XLD Brainpower 886C-6 / LCDS886 - NOT Reversible?

[ElectricXtment]

Hey there,

after reading a lot here, I finally got a Project where i have a question
the rest of the Internet could not answer... So I hope someone here
has some Idea about this issue.

Hardware used:
Controller:
XLD Brainpower 886C-6
Display:
LCDS886
Motor:
Gear-less BLDC with 3 +1 Hall-sensors (A,B,C,S), 350W

So the Issue is that i use the wrong controller for my application.
As I want to use this setup to propel my Dinghy (Little Boat to get
to the big boat anchoring), I definitely need to be able to reverse.

But the Controller is built for bikes so it does not have this option.

So i thought that i will swap 2 Hall sensors and 2 motor phases when
in halt, and reverse. Far from true!
It works. ...sometimes.

It is a 50/50 chance that the motor turns in reverse, the other times
it goes in the same direction as before, but grunting and moaning...

When the direction swap is successful, the motor turns approx.
1 Pole in the wrong direction for a blink, then rotates in the desired
direction.

So my conclusion is that the controller somehow remembers the
orientation and gets a logic-hiccup when it fails to reverse...?
Guestimations...

The other Idea I had was that the Hall-sensors may be oriented in
a direction, not symmetrical to the poles and the controller has to
"learn" in what direction this is... Possible or unlikely?

So I guess that i will have to use this controller in some other project
if there is nothing to be done about this...

BUT! If there is some possibility to hack the Protocol between the LCD
and the Controller... Maybe one could tell the Controller to do things
it was not supposed to. Any hints about the Protocol? Read somewhere
it should be UART, but what are those guys chatting?

So I hope someone has some pokes in the right direction... :wink:

Cheers!

 

[amberwolf]

First, you said it's a gearless motor, but it has a separate speed sensor hall wire? I've only ever seen geared hubs (which generally can't reverse) with that configuration (because the shell has the speed sensor, since if you're coasting and not using hte motor the regular halls won't detect any motion, as the motor isnt' spinning).

Which specific motor are you using, so we can eliminate this as a possibility?

So i thought that i will swap 2 Hall sensors and 2 motor phases when
in halt, and reverse. Far from true!
It works. ...sometimes.

It is a 50/50 chance that the motor turns in reverse, the other times
it goes in the same direction as before, but grunting and moaning...

NExt, for the above, this means it is a wrong phase/hall combo. You can indeed swap any two phase wires to reverse a motor, but you have to swap the appropriate halls to go with that, and they may not be the same color ones. So you have to find the right reverse combination to work with it. Then your swapping switch/etc would work.


FWIW, if it has "intelligent identification" or "learn" wires, it's very common that simply stopping the motor, then connecting that jumper together will typically reverse the direction, then disconnect the jumper and now the new direction is reversed. So it *can* work as a kind of gear shift, but only if hte motor is completely motionless when doing it. If the prop is forcing the motor to spin at all, it may simply spin in the same direction. Might need a "brake" to stop the prop from doing that.

 

[ElectricXtment]

Hey amberwolf,

thx for the Reply!

jup, it is a gearless BLDC... I am not sure about the brand/model,
was a used one from craigslist, no OEM stickers or hints.

I also opened the motor casing and found this configuration inside.
Ylw=A Blu=B Grn=C and the ominous S=??? I dit not remove the PCB
to see where the traces of S go, but on the Oscilloscope it showed
flanks with approx. the same amplitude as the other Hall Sensors.

I also tried with/without the White S cable: No difference...

Here's a Pic:

NExt, for the above, this means it is a wrong phase/hall combo.

Well, why would this be so? It turns perfectly if I leave the reversing
switch alone(Even in Reverse!!!). Repeatably. The hiccup only happens
(Occasionally, 50%) after a direction swap.

About the "Intelligent identification:
Yes, I suppose this could work, but it would need something like 1~2sec
for the system to switch. And if the User by mistake pushes the throttle
more than one time...fail...and collision...meh.

I would soooo love to tell the controller to reverse over UART, that
would be so sweet...

PS: I played around with the Identification feature and it will change
the rotational direction flawlessly until I get bored... So I guess this
is another Indication that I got my wiring done correctly. Right?

 

[amberwolf]

I also opened the motor casing and found this configuration inside.
Ylw=A Blu=B Grn=C and the ominous S=??? I dit not remove the PCB
to see where the traces of S go, but on the Oscilloscope it showed
flanks with approx. the same amplitude as the other Hall Sensors.

I don't see anywhere that white wire goes to a separate sensor (there would have to be three more leads coming off the board going to another hall sensor). So it is probably just a parallel tap off one of the other sensors. If your scope is a dual trace, setup channel A on the white wire, and use B to see if it matches waveform and timing with one of the other three.

Well, why would this be so? It turns perfectly if I leave the reversing
switch alone(Even in Reverse!!!). Repeatably. The hiccup only happens
(Occasionally, 50%) after a direction swap.

