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

[John in CR]

Please, proper overkill on the dropouts this time, and no wrench is deep enough to qualify. You've dodged the bullet enough times already...even Neo finally caught one. We want to keep you around here.

John

 

[no mods]

BUHAHAHA !! @ vid. that ish could be on some sitcom show. Your toy r us brakes are audible like a mile away...


im seriously a dork laughing at ur internet video out loud!

btw what happens if u separate both systems? when u get it back together see if you can run 2 throttles so each system is completely Independent just for trouble shooting purposes that way u at least have a solid base to work from.

** i was thinking about quitting and doing something else with my life... but now im gonna continue to put in 60hrs a week as long as i can just so i can build stupid stuff with lipo.

 

[methods]

Lipo is awesome...

The video does not do the bike justice - It feels way faster than it looks
Like the first time you fly an RC plane with a Lipo pack instead of a busted old 7S NiCd pack...

I cant wait to build my Astro bike.

- methods

 

[icecube57]

HERE IS THE SAME VIDEO AT UNDER 10MB.

 

[Arlo1]

Thats to cool I just got out of my funk watching that! I need to build something like that for sure! Thanks for posting video. Wow to funny.

 

[Doctorbass]

The video does not do the bike justice - It feels way faster than it looks
- methods

I totally agree with that..

A secret to render the best perceivable speed by the viewer is to have a wide angle on the camera!!... that viewer to also see the side of the street moving faster...

I played in loop the sound of the spun dropout at 1:10 of the video... OUCH !!!!! .. .it's great to seee it did not happened at 47mph!!!

Doc

 

[steveo]

Hey methods

Thanks for posting a video!!!.. your lucky you didn't kill yourself with the front motor spinning the drop outs..!!!

I will probably make my project bike this winter ... 2wd bmx with star/delta golden motors!!

I just need some funds to start the project..

-steveo

 

[methods]

Update:

I already welded up the drops and rode the bike to the store with the trailer to get groceries.
I have pictures but I need to download them from my phone.
Will post later.

I also found the misfire problem!
I technically dont *know* what the problem is but when I show you this video it will all make sense and I think we will solve it.
I will try to get it done tonight....
It is possible that I may need to diode isolate the controller signal inputs from the throttle output. Wait till you see the vid.

-methods

 

[methods]

So I got right to work with another Ghetto-Fabulous-Half-Ass-Methods-Fix....

What more could you need? A framing hammer, monkey wrench, and my trusty Angle Grinder (that I only paid $20 for and wont turn off... I once plugged it in while it was sitting on my bathroom floor and it ran all the way across the tile I had just set a week before... Tania was NOT pleased )



So I didnt tell anybody this but I had actually spun the front hub several months back.
Thats why I went with the rear hub....
Day before yesterday I welded in a couple of 10mm wrenches to serve as drops
As noted - they are not deep enough.

I went back with the grinder and cut them a few mm deeper
I also built out the ends and then grinded it all down.
There is lots of metal to get purchase on now.



By dusk I had the "Double Booster" back on the road!

(FYI: Some may recall that there was "Project Mayhem", the "Ghetto Booster","The Tender Brandson", "The Choppa", "The bonner (still in production)", and a few other one-offs that never got names. Would not want to mix things up now... This bike is the Double Booster ((for obvious reasons)) and the 3220 Astro bike is the Ghetto Booster MkII)



That is the same trailer that I towed my wife away from the church in at our wedding.
btw: Did I tell you guys that I made her say that she would "support the Electric Revolution" in our wedding vows? I swear to god, you can ask anyone - it is EVEN ON VIDEO!

And nothing would be complete without the trademark of methods Ghetto
(which I stole from Doc - making it twice as Ghetto)
... The wrench ...



Talk sh*t if you want - I have never spun an axle that had a pair of wrenches bolted to it. I have spun more axles than most people on this board - but never ejected a single one when it was hooked up with one of these 18mm bad boys :wink: That is some serious "Uncle Rico" work right there. . .

I have to heat them up with a map torch and beat them flat with a 10lb hammer.
Anybody who knows anything about Honda's knows that 18mm wrenches are useless anyway... All you need is 10mm, 12mm, 14mm, 17mm, and 19mm
You can just throw the others away

-methods

 

[methods]

I rode that bad-boy to Safeway and brought home the bacon.
Other than the fact that it had no breaks what so ever - it worked well.

So... I think no_mods hit the nail on the head when he suggested running seperate throttles to eliminate variables. This way we can have them both running at the same time but totally isolated. I did not try this yet - but check this out!

