Problem with KU123 Controller

[Deleted member 4366]

We fitted a KU123 controller to Saneagle's bike. It works OK except it cuts our at 35.5v when his battery is only half empty, so he only gets 10aH out of his 20aH 36v battery.

35.5v is the LVC for 48v. My memory says that the pair of single red wires is a pair of jumpers to switch between 48v and 36v, but I can't find any reference to this unless there was a label on the actual wires, which have been removed. Can anyone confirm this and which way they should be for 36v?

We've tried connected and disconnected directly on the pcb, but it still cuts out at 35.5v, so if you have one of these controllers and your range is a bit low, you need to check for this.

The bike works perfectly with a 36v KU93, which will let thebattery give its full 20aH.

 

[wurly]

What pair of red wires are these Dav8h? if there is a jumper wire i haven't found one. I have a 48V KU123 controller, the only red wires are the main battery and the on/off wire. It might be worth looking at the status of the red led. 8 flashes means lvc i believe, in which case it's a 48volt like you say, not sure if you can change it by removing/adding a jumper wire.

 

[john h]

i have a ku 123 it says on the front of controler 36v connect the two red [short ones] 48V DISCONNECT THOSE TWO WIRES.

ALSO SAYS LVC 27.5/35.75 V

 

[wurly]

Mmmm, i'd be interested as to where those two red wires go. They must select or bypass a resistor on the undervolt circuit. Not overly important to me but, if either of you get a chance can you take a photo of where on the pcb the two red wires terminate?

 

[Deleted member 4366]

OK. Thanks for the replies. We've got it sorted now. I'll post some photos later. There's two pads in the corner of the board marked "XZ" between the voltage splitter resistors for the LVC detection. When the pads are connected, they bring another resistor in parallel to one of them to change the ratio by 33% for 48v. So connected for 36v and disconnected for 48v. When we got the controller, there were two thin red wires run externally to connect/disconnect.

I had soldered the pads to join them, but it didn't work. maybe a track was cracked when I removed the red wires. When I re-soldered them, it worked OK.

This is handy to know because you can set the LVC at any value you want by changing/ adding a resistor. The default values are 1K2 (1201) and 15k (1502) for 48v and then a 43k (4302) is brought in parallel with the 15k to take it down to about 11k for 36v. LVC detection by the main processor happens then at about 2.86v.

 

[Deleted member 4366]

Here's the photo. The resistors are top right:

 

[amigafan2003]

Cool - the 48v mode with a cut off at 35.5v would be ideal for 10s lipo.

 

[Old_Dave]

It seems that not all KU123's are equal, I've just got a KU90 that different from other KU90's, so maybe the designation relates to the watts and number of FET's rather than an an exact design / pcb


Sent from my iPhone using Tapatalk

 

[Deleted member 4366]

It seems that every time I get a controller from BMSBattery there's something slightly different.

 

[wurly]

Although, i suspect the pcb/componants fitted and software are identical.It appears to be the wires and connectors they attach seem to be different, but with the collective knowledge here i reckon' we can get them to work as we want :0)
Its a shame we can't get into the software and change things.

 

[Old_Dave]

Latest Ku90 from BMS spot the differences, lol





previous version posted by Scottyf in another thread was

 

[wurly]

Ok, i concede. My KU123 is the same as Daves but i can see now that PCB designs have probably changed and you never know what you'll get. I'll check my pc first before replying on my phone :0)
I can see my KU93 is very different to the KU90 shown in your photos but what that actually means as regards to their features i don't know? probably different software as well.........

 

[Deleted member 4366]

The older KU60, KU90 and KU120 were different. I don't think that they can handle high speed motors (commutation).The KU120 used to have the self-learning procedue while as the KU123 is automatic.

 

[Jeremy]

These controllers, together with those badged as Hua Tong and Greentime (amongst other names) are made by Wuxi Xichen Micro-Electronic Co (???-???·???) and come in a wide variety of different configurations. The "KU" labelling is just a nominal label that the retailer uses and doesn't mean much in terms of what you might actually get. If you buy a Greentime or Hua Tong you'll get a variation on the same controller.

The annoying thing is that these are all programmable by the vendor, just like the popular Xiechang controllers, but unlike the Xiechang controllers we've not been able to get hold of the programming protocol.

 

[Old_Dave]

I think (well I have 2 that do) that a common cpu in these boards is the STM 8S series, I've seen it rumoured that there is a 'motor development kit' with free software from ST....

But cos I'm not at the pc I'm having a spot of bother finding out if this kit is for the CPU off of the board or if it can be programmed after assembly.

But here's one link on the manufacturers site that could be interesting though

http://www.st.com/internet/evalboard/product/122903.jsp

Havn't found the software yet.


Sent from my iPhone using Tapatalk

 

[Jeremy]

That's curious, because the normal CPU in these controllers is the X8M06, an in-house design from Wuxi Xichen Micro-Electronic Co. I spent a few hours translating from Chinese to try and glean more about this ucontroller and concluded that internally it seemed to be similar to the XCKJ8B116A chip used in the Xiechang controllers.

I strongly suspect that the programming interface is similar, and uses TTL level RS232 to download a set of values into the controllers flash eeprom for things like current limits, speed settings, etc. We know that some vendors can programme these controllers, in a similar way to the way the Xiechang controllers can be programmed, but so far no one has been able to obtain the protocol. We were lucky with the Xiechang controllers, in that Xiechang released the programming software on their website (albeit in Chinese) and someone decompiled it (it was written in a 10 year old version of Visual Basic!) and managed to change the text fields to semi-English. Since then, some others have sniffed the interface during programming and managed to decode the protocol, so there is a now a pretty easy to use open source bit of programming software, XPD, available.

If your KU series controller is using something other than the Wuxi Xichen Micro-Electronic Co X8M06 then that seems to indicate that the supplier may have changed, which is interesting.