build thread: klein mantra hs3540 5kW 70km/h 20" (done)

[izeman]

thank you that you took the effort to post. i thought of a longer shock as well. but i will try pumping it up to 300psi and see how it works. i could have change the geometry and raise the lower support of the shock. but decided against it, as it would have changed the geometry. and i have no idea how much the riding changes when steering angle becomes steeper.
today i let the rear frame powder coated. i will get it back next week. i also still have one week to adjust cable lenght of CA, brake signal wire etc ... so it will be a clean build.
i also installed an on/off switch in the battery case, and upgraded my led light. i think i can post some new pictures tomorrow.
can't wait to finish the build - after about 6mo of building and 6more of thinking and engineering.

 

[izeman]

update 20130409:

got the frame back from powder coating. has got a real nice finish. no need to use filler and a lot of sanding. paint is quite thick.

some detail:

didn't take too long until i found a real problem. i did install the controller and did some test drive, but there was one thing left to do: weld the little support thing that holds the shifting wire. so when thinking about where exactly to weld it i just copied the position from my old frame.
can you spot the problem?

this thing is exactly where the controller is supposed to be bolt on. DAMN IT. i have to remove it, repaint this part and rethink where to install it

 

[izeman]

update 20130416: bike is done. well - almost. there is some fine tuning and ongoing work like painting small parts and making a cover for the cables behind the controller, or hiding the cables a bit (it was all just bundled together for the first long test ride), but over all it's done. i'm quite pleased. rides nice, and seems to have a descent range - though i didn't test it for more than 20km.

some things to change:

.) i will remove the twist throttle and add a thumb throttle. i hurts my ankle really bad to twist this all the time. the new handle bar seems to make that much worse
.) i will see how to get rid of heat. it's REALLY hard to keep that motor below 120°C winding temperature. i like to go 3500W, but that seems to be too much on the long run, even when i let it go easy. once temp raises above a certain limit it will stay there, as temp starts to even in the motor.

this is what i looks like now:

 

[crossbreak]

Nice to see the result! Inspired by yours and merlins builds i also bought a clite. Still have some wheely problems i wanna solve by enlarging wheelbase as well as u did. I now use only 50amps@12s lipo in a 26"wheel and it already gets warm. Got a holesaw and protective finish to drill some holes into it, like described in the hotrod threads.

whish u a lot of fun with yout build, looks like it's built to last, like!

 

[amberwolf]

.) i will remove the twist throttle and add a thumb throttle. i hurts my ankle really bad to twist this all the time. the new handle bar seems to make that much worse

"ankle"? :? Exactly what position do you ride in, anyway?

 

[izeman]

wrist wrist wrist ...

 

[izeman]

it's funny how personal feelings can be missleading. i thought my new ride is "ok", but nothing special, and was a bit disappointed. i thought it was not much faster than my 1500w mac. yesterday i charged my fully with the mac installed, and took it for a ride.
WHAT A DIFFERENCE! these bikes are totally different. my klein is more like a motorcylce. long wheelbase, very soft and comfortable ride. i think the tires help a lot making this a cosy ride. performance is ways better also ... the fully feels so stiff. still going up the hills in the forest like a pro with only 1.5kw power. and heat is not really a problem if you don't overdue it.

but when a friend of mine did a test ride bad things happened. he said it was pulling like a train, but suddenly the throttle refused to work. motor stuttered. now it's not working at all. CA shows sensors not switching (setup -> throttle). connected a sensorless controller to the motor and it worked fine. now i must check if the controllers is dead, or some halls died. just can't imaging why any of the two things would have happened. motor was cold, as well as controller. test ride was under a minute, and CA limited temp to 120°C. temp was around 80°C anyway ...

 

[crossbreak]

Sad to hear that. I drive my HS3540 only sensorless. I even managed to push AWG12 (I think) wire through the axle, after removing the sensor leads. Desoldered these AWG12 wire from a 15Fet Greentime controller, they have relatively thin insulation and are about 3.8mm in dia. I now have only one tiny lead for the temps sensor additional to the phase leads in the axle. I will change my seonsorless 15Fet for a sensorless 24xIRFB3077 Greentime since it's not much larger/ more heavy. Still got no longer wheelbase but this will be the next thing on my list, since it wheelies a lot. Any intend for oil/air cooling? How long does it take to make the CA limit the motor power? 80°C after only one minute?!