(sorry, did not get around to the video - you will have to close your eyes and visualize)

Lets take 4 situations
Front motor off the ground: With only the front controller powered. With both controllers powered
Rear motor off the ground: With only the rear controller powered. With both controllers powered

In the case where only one controller is powered, whichever wheel is off the ground just acts normal. It spins

BUT

If you turn on both controllers and try to run one of the motors in the air.... Shit hits the fan!

I lifted the front wheel off the ground
I turned the throttle just enough to get the front wheel turning while stalling the rear wheel
The front wheel started freaking out and misfiring!

I then did it with the back
I lifted the rear wheel, turned the throttle just enough to get the rear wheel spinning with the front stalled and the REAR freaked out and started misfiring!

I cant think of any situation where a controller would "yank down" the throttle line or anything like that...
The Infineon does do weird things when it is stalled - the Kelly will just plow right through!
So it could be noise through the throttle lines, it could be noise through radiation - either way -> I need to isolate the systems.
They can share a common ground, but only one - at the battery.

I will first try to repeat the above test with two throttles - then move forward.
I want to do as little work as possible because I really only built this "because I said I would" and I am going to move on to the Ghetto Booster MkII soon.

-methods

 

[Arlo1]

Again awesome, love the wrench! But I am doing some diging and can't find much about your controlers? You are buildning them? Is there a sticky or a faq page for them? Where can I buy one or build one from? Before I get to far ahead I need more info on them. Thanks.

 

[diver]

that is just way to kool..I love it

 

[karma]

wow that bike fly's good thing the wheel spun out going slow. the sound of slapping metal was enough :wink:

 

[Doctorbass]

MEthods, Your uinvestigation become bigger than you expected I guess!... let's try the throttle line insulation with diode or capacitor and see what it does!

About the "Doc Wrench Method" I never spun any axel and always built the dropout twice stronger as i would ... to ensure... The wrench help for two things:

-that avoid the nut to loosen and make the dropout loosing support by lateral friction with the nut
-that protect you in the case where it would spun anyway.. the wrench help keeping the wheel on the fork!... But to ensure this you need more fixture from the fork and the wrench.. not only one.. but at least 2 or 3 :wink:

I still hope the X5 will one day come with better axel to dropout solution!

Doc

 

[methods]

...I am doing some diging and can't find much about your controlers? You are buildning them? Is there a sticky or a faq page for them? Where can I buy one or build one from? Before I get to far ahead I need more info on them. Thanks.

The good news is that (as of Aug 18) you only have 37 pages to read

18 Fet Thread

-methods

 

[methods]

About the "Doc Wrench Method" I never spun any axel and always built the dropout twice stronger as i would ... to ensure... The wrench help for two things:
Doc

Hey, you dont have to convince me Doc!
I like the idea. Cheap insurance.

-methods

 

[methods]

Success !!

I hooked up a thumb throttle to the front hub and a twist throttle to the rear
Absolutely perfect - no problems.

oh... and btw -> It is frigging AWSOME!

I can take the rear motor up to top speed like normal, then when I want to sh*t my pants I roll my thumb into the turbo button
Man - you can really fell the difference in power! Lurches forward from 36mph

Good times!

Thanks for the help everybody
I will post my final solution after I figure out what combination of caps/diodes/grounds/shorter leads allows me to run 1 throttle between the two
For now - I will double-throttle it.

-methods

 

[ZapPat]

I just checked out how my own 2wd setup reacts in the situations you tested above, Methods, and it turns out that my two controllers have the same symtoms as yours (although I never noticed mine doing it during normal driving, just when stalled). So I pulled out the multimeter to take some DC and AC voltage measurements on both the throttle and DC battery lines, and compared the normal operation readings to the "buggy" readings - they don't look to be the source of the problem after all (although a scope would be better to check these signals if the problem isn't found elsewhere, as the gliches may be way too short to catch on a multimeter).

I agree that there are some startup glitches with these controllers (I am using 116 processor-based controllers as you are), and it's these glitches that are causing the problems in the other controller. However, this does not seem to happen to me during normal riding conditions as you described at first, but it does happen when I do the standstill stall tests.

On funny thing about this glitch is that my older 12 FET infinion has the same "chunky" bug show up anytime I have the throttle at the bare minimum position, moving or not. And this is only in a one controller setup... :? And even though they changed processors, they are obviously using the same codebase in both the new 116's and the old infinions.

I think I'm going to try and feed the twist throttle with only one of the controller's 5V supply and GND, and feed the signal only to the other controller. That throttle-ground / power-ground loop might be what's causing that problem even though I tried to keep the power ground point close to both controllers... it's still not the same ground plane.

 

[John in CR]

Methods,

25hp 2wd bike with no brakes and an angle grinder with a switch that's always on...You definitely with the prize. I just hope it doesn't turn into a Darwin Award.

