Hire E-Bike build thread

[goatman]

heat rises, if the heat sink is above the mosfet it will suck the heat out of the mosfet and transfer it to the case to be shed

right now the heatsink is below the mosfet, just give it a flip so the screws are up

 

[amberwolf]

heat rises, if the heat sink is above the mosfet it will suck the heat out of the mosfet and transfer it to the case to be shed

right now the heatsink is below the mosfet, just give it a flip so the screws are up

That only applies if using air to convect the heat around, and only because hot air is less dense than cooler air, and thus rises, because *heat* doesn't rise. *hot air* rises.

When things are in direct contact with each other, like heatsinks on FETs, etc., that's conduction and heat passes directly into a cooler material from a hotter one (whichever direction that happens to be).


So you don't have to flip the controller around--it depends on your airflow. The way that will make a noticeable difference is if there is a passive air intake vent below the controller, and a passive air exit above it...but it would actually be more effective in that case to have the heatsink pointed *down* at the intake vent, so the cooler air flows directly over the heatsink.

If there is active airflow (like if the case is open and you're riding), then having the heatsink closer to the airflow itself, wherever that is, will help cool it.

 

[goatman]

your running 14s5p of ga. how many amps is the controller? its a 1500 watt leaf motor so 40 amps? 8amps/cell max

sometimes you want heat. where i live the average monthly temperature is 15 celsius or less 9 months of the year, similar to the UK.

i keep my batteries in a plastic DeWalt tool case and if its 25c outside my bike flies and the batteries get real nice and warm like 35c. in winter i put cardboard under my batteries and a shirt on them to keep them warmish and the bike will be sluggish.

batteries ive been testing like heat, 38 celsius max.

you might want that controller in the case to keep the batteries warm.

 

[Hickbeard]

@goatman yeah 14s5p with NCRs. Standard 1500w leaf motor and 40a controller.

That's interesting about the heat.

I guess in the winter I can keep the lid closed. I will eventually get some thermal monitoring sorted.

Sent from my CLT-L09 using Tapatalk

 

[Hickbeard]

Few test rides round the site I work and all was good.

First ride out on the road and this happened.

Snapped the welds on the basket and bent the support on the front.

I think contributing factors were poor design, inadequate materials and solid front tyre.

Remedies will be to mount the battery in the frame as low as possible and proper air filled tyre. Same as on the back.

On a side note. When riding it topped out at about 50kmh. The display read 55v and 1600w. This is slower than I was expecting (but honestly plenty fast enough without any brakes). Need to investigate what's going on electrically.

Sent from my CLT-L09 using Tapatalk

 

[Hickbeard]

Looking at options for locations for the controller and battery.

What could be possible issues with putting controller on the bottom under crank, other than obvious impact.

Leaf controllers are waterproof right?

Sent from my CLT-L09 using Tapatalk

 

[amberwolf]

Essentially, unless it's potted like the Phaserunner and some others, so it can't be opened up, it's not waterproof.

If it's IP67 rated, you can consider it waterproof for the purposes of riding a bike around in the rain.


Even then, if your connectors are not specifically waterproof, you'll wick water into the electronics via the wiring itself, both between wires in multiwire cables, and within the actual conductors inside the individual wire's insulation.

 

[thundercamel]

I'd call it water resistant, with rubber on both sides. Holes for the wires through the rubber might be questionable.

 

[amberwolf]

There's someone in the Nucular sale thread with a month-old controller that doesn't get used in the rain, etc., that was I think at least half full of water, when he opened it up after it blew up. I think it ws mounted on the downtube. That' controller has the rubber seals, etc., too. :/

If you want to mount it where it's going to get soaked (like behind the front wheel, in front of the rear wheel, bottom of the bike, etc), expect that it's going to get water in it--if nothing else, it's going to wick in there thru the wiring, and the connectors themselves are going to get wet, and corrode, too.

 

[goatman]

reminds me of a video by grin about water going through motor cables, skip to about 25minutes

https://www.youtube.com/watch?time_continue=1644&v=gwlbAJLzI_w&feature=emb_logo

 

[Hickbeard]

I understand wicking through wires. Our HV transmission lines leak gas through the actual cables themselves. Like you say in between individual strands. Right pain.

I wonder if I could heatshrink the actual controller. [emoji848]



Sent from my CLT-L09 using Tapatalk

 

[thundercamel]

Well for what it's worth, the controller on my wife's bike has been mounted between the seat tube and rear wheel for over a year now. Just opened it up to do a shunt mod and LVC mod, and it's clean as new in there. Granted this bike doesn't see rain, and only a few puddles.

 

[Hickbeard]

@thundercamel

You could eat your dinner off that, well maybe not. But looks clean.

Sent from my CLT-L09 using Tapatalk

 

[Hickbeard]

Re did the battery.

Much happier with this one. Looks a lot cleaner. Didn't rush it this time.

Sent from my CLT-L09 using Tapatalk

 

[amberwolf]

I wonder if I could heatshrink the actual controller.

Where will the heat get out, with the heatsink fins all covered up?

 

[agniusm]

Have not checked on this for a while. Is there a reason you have not folded tabs and bolted through the holes on the modules?

 

[Hickbeard]

Have not checked on this for a while. Is there a reason you have not folded tabs and bolted through the holes on the modules?

Hey dude.

Yeah I originally had the battery in two banks, side by side. So leaving them up made them easier to connect together.

