Burning a crystalyte 48V20A Controler

[Lagoethe]

Hello,

I bought a crystalyte controler 48V20A + twist + crystalyte motor n°404 + ping battery 48V20A.
I built my little bicycle but without pedals.
Then I turned the twist, and...
I heard a nice sound and my controller burnt on the drain of one of the mosfet.
So my question is: Why?, and how to avoid to burn another one?

PS: I burnt it while loaded, the wheel turned well while it wasn't fixed on the bike.

Good day

 

[liveforphysics]

You likely had one of the hall/phase combos that we call a "false positive". The wheel will spin, but it's got the wrong hall sequence so it pulls a huge amount of current just to try to free spin, and it burns up controllers.

 

[Lagoethe]

So what are the solutions?
-Being sure no to have false positive. I don't even know how this is possible (having a false positive).May it be a short-cut to a hall phase and the +5V or ground?
-Avoid huge current. I could put a current sensor on motor phases, check if the current is higher than 50A for example and cutoff if higher?

Last question: How can I be sure that it's was false positive?

Thanks

 

[fifthmass]

Hi,

I may have had something similar ALMOST happen to me. I say almost because my controller is rated for 40 amps instead of 20a and the extra capacity prob saved my a$$. I have noted on a few occasions a high current spike when starting from a stop. I am talking about 25-30 amps or so. It does not happen very frequently.

My only advise is to treat any controller as a pedal first controller; i.e. pedal up to walking pace or so then hit the juice. It should keep the issue from reoccurring. In my case it made the spikes go away.

take care and be safe,
Bill