and simplest ones.

I remember one time we had a Toyota under warranty that would blink the right rear blinker a teeny bit when you put the left blinker on. Everything checked out, and the factory rep finally said to put a complete wiring harness in it, which is a big deal. Every single wire in the car had to be replaced from the headlights to the taillights. Under the headliner inside, on the engine, in the dash, just everywhere.

So we did that and no dice, the thing acted exactly like it did before we replaced the wiring harness. One of the other mechanic's helpers took the rear taillight bulb out just to look at it, and there was a very small amount of solder going from one contact to another. You could barely see it, it was thinner than a hair. We put a new taillight bulb in it and everything worked fine.

 

[merdi55]

Similar to Amberwolf's suggestion of possible induction from a parallel wire, Grin Tech asked me to unzip the Erider cable from the frame (steel), and see if that made a difference.

It appears induction from the motor or battery or controller/CA cables caused a current in the frame, which then induced garbage readings through the erider cable.

I added a bit of rubber foam insulation between the erider cable and the steel frame, and the problem was solved.

I bet if I had been more thorough in completely reworking the wiring like Amberwolf had suggested, I may have discovered the problem myself.

I hope this helps someone in the future. Careful with those steel frames.

 

[bikeboy999]

I am still trying to figure out how to use my Erider bottom bracket torque sensor. I have an aluminum frame, I have not clipped the wire to the frame. I am running it up with cables from controller and throttle to the CA 3.14. Where the cables are all wrapped in a velcro sleeve just below the CA unit. I am so new to the ebike, that today was the first ride. The throttle worked, but the numbers when I was trying to use the torque sensing to set the assist. Will try with the above numbers tomorrow.

Aluminum frame and wire interference induction?


Brian

 

[bikeboy999]

Setup is as follows.

GMAC 10 with LT end.
Phaserunner LT with MT60 to anderson connector
8X Ligo batteries
Throttle with extension cable running around the trike instead of coiling/wrapping with zip ties
CA3.14 with built in multi function switch
E-rider bb 68 5 with extension (wrapped in velcro with other cables just under the unit) to complete the missing 2 inches

The cable from the Erider is plastic wrapped around the front derailleur then makes 90 turn where it joins the CA cable from the Controller and the Throttle cable.

Brian

 

[amberwolf]

I am running it up with cables from controller and throttle to the CA 3.14. Where the cables are all wrapped in a velcro sleeve just below the CA unit.

Then as the topic itself resolved, separate the cable that is from the part that isnt' working as expected from all the other cables.

If this ifxes it, then you have interference, and can shield against it or space it away from other cables as necessary, just as the OP did in his last post above here.

If it does not, then you have a different problem that we can troubleshoot.

The throttle worked, but the numbers when I was trying to use the torque sensing to set the assist.

Could you please complete the above sentence, so that we know what you intended to say? As it stands, it doesn't convey useful information about the torque sensing/assist.

Will try with the above numbers tomorrow.

What "above numbers"?

8X Ligo batteries

In what configuration? Presumably not all in series; possibly all in parallel, or some other configuration?

Throttle with extension cable running around the trike instead of coiling/wrapping with zip ties

What exactly does "running around the trike" mean?

 

[bikeboy999]

It works, apparently running the software to the Phaserunner was something I was supposed to do first. Now all I need to do is figure out how to set it for a spinner cyclist not a masher.

Brian
PS was just about to see if I could delete the above posts, the heat must be getting to me.

 

[bikeboy999]

Running around the trike,a I meant by using the full length of cable instead of wrapping it and tucking it some where.

BTW, thanks.

 

[amberwolf]

Running around the trike,a I meant by using the full length of cable instead of wrapping it and tucking it some where.

That still doesn't describe how it's used or routed, whcih is the important part of the answer, in reference to your posted problem.

Somethign I would highly recommend for future troubleshooting posts is to always give as much detail as you can, as precisely as you can describe it. Often general information with vague or nonexistent descriptions is insufficient to help someone fix things.