GM Motor Controller BLDC Zhidong 500w 36 v (Regen) - Overvolt option

[nitecheck]

Hi - I noticed on the forum that some ppl have posted that they have overvolted this controller to 48v and although the motor performed well at the new voltage - the controller actually becomes damaged! I'm a little shocked at this - well more than a little  I have a cute little 9continent motor 250w 36v 7ah battery pack - with the base no name controller that came with motor... the controller is happily running at 48v for over a 1000kms & is still doing fine.

(I did buy a really nice high end 48v controller - should things go pear shape) - but 12 months later the 48v controller still sits in my PC spares cupboard gathering dust & my cheapie 36v generic controller still powers along (even through the odd 40C summer days) - without ill effect!

I was looking forward to running 48v on the new GM motor (my old battery packs & chargers are configured for 48v etc) & was hoping to simply plug in a pack & get going. (I did get a 36v charger & 3 12v 12AH AGM's with the kit - so I'm not out in the cold on this.) And could probably get along with a 36v system with higher AH - but - well I do not want to. 

Really??? - can these controllers not handle a 48v battery pack with out being damaged?

Any comments would be appreciated...

nitecheck

[Electrobent]

what do mean by "the new GM motor?"

Maybe its an issue of wattage and the people with the problems have the 500W motors while your 250W motor does not suck enough through the controller to melt it?

[nitecheck]

ahhh..."new" to me that is 

food for thought re: the 250w motor not drawing enough power to cook the controller at 48v. I just assumed the technology of a 500W controller would be built on the same sliding scale as it's smaller cousins. There a moderate increase in voltage would be within limits of the controller. Suppose I better check the how much power this motor can draw & check the spec's of the controller before taking the gamble.

[philf]

I've modified all four of the controllers I have (three on bikes, and one spare) - if only to add a power switch and bring out the status LEDs from inside (and add heat sink compound to three out of four of them - they didn't have ANY).

The interesting thing about the board design is that they're doing something really cheesy to get the 5V supply for the microcontroller from the incoming battery voltage.  The supply is based on standard 78xx linear regulators, but because a 7805 can't take 36 (or 48) volts directly, they actually CASCADE multiple 78xx parts.  On the 36V version, they put a 7815 in front of the 7805.  And, guess what?  There is a place on the board for ANOTHER 78xx regulator which would be in front of the 7815.  My guess is a 7824 - which would allow you to supply 48V without popping anything.  There is a single power resistor in series with this arrangement, and there is a space on the board for a second one that ties directly into how the space for the 7824 is laid out.

The transistors which drive the motor are NEC K3435B's (which is generically known as a 2SK3435).  These parts can handle 60V max.  My guess is that the only difference between a 24V, 36V, or 48V controller is how they've populated the section of the board that provides the TTL power (which regulators/resistors are present) and the value of the resistor that provides the low voltage cut-off.

If anyone's "fried" a controller by "overvolting" it - I'll wager that the only thing popped is the 7824.

I've got my spare controller open on my bench right now.  If anyone would like to see a picture of what I'm talking about, I'd be happy to post one.

[Dave]

The interesting thing about the board design is that they're doing something really cheesy to get the 5V supply for the microcontroller from the incoming battery voltage.  The supply is based on standard 78xx linear regulators, but because a 7805 can't take 36 (or 48) volts directly, they actually CASCADE multiple 78xx parts.  On the 36V version, they put a 7815 in front of the 7805.  And, guess what?  There is a place on the board for ANOTHER 78xx regulator which would be in front of the 7815.  My guess is a 7824 - which would allow you to supply 48V without popping anything.  There is a single power resistor in series with this arrangement, and there is a space on the board for a second one that ties directly into how the space for the 7824 is laid out.

I've often wondered about the controllers, especially after seeing what a guy did with an off-the-shelf radio control motor & controller (wrote about it here.

The RC controllers are a fraction of the size of the Golden controller I bought with my kit, can be reconfigured by hooking them up to a PC via USB, are able to scale up the voltage without any fear of burning them up and have no heat dissipation issues. What is being sold by Golden isn't exactly state of the art. I'm no electrical engineer, but it seems that there is a lot of room for improvement when it comes to the controllers.

A while ago someone on this board was talking about designing a better controller; anyone know if anything came of that?

[Mel in HI]

State of the art usually carries a matching price tag. 

[Dave]

State of the art usually carries a matching price tag. 

True, but even the higher priced ebike motors don't seem to have progressed beyond the mid '90's with their controller technology. The Phoenix controllers cost probably $100 more than the Golden units, but it would be nice to at least have something like that as an option when buying an ebike motor kit. I'd sure pay the premium for the smaller size and the programmability.

[philf]

I've been delving further into the controllers supplied by GM, and have replaced the internals of the power supply with a switching affair that can run the microcontroller on anything from 7 - 60V.  The LVC resistor is still an issue if you want to run your motor below 34V (this was a 36V controller), but I was stunned to see that there is one other component that is affected by changing the power input...  The power LED.  I've always just assumed that it was on the output side of the 5V supply.  It isn't.  It's tied directly to the input power, with nothing more than a 10K resistor inline.

This might be one more component to fail, if you up the voltage going in.  I'll post pictures of the modified controller with the LM2575HVT-based switching supply once I've tested everything thoroughly (there is an ulterior motive for changing the supply - I want to run a couple of other goodies on 5V, too, and thought it was more efficient to fix the basic supply than add another separate regulator for my add-ons).

I can already report that the switching unit runs as cool as a cucumber, and uses about 60% less battery at idle than the factor supplied linear cascade...