Kelly Controls KEB48400 weird error code?

[Jhondius]

Hi all

First time poster here.
I searched and read manuals etc. No solution yet.

I have a
Kelley controls keb 48400 controller
2kw 48v motor bldc bought used, new condition, chineese, no further info but lookin good

I measured the hall sensors of the motor and they react to rotating the mtor by hand.

The supply will be a lifepo4 unit 48v 30ah but it will be a while before it arrives.

Testing the setup with a switched 48v powersupply of 36watts.

The controller seems to work but all hooked up it does a led error code that is not in the manual.
I will post a video because it blinks weird.

Why do i think the controller is okay?
When i power i t up with nothing attached it blinks 4,2 hall sensor fault. (Correct: not attached)
I can also connect the rs232 port via usb to pc, start the software and it connects and shows the values.

But when i connect the hall sensors the 4.2 code changes into....well... red and green blinking that is no regualr code in the manual.

My suspicions:
It crashes and reboots all the time (there is a code for too frequent resets and this is not is however)
Or
I draws more the 36w my test supply can provide during some process and goes weird. But this is strange. The manual says it uses max 0.4a in rest and that is about half of what the supply can deliver.

But my hopes are that there are cracks on here that can help me.

Ill try to post the video next

 

[amberwolf]

i don't have direct answers to your questions, but i do have some questions that might help lead to answers, if they're relevant.

first, you say the controller seems to work. is it working in a sensorless mode to spin the motor, but just not with the halls connected? if not, how exactly does it work to spin the motor? (because i wouldn't call any controller "working" at all unless it will spin the motor under your control).

next:

I have a
Kelley controls keb 48400 controller
2kw 48v motor bldc bought used, new condition, chineese, no further info but lookin good

were both bought used, together? or is only the motor used, and the controller new from kelly?

if they were used together, is there any information on why the seller parted with them? if it's from a diy project, do they have a project page or thread somewhere on the internet? (sometimes the info they have is relevant to the problem being experienced...and sometimes people have a wierd problem with something that they get tired of, and sell the parts to make it someone else's problem, naturally without telling them about it).

one possible problem is that it is brain-damaged: some kelly models cannot be connected to the computer when the motor is rotating, at all, or they will be damaged (usualy bricked completely), and unfixable. kelly won't generally do anything to help when this happens, as they claim that their manuals clearly state not to do it...but it happens a lot, and the manuals are *not* clear about that, to many readers of them. sometimes such damaged controllers are resold by frustrated owners to get some money to buy a new controller from a different company that cares (or at least, doesnt' have this known huge bug in their software/hardware that they simply can't be bothered to fix)....sometimes these are resold for just parts, but i expect some of them try to get more money for them by not telling prospective buyers about the problem.


next:

on the power supply, is it a completely clean output, with no electrical noise? it's important because there wouldnt' be any from a battery, so the controller design may not take into account certain types of electrical noise that an smps might generate (evne though it is itself a similar type of converter and will be making it's own noise in operation). that might cause unexpected problems.

if you have a voltmeter on the psu's output, or an ammeter, or both, you may be able to see any momentary overload draw by the controller that drops the voltage below what the kelly expects.

 

[Jhondius]

Thanks amberwolf.

Ill try to answer in the same order.

first: no i have not seen the controller working with any motor.
I dont think there is a sensorless mode, have not found anything about that in the manual.
So you are right: it is not working, i only hope so. The things it _does_ do rs port and leds, indicate only that it is not bricked.

2nd
Yes bougt togetger, both used, but visually looking good.
The seller told me he buys chineese electrical boats, converts them to diesel, and sells the in africa.
He said, yes thats crazy, but the market exists and he needs to make a living. He was from this small town in holland where there are a lot of shipyards. He also had many units like this for sale. And tried to interest me for other electric boatwork.
I can see that it is removed from a boat. He told me he tried the the boat while it was still electric. So working i guess.
It is a bit of a gamble, but the price was good, and the storyand man seemed to check out.

BTW I have the kelly controls throttle switch too, but that seems to be dead because i measured it (ohms) and infiite resistance between all leads. Dont know it that means it could never have worked or just loose connection or something.

I know about not connecting the motor while connecting to the rs port. I do read manuals before i act learned that the hard way.
So it was not connected then. Also it seems not to be bricked.
BTW kelly has this section in their online FAQ that says:
How do i know if my controller works?
test1: does the green led burn continuous with no motor attached no matter what the red does? (yes)
test2: measure resistance between B- and ABC motor terminals, It should go rise to kOhm.
https://www.kellycontrollers.eu/kelly-faqs

----------
HOW TO CHECK IF CONTROLLER WAS DAMAGED ?
Firstly, you should see Green Light solid ON, Red Light flashing or off if the controller is powered.
Then, please measure how much is J2 pin7 or 5V out. You may measure the resistance of the controller, unpowered, from B+ to Phase A, B, C, and from Phase A, B, C to B- if it is a brushless motor controller, measure from B+ to M, and from M to B- if brushed motor controller.Please swap Red probe and Black probe on the terminals when you measure the resistance to avoid the discharge from the capacitors. You may remove the controller from the motor prior to measure the resistance. You should see the resistance gradually up to Kohm rating. Otherwise the controller got damaged.
--------

Well t tried this as i understand it, and there was infinite ohms between all. Not rising to kilo. But i am not shure i did the correctly.
Maybe someone with a known working controller can try?

