20KW 2WD having misfires (but can do a double burnout)

[methods]

Well - first the good.

I can do an insane 2 wheel burnout on this bike
I am talking about both the front and rear wheels coming out side to side!

*Frame: Electra Cruiser
*Controller(s): Two 100V 100A 18fet controllers - current limit at 150A (so.... errr... no current limit :x )
*Battery: 24S 3P 15Ah 20C Lipo with active cell level LVC protection. 89V under full load on both controllers.
*Front Motor: 9C 9x7 in a 26" rim with dirt tires
*Rear Motor: 9C 9x7 in a 26" rim with dirt tires
*Throttle: Standard hall throttle ( splitting a pot throttle throws off the tuning due to the parallel resistance - need to re-tune)
*Brakes: Only a front V-Brake, ziptied onto the fork (I am not joking)
*Peak Current: With 1 motor, 160A. With 2 motors each of them sees around 100A - more if I can get traction.

With only the rear hub it halls ass near the level of a 5305 (for short bursts - then of course it overheats)
With only the front motor it just does lame burnouts
With both motors I can leap off the line with no cogging noise what so ever.
I have to center my weight perfectly - too far back and the front tire comes out. Too far forward and the rear tire comes out. Right in the middle - double burnout!

The Problem:
When I have both motors hooked up the rear motor "misfires"
At any steady state speed - 2mph - 5mph - 10mph - 15mph the rear motor makes random loud noises - like it is misfireing
Under acceleration there is absolutely no trouble
With only the rear motor there is not trouble at all.
With only the rear motor I can pedal, hit the brakes, go up hill, down hill, never a misfire

With both controllers hooked up and one in the off position, the other one can only reach half throttle.
With both on I am SURE I am getting full power (there is no question here - it flipping smokes the tires off)

But I turn both on.... it cruses fine for a second and as soon as the back EMF equals the controller voltage it starts random misfiring
Only the rear, not the front.

Any ideas?

-methods

 

[karma]

lol love the front brake :wink:

with two controllers working off the line how much does the voltage drop on the pack?

 

[Arlo1]

Hve you tried to move the controlers away from each other? They could be sending rfi from one to the other!

 

[methods]

with two controllers working off the line how much does the voltage drop on the pack?

Depends on where I am at. With a fresh charge I can pull 100A on each controller and keep it above 92V
After 10Ah the pack will sag down to 88.2V under full load.
These are brand new 20C Lipo

I just went to the store to get milk and did 49mph uphill at 93A on each controller. I know "up hill" is a vague term but I was accelerating the entire time till I maxed out the KV of the motor(s).

Hve you tried to move the controlers away from each other? They could be sending rfi from one to the other!

That is a very good point - I will do that straight away!


Observation:

On this ride to the store (only my second ride) I noted that *any* time the back EMF matched the drive voltage the rear motor would violently misfire. If at any point I backed off the throttle a little, somewhere in that "backoff" the motor would set off. It does not necessarily jerk the motor - It just sounds like a motor that is stalled out. The controllers I am using act funny when they think the motor is stalled. First they back off and go "thunk thunk thunk" then they make a loud squealing noise (like they are trying to ping out and read the phases)

I think I am going to double-check the hall signals on that rear controller.
I had issues several months back when I had too much hall wire hanging round. The signal was getting weak (or acting as an antenna) and I was having issues
(though... at the moment I cant remember how those issues materialized)

I am putting my money on a crappy hall signal.
Either poor contacts, too long of wire, sagging +5V line, or some other nonsense.

Oh! I know - I should swap the controllers and see if the problem follows the controller or the motor!

-methods

 

[aaronski]

Thats a troubleshooting step most of us dont have the luxury of. I'm jealous. :wink:

 

[Arlo1]

Are your hall sensor wires sheilded? I am reading and realizing how much I love your build thats insane power! If I could have found that kind of equipment before I started mine I would have went with something like that for sure! I have seen some crazy stuff with car electronics and RFI. So cant wait to see you get it! Post Video please!

 

[ZapPat]

Wow, that thing sounds crazy! I want to boost mine up still a bit more, but I am not sure about even wanting to be able to do a *double* burnout on pavement!

2WD is excellent for offroad use, I hope you have access to some good trails. From what I saw on a couple pics you posted a while ago you look to have some nice places to ride that thing.

