150v xie cheng failure thread

whatever

100 kW
Joined
Jun 3, 2010
Messages
1,297
what happened to knuckles thread warning about failures of the 150v xie cheng controllers, any more info available out there?
 
Looks like he (or someone) deleted it after it was moved to this forum out of the reference area. There was no technical information in it, anyway, only the "warning" that they would "blow up". When people asked for details, no replies were forthcoming with them (in that thread, at least).
 
It would be to do with the stepdown voltage regulator IIRC. The component is specc'ed for an absolute maximum rating less than what it is being used for. My guess is that the internal pass transistor breaks down (messy). This is why at these voltages, unisolated linear or buck supplies are ill-advised. Flyback is much more sensible.
 
Ah, well, that stepdown regulator (if it's the LM317 common to many controllers I'm thinking of) is not even really spec'd for use in the 36V controllers, as their packs are often 45V or higher hot off the charger or during charge. (my NiMH pack is 43.something and charge voltage is a tad over 45V at peak); the max voltage on it is supposed to be something like 40V I think.

Or maybe I am misremembering and it is 40V difference from Vin to Vout, in which case at say 12Vout it'd be max of 52Vin, so say, a 48V pack would be too high for it.

So the voltage dividers are used before it, and although wasteful of power and creating significant heat, they do work to cut down the input voltage to a safe level for it, AFAICR.
 
These ones I think use a 100V buck controller IC rather than the 317. The pre-resistors are just a poor hack to reduce the differential voltage.
 
Ah; Well, at >100V I guess it's no different than the 317 would be at mid-range voltages. :( Still, if it works and the heat can be gotten rid of, and you don't mind the wasted power, it *should* work.

Better to replace it with an SMPS but that would cost more (most likely).
 
All XieChang controller boards are designed for no more than 72 volts nominal.
They can handle maybe 100 volts max before the need to "protect" the R10 and R11 resistors from overheating.

In addition, the components of the switching circuit are themselves rated at about 100-120 volts max.
Specifically, the standard SS110 schottky diode (D8) is rated at 100 volts max.
The tiny switching transistors (Q13, Q17 and Q18) are rated at maybe 120 volts max.

Finally, the five tiny resistors (R116 thru R120) are subject to burnout due to initial "turn-on" current surge.

I've posted a simple "high voltage" solution to address all these issues.
The solution has been tested and works perfectly on my own controllers at and above 150 volts.

If this solution (or an alternate solution) is not used for 150 volt operation, it is predicted (and already proven)
that premature failure of these components will certainly occur.

Vulnerable.jpg
 
thanks for clearing it up knuckles thought you had disappeared!
this is the thread you are referring to?
http://endless-sphere.com/forums/vi...65&hilit=ALL+NEW+150V+30A+XieChang+CONTROLLER
and in particular the knuckles "master mod" pic as per below?
( or is there another thread I have missed?)

I'm not up to the stage of messing with over 100v , I'm still below 90v, but am trying to get my head around mods required for higher voltages with xie controllers. It seemed important to resurrect the missing thread warning of these issues.
thanks very muchly, hoping you will continue to post your researches they are invaluable.
 
Back
Top