Powerfull Emoto + Rain = nasty crash

trazor

100 W
Joined
Jan 13, 2015
Messages
171
I was riding my OpenEMoto build on the field and a strong rain started. My bike and myself were very wet, and at some point the bike stopped. I've a kelly controller and the red led signalled 3-3 error code (later, safe at home I learnt that the code means "Throttle short circuit or disconnected"). I had reset the controller and continued ride to home and then this happened:

[youtube]eYrMMuwD9jI[/youtube] (360 video, so pan around)

I felt that the bike went full power without my input. I never thought that water and hall sensors will be a risk. I think that probably some dirt and water made the 5v line to connect to the signal line. This is a very dangerous condition!
I don't have a brake sensor, but will install one soon.

This bike is using the same motor/battery than my old build, it is very powerful (check here https://www.youtube.com/watch?v=-HDfVYKeUqQ or https://www.youtube.com/watch?v=7YO6QAL7AyU&t=56s )
This is not a rookie throttle mistake, I've been riding this for long.

Now I've an injuried knee, my 360 camera is cracked, and my Marzocchi 380 is broken in a very strange way: the left bottle is bent, so the stanchion cant slide. No cracks or signal of hits in the bottle.
I'm angry and sad.

Something like this happened before? Do you have any ideas on how to prevent this?
 
Yep, many of us riding powerful bikes had a throttle short experience, with various damages.

It can be prevented. First with proper connector waterproofing and wire routing, that should be applicable to all of us. Then, many controllers, and the CA, can be set for throttle response curve and voltage range. Since a short does return the full voltage of the controls circuit (generally 5v), setting the max throttle response voltage to a lower value (eg. 4.6v), will prevent that a short can be read by the controller program as full throttle signal.
 
Sorry to hear about your crash, it could have been a lot worse as you could have ended up under the following car. As above, unfortunately, you're not alone in experiencing this type of failure. Depending on throttle type I've also read posts (John CR, I think) about adding a resistor to clip max throttle value.
Most cheap throttles are not IP rated, you have a few options to prevent similar fail mode in future. Regarding the fork perhaps you were really unlucky and the sliding bike hit the curb/ gutter on the opposite side.
 
Water in the throttle can short from wire to wire, bypassing the hall switch.

When caught in rain, I have a half twist throttle, which is easy to cover with half my hand. I run the twist with just my pinky finger, and cover that crack between grip and throttle body with my hand keeping it dry inside.

To park in rain, cover it with a small baggie that you should always carry if likely to need it.
 
dogman dan said:
Water in the throttle can short from wire to wire, bypassing the hall switch.

When caught in rain, I have a half twist throttle, which is easy to cover with half my hand. I run the twist with just my pinky finger, and cover that crack between grip and throttle body with my hand keeping it dry inside.

To park in rain, cover it with a small baggie that you should always carry if likely to need it.

A judicious application of some silicone goop around where the wires attach to the hall sensor would go a long way to preventing most moisture induced shorts at the handlebar end.
 
Any newly installed kit I place involves opening the E-throttle and judiciously applying a waterproofing spray to the innards. Many of us have been in the same boat as you. It sucks not being in control. :?
 
The painted white lines in the road are more slippery than the asphalt when wet. I've often experienced the tire spin up going over them on bikes that had no throttle malfunction. Ebikes spin up the wheel really fast compared to ICE bikes of similar power levels.

That said, it's also possible water can trigger going WOT at anytime, be glad you weren't at an intersection with heavy cross traffic, any wreck you can tell about afterwards was a kind one.

-Luke
 
liveforphysics said:
The painted white lines in the road are more slippery than the asphalt when wet. I've often experienced the tire spin up going over them on bikes that had no throttle malfunction. Ebikes spin up the wheel really fast compared to ICE bikes of similar power levels.

That said, it's also possible water can trigger going WOT at anytime, be glad you weren't at an intersection with heavy cross traffic, any wreck you can tell about afterwards was a kind one.

-Luke

I agree with you about the slippery road paint. That can cause a highside or more probably a lowside fall. This wasn't the case. The bike throwed me in a extreme wheely.

I wonder how the "big brands" prevent this? Are the potentiometer kind of throttle more reliable?
 
Thanks all for your ideas.
My knee pain is getting better.
I've fixed my Fork in the way a surgeon cures a really bad cancer. I've removed an important part of the fork. One of the bushings had to go. Now it slides ok, but without a bushing it is going to degrade soon. That is better than no fork.

For now, I'm not going to use the bike in rain, and will design and install a brake cut on the rear.

Thanks!
 
To my knowledge the CA has a throttle short protection by detecting the short as a fault ( ie. 5V signal as mentioned above).

Can anyone comment of this being a cure for such a problem?

I ride on cliff edge sometimes and this type of a failure could be lethal.... :shock:
 
12-C said:
To my knowledge the CA has a throttle short protection by detecting the short as a fault ( ie. 5V signal as mentioned above).

Can anyone comment of this being a cure for such a problem?