As for your problem, Methods, I am ready to bet that it's your excessively long common ground wire between controllers that's doing it. On your photos I can see the point where your main battery wire splits into two sets, and this point is much too far away from the two controllers. Remember that the total voltage drop along both those controller's non-common ground wires becomes a voltage offset to anything common to both controllers (worse with current and you are using a lot!)... like the throttle signal as you are using only one for both. Current spikes in either/both controllers end up as noise, so you might get erratic throttle behavior. I also tied both +5V throttle supplies together on my own setup, although I haven't though it through yet if this is a good move or not. It seems to work OK for me, but at about 50V/37A times two, not 100V/100A+ times two!

On thing that does nag me is the fact that they work well when not on BEMF equilibrium... odd. Did you make sure you disabled that "bar protect' feature in both controllers?

You can keep both controllers side by side without problems in my opinion, as the housings help to isolate each circuit. Even if they were bare PCB's side by side I wouldn't worry much, since they already each produce lots of their own crap and thus are designed to take it.

Do you think you could get your physicist sweety to photograph/film you doing a 2WD burnout?

Pat

 

[methods]

Are your hall sensor wires sheilded?

The wire I chose for the hall sensors does have a shield on it but I did not hook it up.
Perhaps I should go into the controller and do that.
I am thinking that I will do a single ended ground at the controller side. I will leave the motor side detached.

---------------------------------------- Zap-------------------------

Wow, that thing sounds crazy! I want to boost mine up still a bit more, but I am not sure about even wanting to be able to do a *double* burnout on pavement!

On my first ride I went over to my buddies how to do a "drive by". He was out in his driveway and I yelled "TWO WHEEL DRIVE" then gunned it. I was leaned all the way over the front bars in anticipation of the front burnout only to feel my rear end swinging side to side

2WD is excellent for offroad use, I hope you have access to some good trails. From what I saw on a couple pics you posted a while ago you look to have some nice places to ride that thing.

Tons of off-road around here. There are also BMX tracks.

As for your problem, Methods, I am ready to bet that it's your excessively long common ground wire between controllers that's doing it. On your photos I can see the point where your main battery wire splits into two sets, and this point is much too far away from the two controllers. Remember that the total voltage drop along both those controller's non-common ground wires becomes a voltage offset to anything common to both controllers (worse with current and you are using a lot!)... like the throttle signal as you are using only one for both. Current spikes in either/both controllers end up as noise, so you might get erratic throttle behavior. I also tied both +5V throttle supplies together on my own setup, although I haven't though it through yet if this is a good move or not. It seems to work OK for me, but at about 50V/37A times two, not 100V/100A+ times two!

I agree that the DC lines are way too long. I am only using 10AWG and I am certainly developing an offset on those ground wires.
I cut the wires on all my controllers at 2' for DC power and 4' for the phase wires. The idea is to be able to move my setups around from bike to bike.
In this case I may need to cut the wires - I can extend them later if I need to.

It should be noted that the controllers are identical - with identical offsets, but I can see where there is a problem when I have a common ground back at the battery split then I tie ground together again at the throttle lines.... Can anybody say ground loop :lol:

On thing that does nag me is the fact that they work well when not on BEMF equilibrium... odd. Did you make sure you disabled that "bar protect' feature in both controllers?

What the hell is "Bar Protect"? Sounds like the alarm?
I can say that the controllers were perfectly when they are the only thing hooked up.

I really thing this is a problem to do with the hall sensors. I had a problem like this before and when I shortened up the hall wires it went away.
Maybe tightening up the grounds will have the same effect... I would never build something like this at work
At home I have so little time that it is all I can do to find time to throw something together before I have to do some kind of shit-work like mow the lawn or deal with an ant problem.

You can keep both controllers side by side without problems in my opinion, as the housings help to isolate each circuit. Even if they were bare PCB's side by side I wouldn't worry much, since they already each produce lots of their own crap and thus are designed to take it.

Do you think you could get your physicist sweety to photograph/film you doing a 2WD burnout?
Pat

I knew before posting an outrageous claim like "2WD Burnout at 20KW" there was zero chance of a video not being required.
I will see if I can get the wife to shoot me after work today.

 

[liveforphysics]

Awesome Methy! Get some vids!

 

[fechter]

It could be possible the rear controller is doing some kind of regen braking when it is being driven faster than its Kv.

I would agree the path between the two controller main negative wires needs to be as short as possible to prevent ground loop currents in the throttle lines. Try adding a heavy jumper between the two negative wires close to where they come out of the controller.