Your build has inspired me a lot, thx for posting this stuff!

Here is a pic of these AWG12 wires:

compared to the stock wires:
http://endless-sphere.com/forums/download/file.php?id=108139

 

[izeman]

could you please repost the pictures? they have vanished.

for every 12awg wire you want to use, you would have to remove insulation. then use the smallest heat shrink tube that will fit, and shrink it. that way it will quite easily fit through the axle.

CA starts to limit at 90°C and stops power to the motor at 130°C, trying to maintain 120°C by limiting the power fed to the motor. this works REALLY well, but of course could stop you from reaching the top of the hill - but safe a motor instead.

reading justin's thread about motor cooling/heating i may have put the temp probe in a bad position. i put i directly below the connection between phase wire and winding. this is the place which will get hot very quickly, well before the stator reaches some serious temperature. this is maybe too conservative.

 

[izeman]

found the culprit. broken hall sensors. and by broken, i mean REALLY broken. the left one had one broken wire (hanging loose), and the middle one had a wire's insulation rubbed off. i hope you can see this in those CRAPPY blackf****nberry pictures. i hate that phone ....

glad i have three halls left. will swap all of them for new ones, order some more - just in case, and hope that the controller did not blow the fets because of that.
but the question remains: how can this happen? why could the case rub on the wires? there should be enough space. i will epoxy those wire to the windings now and listen for some rubbing noise when reassembled.

 

[izeman]

replaced the halls. epoxyd them to stop wires from moving around. there is really little room between wires and case.

but the motor is running fine again. there is something rubbing inside. will have to find out why. when i reassambled it, it saw that the position of the axle can vary quite a bit depending on which side i close first. the axle can be moved some 2mm. how can this be avoided. will add another thread for that.

i will buy some new bearings and then start with oil cooling.

i thought that the metal epoxy would be more heat resistant, but i found out that it only can withstand 120°C constantly and 150°C peaks. as the temp monitoring of the CA is set to 120°C max this should be ok. what do you think?

 

[crossbreak]

i thought that the metal epoxy would be more heat resistant, but i found out that it only can withstand 120°C constantly and 150°C peaks. as the temp monitoring of the CA is set to 120°C max this should be ok. what do you think?

This has concerned me, too. I use "Epoxy Putty", the grey stuff.

Epoxy is no thermoplastic. It does not get waxy if it gets hot I guess. As long as there are no great forces on the material, everything will work fine I guess. I'll have a talk to some friend who studies Plastic Chemistry Science at my local university to confirm this.

Hope that oil coolant can keep temp below that. I will start with thin silicon oil made for transformators, Seems to be perfect for our purpose: http://en.wikipedia.org/wiki/Transformer_oil

Btw, I am impressed, that a 20" rim has such a great effect on performance, compared to a 26" rim. I entered both systems in the ebikes.ca simulator, having the same no-load speed and the same battery power draw There are worlds in between

 

[izeman]

i'm not sure that you entered equivalent data for both motors. volts and amps are different. so power outcome and overheat time is different too. not that easy to compare. you should set the same voltage, and same amps and see how fast or how steep you can go. at least this is how i did it.
it's funny that you can reach the same top speed with both motors and it's more efficient in the 20" rim. at least this is what the charts says: from my personal experience it SUPER EASY to kill that motor even in a 20" rim within minutes if you find a hill steep enough (+10%).
btw: you should enter 21.6" for the wheel, as this is the exact bicycle equivalent for the 20" bmx rim with the moped tire i used.