Now I've changed to one long design but I've epoxyed the modules together so can't open them up to put square nuts in.

Sent from my CLT-L09 using Tapatalk

 

[Hickbeard]

I wonder if I could heatshrink the actual controller.

Where will the heat get out, with the heatsink fins all covered up?

Don't know. I guess nowhere so bad idea. [emoji848][emoji23][emoji2357]

Sent from my CLT-L09 using Tapatalk

 

[methods]

Sick build, lots of good stuff in there. Looking forward to reading it later.

I did not have time to read (only skimmed your pics) but I think I noticed your controller in the case with your cells. In another thread I think I heard you say 40A continuous, and above I saw pictures of a soldered shunt (you look like the modder type)

You may find that airflow over the controller is VERY IMPORTANT.

Or not

But... if your controller starts getting hotter than 60C (too hot to hold your fingers on) then you will accelerate the aging of the components (especially the large electrolytic caps) and you will eventually find catastrophic failure.

Heat builds up quickly in an air tight case.

(source: http://www.schindlerengineering.com/public/2020/05/17/i-tripped-while-jogging/)

Nice case btw... Mine leaked like hell so I tossed it.

(source: http://www.schindlerengineering.com/public/product/original-v0-prototype-the-ip67-ardruino/)

-methods

 

[Hickbeard]

@methods

Haha nice dude. Matchy Matchy. I thought pelicases were meant to be military standard waterproof and such. Was it your mods that made it leak.

The case is gone. Didn't work out as I'd hoped. Couldn't secure it to the front so more traditional e bike battery placement.

Going to make a fibreglass enclosure for it, and maybe the controller too.

Sent from my CLT-L09 using Tapatalk

 

[methods]

My case was a $5 GoodWill (used store) special. It leaked at the hinge. If it were greased it probably would have been ok, but I jumped off a 20' cliff with it dry, plunged about 2 meters, and it came up with a few dribbles.

Moved to ABS after that.

http://www.schindlerengineering.com/public/product/ip67-c-o-t-s-modification-housings/

I am looking at Rovers and Thruster Powered rigs. ABS and PVC are the win for that. Notice the flexible membrane... this helps alleviate the pressure differential. Trapped air is our enemy, so most people back-fill with oil, wax, or anything that will replace something that compresses with something that wont.

-methods

 

[Hickbeard]

Couple updates. Although not much physical work done.

Got my hydraulic brakes today. Clark M2s. 180mm on front, 160mm on back. Fitted back disc. Absolute pain to get rear wheel back on. Really tight fit.

Fitted battery housing to frame and located where controller is going.

The tape is rough outline of where the fiberglass enclosure will cover.

Thinking of running a vent into the battery housing. What are peoples thoughts on something like this hose? Running from top underneath then up the bottom. [emoji1787][emoji44][emoji1787][emoji2357] Then some vents at top to allow hot air out. Will give decent airflow?

Then will trim down lid, fibreglass it and fit hinge at top, rubber seal and lock mech at bottom.

Sent from my CLT-L09 using Tapatalk

 

[methods]

Some pics did not render.
If you let water in, then let it back out.

I would avoid dependence on fans on an ebike.

My personal testing has shown that if you keep a sizable air gap between energized conductors you can run wet at high voltage (100V) for many years. The moment you bridge with liquid - science experiment - involving crystals.

I saw little ingress into my Hobby Packs, but the 100mil spaced JST-XH ballance taps are all destroyed. Inside the hobby packs the path between electrodes was a tough one, so never ate one up.

... The top of an 18650 is a breeding ground for electrochemistry experiment. Just take a teaspoon of water, run it down the side of your setup (to absorb dust and salt) then pool it on top of a fully charged battery.

-methods

 

[Hickbeard]

Wasn't thinking of fans. More like an air scoop at the front running to bottom, with space for water to escape and a vent at the top for hot air.

Not sure about pics. All work for me [emoji1787][emoji2357]

Sent from my CLT-L09 using Tapatalk

 

[methods]

Moisture is not your friend.

Let us know how it turns out. Sometimes no impact, sometimes some white dust, sometimes green crystals, sometimes brown goo.

If there are any exposed electronics with tight spacing (less than a few mm) conformal coat them or otherwise keep them free of moisture/dust buildup. Grease works but collects conductive dust and is a nasty mess.

...

If you play with electrochemistry, say by taking a glass of tap water with two exposed copper wires and a variable power supply ...
you will get a really good feel for the impact of higher voltages.

1V
10V
100V

Probably wont see much going on. Now shake a little salt in there

1V
10V
100V

Shake some more if you need it.

At something like 1-2V.... and very little salt.... the affects take a very long time. Maybe the copper just discolors a bit. If you ramp that up to something over about 3V.... you will start to see degradation. Maybe over minutes or hours. If you then crank that up to 10V... it takes only minutes. If you crank that to 100V - seconds.

Lipo bikes tend to have tight spots with a lot of potential across a small gap. If any condensation (along with absorbed dust) manages to collect you can see rapid degradation to your reliability.

Throttles start acting funny
Controllers get unreliable
Packs start going out of ballance and eating balance taps

Stuff like that. Dialectric grease helps tremendously for connector pins, and anything that does not require conductivity should be drenched in conformal coating. Silicone or Acrylic work fine, I use MG Chemicals spray on. Cleaning first with a harsh solvant like MG Chemicals Heavy Duty Flux Remover helps to keep the bond good for years.

Dont get it in your eyes

-methods