 

[no mods]

makes since it would be hall related brings be back to the days of trying to get my MSD (my spark died) dis to fire my coil on plugs for the turbo car. a bit different but every way u can shield isolate or shorten seem to make a difference.

im no electronics "ace" but those r my two cents from my shadetree experience...

and while ur at it why dont u log actual signal voltages going into each controller... maybe there is some weird back feed crap happening that can be taken care off before you dig to deep for dem' gremlins

but again.. with a pipe in my mouth i aint no expert'

peace.

 

[GGoodrum]

That is a crazy setup, for sure. Amazing that the pack is so small, for something that is 100V and 15Ah. I'm using half that (50V/15Ah...) on my 3220-equipped folding bike and it blows me away how light it is, even compared to my a123-based setups. You and Luke are turning me into a convert.

Oh, and I too love the front zippy-mounted front brake. Very ghetto.

-- Gary

 

[ZapPat]

It should be noted that the controllers are identical - with identical offsets, but I can see where there is a problem when I have a common ground back at the battery split then I tie ground together again at the throttle lines.... Can anybody say ground loop :lol:

Identical offsets ONLY if they were pulling the same currents at the same times, but their communtation events are not synchronized an so you will end up with varying offsets at different moments. High frequency unsynched PWM switching events should not be such a big noise problem because each controller's cap banks should be able to filter most of this out.

What the hell is "Bar Protect"? Sounds like the alarm?
I can say that the controllers were perfectly when they are the only thing hooked up.

Bad memory, Methods? :wink: Remember I posted a month or so ago that this programming variable is used to turn on/off the throttle OV safety cutoff feature (something like cutoff output if throttle > 4.2V or so). I'm thinking that your ground loop noise might be giving you spikes that could trigger this protection feature once in a while if it is activated in either controller. But probably not though since it doesn't do the problem all the time (and exclusively) in full throttle position.

I really thing this is a problem to do with the hall sensors. I had a problem like this before and when I shortened up the hall wires it went away.
Maybe tightening up the grounds will have the same effect... I would never build something like this at work
At home I have so little time that it is all I can do to find time to throw something together before I have to do some kind of shit-work like mow the lawn or deal with an ant problem.

I have a really long hall sensor wire extension myself (I have to wrap it arround the frame lots to use up the slack) and it doesn't cause me problems. But then again I'm not creating anywhere near the EMI that you are making there. One thing I do is braid the three phase wires together on all my setups which may help a bit, and of course Arlo1's suggestion of a hall wire shield should help much too.


Maybe a good thing to do would be to try the solutions one at a time so that we could know what the problem was after all? I for one would be curious!

Good luck with the troubleshooting!

 

[chet]

What happens when you swap controllers? Also, could both controllers on one battery cause problems?

 

[Ypedal]

Aaahahaha.a.. *cough up pepsi thru my nose..

Front brake setup.. all go.. very little whoa.. ( i admit.. i've got sneakers with bare soles from similar test rigs.. )

Video man video !! pls !

 

[stator]

Which controller do you have the CA plugged into? Maybe try swapping that first. You have the CA connected directly to one of the hall sensors, correct?

--Bill

 

[ZapPat]

Aaahahaha.a.. *cough up pepsi thru my nose..
Front brake setup.. all go.. very little whoa.. ( i admit.. i've got sneakers with bare soles from similar test rigs.. )

I hope you also have an ebrake setup on that thing. I trust these more than regular v brakes now, let alone tie-wraped on vbrakes!

 

[Ypedal]

Ebrakes ? .. yep.. i got those.. nicely tucked away in a parts bin ! :wink:

 

[methods]

It should be noted that the controllers are identical - with identical offsets, but I can see where there is a problem when I have a common ground back at the battery split then I tie ground together again at the throttle lines.... Can anybody say ground loop :lol:

Identical offsets ONLY if they were pulling the same currents at the same times, but their communtation events are not synchronized an so you will end up with varying offsets at different moments. High frequency unsynched PWM switching events should not be such a big noise problem because each controller's cap banks should be able to filter most of this out.

I meant to say identical length wires.

Which controller do you have the CA plugged into? Maybe try swapping that first. You have the CA connected directly to one of the hall sensors, correct?

--Bill

Hmmm... Good point!
I will swap the CA - it is plugged into the rear.