That's probably because reverse doesn't always swap the same hall and phase wires. It may use a completely different combination than forward would. Why you get a hiccup and not always a problem, I don't know.

The basic key to finding a good reverse combo, once you have a working forward one, is to swap any two phase wires, then swap halls around until you find the good reverse combo (you may have to swap all three!).

Once you have a good reverse combo, you then need to setup your reversing switch to swap those two phases, and whichever halls (could be all three!) are necessary to recreate that correct reverse combo.

About the "Intelligent identification:
Yes, I suppose this could work, but it would need something like 1~2sec
for the system to switch. And if the User by mistake pushes the throttle
more than one time...fail...and collision...meh.

Well, it's the only way other than finding the right reverse combo I know of for a controller without a reverse function.

I would soooo love to tell the controller to reverse over UART, that
would be so sweet...

I don't know a way to do that. If you want to look at the various open-source firmware threads for some displays and controllers, you may find someone has discovered how to do this in the github / repositories linked from them, and this might help you see how this would work for your own (keeping in mind that there are no standards about protocols or functions, etc etc.). There are also some threads about controller/display protocols. I don't recall reading of any way to change directions on any controller via the display, but it doens't mean the ability isn't there--just that it hasnt' been used yet.


There may well be a pad on the board inside the controller that when grounded will reverse rotation, but they aren't marked in plain english--just cryptic abbreviations that arent' consistently used between cotnroller brands and models. You could try looking up other controller threads where controllers are opened up and pads tested and labelled, but no guarantee they'll match yours. Like this one
https://endless-sphere.com/forums/viewtopic.php?t=68501#p1037209
that says "E4" is reverse.

PS: I played around with the Identification feature and it will change
the rotational direction flawlessly until I get bored... So I guess this
is another Indication that I got my wiring done correctly. Right?

No. The point of the identification is that it doesnt' care what your wiring is--it figures it out *for you*. As long as you have power and ground to the halls, and all three hall signal wires from the motor are each hooked up to a corresponding wire on the controller, and all three phase wires from the motor are hooked up to a corresponding wire on the controller, it doesn't matter if the colors match or what order they're in, the IID will deal with it.

 

[QQ0702]

Hello,
I am currently working on a motorized wheelchair project, and I use a XunLida Xld 36/48 v 350 W controller, all the connections are correct and I use the HIGH EBS functions and also Intellegent identification to reverse the direction orientation, except that this reverse function works for the first 3 minutes and then does not work all at once. I wonder if there is someone who has already encountered this problem and if there is a solution for this kind of problem. thank you so much

 

[QQ0702]

Hello,
I am currently working on a motorized wheelchair project, and I use a XunLida Xld 36/48 v 350 W controller, all the connections are correct and I use the HIGH EBS functions and also Intellegent identification to reverse the direction orientation, except that this reverse function works for the first 3 minutes and then does not work all at once. I wonder if there is someone who has already encountered this problem and if there is a solution for this kind of problem. thank you so much

 

[QQ0702]

Hello,
I am currently working on a motorized wheelchair project, and I use a XunLida Xld 36/48 v 350 W controller, all the connections are correct and I use the HIGH EBS functions and also Intellegent identification to reverse the direction orientation, except that this reverse function works for the first 3 minutes and then does not work all at once. I wonder if there is someone who has already encountered this problem and if there is a solution for this kind of problem. thank you so much

 

[ElectricXtment]

all the connections are correct and I use the HIGH EBS functions and also Intellegent identification to reverse the direction orientation, except that this reverse function works for the first 3 minutes and then does not work all at once.

@QQ0702:
Would you mind explaining in detail how you got this working?
3 Minutes would work fine enough for me!

@amberwolf:
I'll check the signal of the "C" output and post it.

That's probably because reverse doesn't always swap the same hall and phase wires.

Well, I do. I always swap the same Hall-Signals. And many times it works...

There may well be a pad on the board inside the controller that when grounded will reverse rotation,

Thanks for the hint! I opened the controller and unfortunately the E4 Pin is not to be found...
The DS Pin makes the controller stop, nothing more...meeeh...

I also got a hint that inverting the hall signals a&c could help. I'll try this and
share the results.

Cheers!

 

[shaman]

Ok. So your problem is that you’re not using a CFOC or CFOC2 controller.

 

[QQ0702]

thank you for your answer, me it is with a 5V relay that I make the change of direction, I put the relay in the working position for 1 second then at rest, this function allows me to change the direction normally, but I explain , if I do this operation constantly, the motor changes direction of rotation, but if I do not reverse direction using the two green wires which effectively allows to change direction for 3 minutes straight, this functionality does not work more and I can no longer make a change of direction.

 

[ElectricXtment]

Could you post a simple schematic of your setup for better understanding?

 

[QQ0702]

hello, here are my connections and the different elements that I use

 

[QQ0702]

some one here have a datasheet of this XLD controlleer or any datasheet of this type of controllers ?