I ride on cliff edge sometimes and this type of a failure could be lethal.... :shock:

Controllers throttle range usually go from 1 to 4 volts. Detecting 5 volts is obviosly an error condition. Minutes previous to my crash, I had the Kelly controller signaling it one time. Second time I had not same luck. I don't think this error detection method is reliable. If I understand ohms law correctly, if there is a short (or low resistance) between ground and power cable, the voltage can drop to controller accepted range. Signal, ground and power cable are usually milimeters apart.

Can somebody with real knowledge about electricity behaviour confirm?
 
12-C said:
To my knowledge the CA has a throttle short protection by detecting the short as a fault ( ie. 5V signal as mentioned above).

Can anyone comment of this being a cure for such a problem?

I ride on cliff edge sometimes and this type of a failure could be lethal.... :shock:
Yes it is a safety, and it does exist in the programing parameters of some controllers too. But it is not 100% reliable, as Trazor did mention. Waterproofing should be part of periodic maintenance.

I should mention 2 other situations that I have experienced. A very unlucky throttle short direct on the controller board, that was caused by salted water that had got inside, and dried causing a short. A mechanical damage after the parked bike fell on the right side, that had caused the throttle to stick in full position.

Today I am careful about that, switching my bike power on while I am still aside of it. Everytime an unvolontary full throttle occurs, is a bad experience even when suffering no damages.
 
Sorry you got hurt. I have been hurt on my ebike,.... not good.

What happened to you, as others have said was water in your throttle, or the wiring from your throttle.

On my bike I usually get about .8v when the throttle is in off position and about 4.8v when I am WOT....depending. I say "about" because it depends on the controller and throttle combination. The quality of the Hall sensor in your throttle is the key to the return voltage from your controller. Most e-bike throttles are cheaply made and you, unfortunately, were the unwitting victim of water in your wiring system.

(You did get great video of it though! 8) )

It was most probably water in your throttle and it shorted your Hall sensor, although it might also have been water in you wiring harness where your throttle connects to your controller. The short most probably fed 5v throttle voltage to your controller and the bike went WOT because the controller thought you were WOT. Did it go faster then it ever did before? If it did that is because of the full 5v return which we useally don't get from the Hall sensors in our throttles. They tend to max out under 5v.

Nothing worse the a ghost rider controlling your e-bike. :evil:

I have had it before on my bikes when my throttles have broken. I use a throttle with a kill switch on them for that reason. :wink:

The best thing to do is not ride in a rain. However if you need to ride in the rain then, as other have stated, water-proof your system.

As Dogman suggested, simply putting a plastic sandwich bag over your throttle and taping or rubber banding it closed around the handlebars will give you some protection. Perhaps enough to get home anyway. But you may have also gotten water in the connector between your throttle and controller.

If you live in a place where it rains, do your self a favor and research everything on waterproofing an e-bike. It will save you some pain. And that is something we can all appreciate!

:D
 
Another thing that can happen to cause unexpected WOT events is a broken ground on the throttle (or grounding problem inside the connector between throttle and controller).
trazor said:
I wonder how the "big brands" prevent this?
By having the controller fault and shutdown when such a problem is detected.

Yours already does this, but apparently not in a sufficient way.

As you note, however, it's possible for the "short" to be insufficient to be full supply voltage, but still be enough to be full throttle voltage. Not much the controller can do to differentiate that in a single-sensor throttle.

However, you could use a dual-sensor throttle, where one sensor and magnet is run clockwise, and the other goes counterclockwise, but both magnets fixed to the moving part of the throttle, so that both of them have separate signals that change at the same time. One starts at higher voltage and goes toward low, and the toher is "normal" and starts at low voltage and goes toward higher voltage.

It's not likely taht *both* of them would have the same fault at teh same time, so a controller that reads them will detect eitehr of them faulitng compared to the other, and shutdown.

YOu can build such a throttle by cutting two throttles up, and merging their mechanics into one with the magnet/sensor seciton of one of them flipped over vs the other. (or custom-machine one and use the bits from two inside it).

You could build a circuit that reads such a throttle, by inverting the voltage from the "second" sensor, then comparing the voltage between teh two. The comparator output (activated anytime there is a difference between the two throttles) would be the "error" signal output, used to drive the "disable" or "shutdown" control line of the controller. Then the voltage from the "normal" throttle is passed to the controller itself as the throttle signal. (or the inverted voltage from the other one).


If this unit is inside the controller case, then it's not possible for a wiring fault between the unit and the controller to cause a failure, as long as teh controller itself is waterproofed or itself inside a waterproof compartment.

However, it's still possible for stuff to get inside the controller, then it could still cause the same problem as at the throttle itself--it's just less likely to, since throttles are more exposed and vulnerable than controllers, typically.