Aaahahaha.a.. *cough up pepsi thru my nose..
Front brake setup.. all go.. very little whoa.. ( i admit.. i've got sneakers with bare soles from similar test rigs.. )

I hope you also have an ebrake setup on that thing. I trust these more than regular v brakes now, let alone tie-wraped on vbrakes!

ebrakes?
lol....

I have a V-Brake

nomods - this one is for you :wink: This is your future. Mod-fix-mod-fix-mod until it is un-ride-able - throw away

So I got a video - a few cool burnouts (Ok, I used one twice ) followed by the walk of shame. My wife likes to point out just how lame I am.
It is pretty funny...

I dont do all that video compression stuff so it is a big file. 35Mb
Please SAVE AS -> dont try to stream from my server. It is too big of a file.

-methods

 

[ZapPat]

So I got a video - a few cool burnouts (Ok, I used one twice ) followed by the walk of shame. My wife likes to point out just how lame I am.
It is pretty funny...
I dont do all that video compression stuff so it is a big file. 35Mb
Please SAVE AS -> dont try to stream from my server. It is too big of a file.
-methods

You are killing those of us who still live on the information dribble that is dialup, but I should be able to download it while I sleep tonight at least. I specially want to see where the shame part comes into the story!

 

[methods]

Why does my ebike always look so busted and lame on video? I was going 47mph at the end of that drive-by. . .
Felt like I was riding a rocketship

(oh boy... I just bent over to get my ass kicked didn't I? )

-methods

P.S.

I specially want to see where the shame part comes into the story!

Not worth the bandwidth
I am sure someone else will crunch it down to like 2mb....

 

[Ypedal]

Duuuuuude !!!! the ultimate camera shot and the cat gets the spotlight ????????

but i gotta say.. that sucker hauls(/ed) pretty good !

*bummed..

 

[PhoenixOSU]

Could the transfer of energy be resulting in one motor going briefly into regen mode while the other is motoring? I can see this causing weirdness.

 

[John in CR]

Methods,

I thought it displayed the acceleration and speed quite well. The burnouts would be better served on dirt, so we can see something fly. Kinda sorry the cat was in frame at the wrong time. Glad you're ok. How thick are those dropouts? Did it snap, or just not deep enough? You gotta make the oval retainer/torque arm/whatever that encircles the axle and bolts to the bike, so the wheel positively cannot leave the bike, and make the dropout itself much deeper too. Imagine that happening when you goose it at 30mph to pass a car. Do you need the wife to calculate the forces in those dropouts? Burning up motors and electronics, while pushing the envelope is cool, but not easily avoidable and predictable mechanical failures that can get you killed.

So what's the damage assessment? Wiring ok, since it looks like the connects on the fork released pretty well? No smoking shorted cells on the batt pack?

...and what do you mean no bike to get to work? Rear motor is fine, so just zip tie a normal front wheel on and be on your way.

Seriously, glad you're okay. Hope you learned something despite getting off so easy.

John

 

[methods]

To answer a few of the questions above:

Regen is not enabled
The motors are the same KV with almost the same tire diameter

If it were simply that the front motor was pulling or pushing the rear motor into a bad state (say a regen state) I would argue that it should do that when I pedal / apply the brakes with only the rear hub hooked up.
I dont recall having any troubles like this... but maybe the front motor exacerbates a condition that I just don't notice when pedaling.

If I ride the bike in a perpetual state of acceleration (slow ramp up, snap off throttle, .... slow ramp up, snap off throttle) I can ride with no misfires
If I am going 20mph and I just barely crack the throttle (to the point where it should not add forward momentum) it will misfire - so I guess it is just when the back EMF is >= the forward emf?

Anyhow - I have spun many hubs and this one was no big deal. I dropped a wheel out of the front hub going 30mph and slid into an intersection once.
The front dropout had spun a while back so yesterday I welded up new drops out of a pair of 10mm wrenches
I think the problem is that I did not grind the face of the drops (i.e. the part facing the nut)
A wrench tapers as it goes out... This taper probably caused the axle to sqeeeeeeze out.

When I get home I am going to put a reverse taper - The fattest part will be the tips and it will get skinner as it goes up.
This will keep the axle in.


I think I know why other peoples videos look so much faster...
If the camera is steady and the bike blasts by it looks a lot faster than if the camera tracks the bike.
I am going to try that when I get home (errr... after I finish fixing the front forks)

I ended up stealing my wifes bike and riding that back to work

-methods