flathill wrote:Nice little controller but I think you are making a big mistake by potting the whole thing is epoxy.
I totally understand, technically, this is the most robust solution as far as vibration and moisture protection go, but I still think it is a design cop-out as now you just made a throw away brick. You can't salvage parts from a broken controller. You can't repair a broken controller. You can't hack a controller.
Well maybe it would be wasteful to let controller get destroyed by moisture or sea salt in the air? Any way, if controller is reliable, and I believe it is, epoxy is a great way to keep it out from bad outside conditions.
Hey Flathill, I sympathize with where you are coming from, but suggest it is premature to say this is a "big mistake"
. Potting isn't exactly a cop-out, it has more than its share of challenges as well. But yeah making an openable and serviceable controller that ALSO has the same level of vibration resistance, environmental ingress resistance, and thermal dissipation capabilties as a potted unit is indeed a massive undertaking. This is what we tried to do in the first run of 50pcs on the start of the thread, and arguably it was a bit of a failure. Even with the gore-vent to alleviate pressure differentials they can still wind up with water ingress, there's little chance they would stand up to the vibrations of being on a skateboard say, and there was absolutely no way for us assemble/manufacture them in this method and come close to breaking even.
To your points about repairing and hacking, the goal is to make a part that requires neither of those. People only hack controllers because lots "economic decisions" in the component selection and assembly of china supplied stock devices so there is easy room for improvement. If a device is already using the top of the line mofsets and high ripple bus capacitors, then there is nothing left to hack from the hardware side that would provide gains. And in terms of repairs, I've personally dealt with 100's and 100's of failed motor controllers over the years and am sick of it. At least 95% of those fall into 3 categories:
a) Water ingress / corrosion damage
b) Vibration damage (wires breaking off the PCB, capacitor legs breaking, mosfet screws scoming loose, )
c) Mosfet failures.
A and B would have never happened had the controllers been potted, and C can be almost completely eliminated with properly robust power electronics engineering. You won't see mosfets randomly fail on a controller that has thermal rollback, pulse by pulse current limiting, a finely tuned gate driver circuit etc. You can take one of these controllers running at full power and then suddenly short all the phase wires together and the device wouldn't suffer a blip. You can have it doing regen at full current then yank out the battery pack and before the voltage has a chance to rise up and damage things all the regen will cease.
Even with a, b, and c, taken care of, there inevitably will still be a few random mystery failures that result in pretty looking bricks. But if we can get this down to << 1%, ideally <0.1%, then overall people will be much more content and less troubled than the current situation of controllers with high single digit failure rates, even if those failures are sometimes repairable.
Properly ruggidized electronics are potted. Properly designed electronics have an extremely low failure rate, particularly when well potted to defend against vibration and humidity.
Precisely. We can learn a lot from automotive here, since that is one application where vibration and exposure are similar to what an ebike can see. In that world, almost everything is potted.
Previously competed in the Suntrip race on a back to back tandem solar powered row/cycle trike
. 550 watt solar roof, dual Grin All Axle hub motors, dual Phaserunner controllers, 12 LiGo batteries, and a whole wack of gear.
Now back in Vancouver learning to be a dad with my Big Dummy Frame (yes This One
, thanks ES!) with GMAC 10T rear hub motor, Phaserunner controller, and 52V 19Ah EM3EV pack
My website: http://www.ebikes.ca
Please contact via email, firstname.lastname@example.org
, rather than PMs, which are disabled