Waterproofing the throttle might be another way, but it's VERY hard to *completely* waterproof stuff.

liveforphysics said:
The painted white lines in the road are more slippery than the asphalt when wet. I've often experienced the tire spin up going over them on bikes that had no throttle malfunction. Ebikes spin up the wheel really fast compared to ICE bikes of similar power levels.
I think that's half of what caused my ankle-breaking crash at the Death Race (though it was a crack-repair-strip on the asphalt rather than marker strip, and it was dry). Never know for sure, though.


12-C said:
To my knowledge the CA has a throttle short protection by detecting the short as a fault ( ie. 5V signal as mentioned above).

Can anyone comment of this being a cure for such a problem?
Sure--teh controller the OP has even did this (faulted and shutdown). But it's not perfect, and only works for over-limit voltages. Doesn't detect an undesired WOT taht is within the normal range.
 
When I started going out on rainy days, looking for crosswalk stripes to slide better on the motorcycle, I knew it was time to stop riding powerful bikes for about 30 years. :roll: Finally I'm old enough to trust with a motorcycle again.

I took up something safer then in the 80's, extreme skiing. :lol:

Powerful e bikes are such fun, but in the tire department, they really don't cut it. So easy to lay down e bikes when it gets slick, compared to a moto tire.

Part of why I got that scooter, now my need for speed is satisfied with adequate tires and brakes.
 
I know this thread is kinda old, but anyhow, if it can help future readers...

I had this very same issue twice with regular throttles. The first time it just died during heavy rain and I had to wait for it to dry. The second time it went full blast when I was sitting at a red light (luckily my scooter was still quite underpowered at that time).

So I decided I never wanted to let this issue ever happen again and bought a cable throttle, just like that:
https://item.taobao.com/item.htm?spm=a230r.1.14.225.7dee44e0J8RLLb&id=560097662979&ns=1&abbucket=6#detail

With those things, you have a somewhat regular motorbike handle mounted on needle bearings, linked via a metal cable to the potentiometer box. It allows you to put this potentiometer somewhere inside your bike where it will never see water, plus it also gives you the feeling of a regular motorbike throttle, which is really quite nice.
In addition I used waterproof connectors on all the throttle harness + tape + programmed my controller to 4.5 V max.

Never had a problem with rain ever again, even running under hardcore storms.
 
Production EV's using twist throttles generally have an upward sweeping hall effect with downward sweeping hall effect, and a micro-switch for fully closed throttle. Takes 5 wires vs 3 wires, and requires a controller capable of taking that type of input, but it's worth it for high powered stuff IMHO. If the signal sweeps don't correlate, the controller faults vs accepting the throttle command, and if it ever gets the micro-switch signal it stops throttle command even if the halls do correlate. The halls VS potentiometers is also nice, as you can have a fully potted assembly with halls as only the magnet is a moving part.

Also, a number of OEM production gas bikes are drive by wire now entirely, BMW and Harley and Yamaha and Triumph and more have non-cable type electrical signal only throttles on various bikes today.
 
Great finding @Dui, ni shuo de dui, thanks. As @Antenor" said, It will be great if the product is available in alibaba or any other western store. Will keep looking.

The system described by @liveforphysics sounds very good too, altough complicated and probably not available as a DIYer product.

After the crash I decided that having a throttle brake cut is a must. I had some crashes after landing drops and this rain episode that can be prevented with a quick brake lever action. The problem is that no good hydraulic brakes provide a switch (I think tektro have some for OEM only).

So, I designed a modular brake switch that can be used on Shimano Deore M615 and Shimano Zee M640, two great hydraulic brakes.
As always, the sources are the OpenEmoto project: https://github.com/eroldan/OpenEMoto/tree/master/BrakeCut

After three months, the system is working flawless.
The point of action is very adjustable and I've mine to trigger at the beginning of pad contact feel.
I use it in the left only because:
- Muscle memory from motorcycle clutch
- Free hand from throttle
- In panic case, blocking rear is less risky than front(I use moto style brakes: right=front)

I feel much safer now.

IMG_20180209_135103.jpg
IMG_20180209_135117.jpg
IMG_20180209_135445.jpg
 
nice ebrake solution..
Magura also makes hydraulic ebrakes with a cutoff switch..
http://www.magura.com/en/components/bike/speedemobilityseries/productdetailpage/?p=2594
 
that you were riding such powerful ebike with no brake switches was crazy.
you dont have to have regen but such cut off switch is a must.
 
miro13car said:
that you were riding such powerful ebike with no brake switches was crazy.
you dont have to have regen but such cut off switch is a must.

brake cutoff switches are a good safety feature but many ppl disable them for aggressive riding such as wheelies, burnouts, dragging the rear brake slightly for increased traction off-road, etc..
 
efMX Trials Electric Freeride said:
nice ebrake solution..
Magura also makes hydraulic ebrakes with a cutoff switch..
http://www.magura.com/en/components/bike/speedemobilityseries/productdetailpage/?p=2594

FYI
I was about to buy one of those but found this instead: https://www.amazon.com/gp/product/B015DJWWM6/ref=oh_aui_detailpage_o01_s00?ie=UTF8&psc=1

i've them on my desk but not installed/tested yet.
 
Back
Top