John

 

[ZapPat]

Results of only using one 5V and one ground connection with the throttle (both signals connected in parallel as before, but one controller's 5V and gnd throttle connection is not used):
Same problem
As before no matter which controller I stall, the other one rata-tat-tats as it spins. Until of course the stalled controller decides that it's stuck and disables it's output, then the other becomes nice and normal again.

Methods - Did you try the same front/rear stall test as before to see if the problem still happens with two throttles? I ask this because from your last post it sounded as though you started with the rear motor only, then added the front throttle afterwards. This would help isolate the problem a lot.

 

[jsplifer]

Ohhhhh Method no!

The part were the pieces of your bike are falling off, on the way to the door, the head down... classic.

Its good you didn't break any bones though.

Now the bike. The bike screams dude! No doubt about it. And confirms my suspicion that dual hubs would be kick a$$. Speed is totally apparent even with camera tracking. Very nice! I hope you get her back up to speed as I will be interested in how you solve some of the hurdles with a dual hub configuration.

Cheers,
jsplifer

 

[methods]

Methods - Did you try the same front/rear stall test as before to see if the problem still happens with two throttles? I ask this because from your last post it sounded as though you started with the rear motor only, then added the front throttle afterwards. This would help isolate the problem a lot.

Glad you are working on this for me now cuz I dont have time.
The only test I did was last night in my underwear (the test that showed failure by lifting a wheel) then this morning all I did was add a second throttle and take off to work.
When I go home for lunch I will try lifting one wheel then the other - but I can tell you already that it wont be a problem. I rode all the way to work operating the two controllers independently and had no issue.

I wish these controllers would not try to be so slick... My Kelly is a brute! You twist that throttle and by god - it goes! It will dump the entire current of the battery into the coils instantly.
This behavior that these infineon type controllers display is a little annoying. . . But we will figure it out soon enough.

-methods

 

[no mods]

I will post my final solution after I figure out what combination of caps/diodes/grounds/shorter leads allows me to run 1 throttle between the two

dude.. what about a progressive throttle system?? that would be Fing dope.. like a speed Dependant boost regulator. i know autronic and aem (and hell maybe crome or uberdata now) have pulse-width modulation for boost in their software. AS you know, by using inputs on speed and or manifold pressure software tells ecu to output 0-5 volts respectively. That signal then goes to whatever electronic solenoid you use and in turn you get more or fewer lbs from gate going to manifold.

wonder if that can be built from hardware alone?? or perhaps even simpler with a "stereo" style potentiometer with one leg to start after the other... seems easy in my head.
1 throttle- 2 diodes 1pot?? that way if it sucks ass you can just retune pot for start points or:
1 throttle- 2 diodes 1pot that runs some curvilinear path

 

[no mods]

ok heres the jackpot!!

speed would be cool TRACTION Dependant would be even cooler. that way you can run full throttle and take the human error out and have a mind boggling amount of shit happening when u explain to someone u have 2 wheel drive bike with speed Dependant traction running on two controllers with 100 volts of power coming from lithium- the same chemistry found in ordinary cell phones.

so... here my thoughts..

1 mag/hall on each wheel
input into CA or maybe it would be easier for standalone unit?
generate 2 output sources ...

here is where my imagination takes off...
have a box, with lots of cool readily available parts inside, take input from the linear throttle output signal- run a comparison algorithm between wheel sensors and reduce power to the "fast" wheel by the minimum % necessary to remedy wheel slip.
its that easy!! <-----RIGHT?? haha maybe it would be easier to hack a abs ecu??

or am i way out in left field and need to go back to my ignoramus pipe-in-mouth shade-tree corncob whittling hobbies?

 

[fechter]

Results of only using one 5V and one ground connection with the throttle (both signals connected in parallel as before, but one controller's 5V and gnd throttle connection is not used):
Same problem
As before no matter which controller I stall, the other one rata-tat-tats as it spins. Until of course the stalled controller decides that it's stuck and disables it's output, then the other becomes nice and normal again.

That is strange :?

I was going to suggest the same test. If it runs OK with two separate throttles, then there must be enough voltage difference in the throttle lines to make a ground loop of sorts. Using a heavy duty connection between the two controllers' main negative power line should keep the ground references close, but apparently it's real sensitive to small glitches. One approach might be to run a heavy wire between the two controllers attached to the throttle ground trace on each one.

As I recall, the input resistance of the throttle line is around 100k. It should be possible to make some kind of buffer between the throttle and the two inputs. Assuming the problem results from noise on the throttle signal line, you might be able to use a simple RC filter on each throttle input. The capacitance would need to be low enough to not slow down the response, but high enough to filter the spikes.

Similarly, you may be able to insert a fairly large inductance in series with each throttle signal line to filter it.

It may only be necessary to filter the input of one controller (controller 2) if the throttle is directly connected to controller 1.