3rd
I dont know about its ouput cleanness. It is a quality supply.
Measuring is a good point.
When i test the supplyś volts, nothing else connected it is straight 49V
When the controller is connected, but no motor, or with a motor, the voltage swings a lot. Between 15 and 47 at least but so fast that my meter does not quite follow. But in the same speed that the leds seem to react.

So yes: the voltage drops below what kelly expects as per manual, top p.5
"Controller supply voltage range, PWR, 18V to 90V"
https://www.kellycontrollers.eu/custom/kellycontroller/image/data/srattached/54ab127617eef727f60aca065e7bd4b4_KellyKEBUserManual.pdf

So it may be that the controller is just starting and quitting all the time. Explains the lights.
The manual says the controller uses <0.5mA so 24Watts.
But maybe the supply (switched) cannot keep up with fast demand.

I think ill have to wait for the battery. I have no other 48v source around.

 

[Jhondius]

video
https://youtu.be/r7EoySwuGrg

 

[DogDipstick]

video
https://youtu.be/r7EoySwuGrg

I think that is trying to power but not cause of lack of. My KEB takes a good 20w+ to power and that is just the power the system.. no load.

Get a stable power source and pls report back.

 

[amberwolf]

So you are right: it is not working, i only hope so. The things it _does_ do rs port and leds, indicate only that it is not bricked.

ok, i was just checking; sometimes i miss important info in a post.

Yes bougt togetger, both used, but visually looking good.

fwiw, a dead unit will usually look exactly like a working one, even internally, unless some part has dramatically self-disassembled, etc.

He told me he tried the the boat while it was still electric. So working i guess.
It is a bit of a gamble, but the price was good, and the storyand man seemed to check out.

it's likely that it is indeed a working unit. i just brought up the possibility because that sort of thing does happen.

BTW I have the kelly controls throttle switch too, but that seems to be dead because i measured it (ohms) and infiite resistance between all leads. Dont know it that means it could never have worked or just loose connection or something.

is the throttle a potentiometer (pot) throttle, or a hall-based throttle? if the latter, you won't get a typical resistance reading from it.

if it *also* has a switch in it that requires the thorttle be turned a certain amoutn before it clicks on, then this will disconnect at least one lead of the control unit inside from the outside world, so it would be open-circuit until this switch engages. (for a pot, it could be the power or signal leads, for a hall throttle it's probably going to be the signal lead).

test1: does the green led burn continuous with no motor attached no matter what the red does? (yes)

if that's what it's supposed to do, then you can clearly see in the vidoe that it is not doing that. the green led fades out periodically and comes back on. that probably means the system is losing power input; that the voltage applied to the battery terminals is dropping below the controller's minimum required voltage.

if that is the case, then the controller could be perfectly fine, but unable to stay on due to the power supply being unable to supply it's load.

When i test the supplyś volts, nothing else connected it is straight 49V
When the controller is connected, but no motor, or with a motor, the voltage swings a lot. Between 15 and 47 at least but so fast that my meter does not quite follow. But in the same speed that the leds seem to react.

see the above--it means the power supply can't maintain a steady output voltage because the load on it (current draw) is too high for it's rating.

I think ill have to wait for the battery. I have no other 48v source around.

that's probably your best bet then.

but...if the controller can take as little as 18v to operate correctly, then if you have access to a couple of common car "12v" batteries, you can put them in series for 24v (26-28v really, if they're fully charged). same thign if you ahve a couple of computer-system uninterruptible power supply systems around--they usually have 12v batteries in them for household units.

they woudln't last very long (minutes?) but you could also series a bunch of "d cell" flashlight batteries. at 1.5v nominal each, 16 of them would give you about the same "24v". they won't run the motor (they *might* start moving it just barely at low throttle settings), but you should be able to do all the other testing.

(the reason for starting with at least 24v is that when it sags in voltage from load spiking, it won't drop below the 18v minimum).

 

[Jhondius]

The battery arrived. Now waiting for the right connector. Then i will test the controller and motor.

 

[Jhondius]

Hi all

I connected the battery (48v 30ah lifepo4 from a escooter) to the kellycontrols keb48400 to the engine 48v/2kw
(And a 50a fuse)

And it WORKS!

So the weird led code was no code but not enough power in the 36w 48v psu to start.

Thanks for all help.

Next steps are wiring and connectors and packaging the switches pots batt etc in a nice and clean way. Also adding a turnigy 160a 0-60v wattmeter to see the batt condition and remaining charge. I have some connector q's in another thread.

After that some constructive work to get the engine in place.

Ill post any worthwhile info on the build here to help fellow builders in the future.


Greets