 

[izeman]

oilcooling mod started today.

took the whole motor apart, and inspected everything. all looking fine, windings are fine, but some rubbing on a winding had occured. it's related to my other thread about axle position relative to the case. the axle can be moved one? mm from side to side. farest left and farest right cause the case to rub at the stator.

here you can see the slightly damaged winding. i put some silicon on it, don't think it could be a problem

this is the high temp resistant silicone. up to 250°C

put it one the wires, and in the axle slot, and covered all hall sensor wires so they won't be in touch with the oil

will let it dry now, and maybe start drilling the hole for the breather.

 

[izeman]

made the breather plug and bleed screw today.

drilled a 5mm hole and cut a m6 thread. inbus screw has a 2.5mm hole and can take a small tube if i decide to use it.

screw installed with a o-ring

inside of the case. the screw is protruding a bit so oil can flow around it

made a bleed screw as well to be able to let the oil flow out in case i need to replace it. the screw is flush with the case inside

sealed the case, and will fill 100ml of atf tomorrow. let's keep fingers cross that it's leakproof.

 

[izeman]

i put 90ml of atf in it. to be honest i don't see THAT much of a difference. i didn't get worse. but not that WOW much better as i was hoping for. it may be that most of the oil was sucked up by the stator. so i may fill in some more.
it could be that the motor still reaches high temps (80°C @3500w up a 10-15% hill, wot after 1km), but maybe then better holds the temp and comes down faster. i will do some more testing.

 

[Observator]

Got around the test it some more after the oilcooling mod?

 

[crossbreak]

grinded a plate for my friction drive today. The Epoxy Putty that I used to fasten a ball bearing in the sheet got waxy After cooling down it was crumbly...really a good to use high temp epoxy in the motor

 

[izeman]

Got around the test it some more after the oilcooling mod?

no more testing. atm i'm rewiring my battery, and installing a PAS sensor. i will add some more oil in the next days, and will report then.

 

[izeman]

grinded a plate for my friction drive today. The Epoxy Putty that I used to fasten a ball bearing in the sheet got waxy After cooling down it was crumbly...really a good to use high temp epoxy in the motor

what product exactly did you use? "standard" epoxy won't stand 120°C.

 

[crossbreak]

what product exactly did you use? "standard" epoxy won't stand 120°C.

i used "Wiko Epoxy Stick", it's made for 120°C short term 150°C. Question was, what happens if it gets too hot? Now we know it

 

[izeman]

what product exactly did you use? "standard" epoxy won't stand 120°C.

i used "Wiko Epoxy Stick", it's made for 120°C short term 150°C. Question was, what happens if it gets too hot? Now we know it

omg. that's the same stuff i used for installing my temp sensor and hall sensors. what was the max temp your motor reached? i know for sure that the windings never exceed 120°C because that's the limit the CA is set to.

edit: maybe jb-weld would be the right stuff to do it. 315°C max temp http://www.amazon.de/JB-Adhesive-Weld-2-Komponeneten-Bindemittel/dp/B003OIHTRM/ref=pd_sim_sbs_hg_2

 

[izeman]

grinded a plate for my friction drive today. The Epoxy Putty that I used to fasten a ball bearing in the sheet got waxy After cooling down it was crumbly...really a good to use high temp epoxy in the motor

hold a second: you grinded the plate to which the bearing was glued with epoxy? grinding can make the metal GLOW. which is at least 600°C. no eopxy would withstand that temp. even if it was not glowing you could have easily reached temperatures of 300°C and above.

 

[crossbreak]

i have no idea how hot my grinder got the alloy plate. Guess around 250°C. Have no CA yet, arrives on Friday, then I can tell how hot it gets :?

could perform some tests using a heat gun...

 

[izeman]

i have no idea how hot my grinder got the alloy plate. Guess around 250°C. Have no CA yet, arrives on Friday, then I can tell how hot it gets :?
could perform some tests using a heat gun...

so i'm not worried any more. what did you expect? every epoxy gets soft and starts crumbling when you overheat it. that's ok. and this is how you remove it if it's unwanted. heat if up and pry it away with a screwdriver or any other tool suitable.
i think "our" epoxy is good enough for the motor internals, if we keep temp limits below 120°C.
nether the less i order some jb weld shipping from usa was twice as expensive as the product itself *ggg*