A one-electronic component solution to fix the drain of the One Z-series

[Filalapat]

Necessary elements :
- CRYDOM CMX100D10 Solid State Relay
- Heat Shrink Tubing
- Velcro adhesive tape
- Equipment for soldering, cutting and stripping the wires.

Warnings:
The electrical modification of certain wheel connections developed here is your own responsibility. I disclaim all liability for the warranty, and all kinds of problems that you would have with your wheel during or after this modification.
The modification detailed here does not modify the power wiring (XT60 connectors), and no high amplitude current flows in the circuits concerned here. This therefore greatly reduces the risk of an untimely cut off of the wheel.

Electrical characteristics measured on the battery communication connector.
In the lack of official technical information from Ninebot, I placed an oscilloscope to visualize the electrical signals that pass through the 2-pin connector.
On the two communication connector wires, respectively yellow and green, I chose (arbitrarily ...) the green as GND reference.
With the wheel started up, the observed signal revealed 20Hz signal with burst of digital frames, sometimes positive, sometimes negative with peak amplitude of about 3V, being renewed every 50ms, but most of the time between the frames, the signal remains at 0V. I did not try to decode them because this is not necessary in order to find a solution to this drain problem
When the wheel is in standby state, I observed a slight voltage between yellow and green wires (do not ask me how much, I forgot to note the value!).
If you disconnect the 2 pin connector, the voltage delivered by the battery on the electrodes of this 2-pin connector is 3.1VDC. (yellow wire, the ground being on the green wire). It is obviously the presence of this DC voltage which drain the battery over the days, through the input impedance of the motherboard circuits.

My idea is to stop the battery communication connection when the wheel is in standby state and restore it when the power is turned on. For that I will use a static relay. This is certainly not the cheapest solution which can be designed for that, but it has the huge advantage that I have already this component. This will be the only electronic device needed.
Thus, I use a CRYDOM CMX100D10 solid state relay. This model is largely oversized for my use here, I know. (picture 4)
The solid state relay switching electrodes (pins 1 & 2) will be used to cut the yellow wire to the 2-pin battery connector.
I chose to take the 5VDC supply voltage that power supply the bluetooth audio card to power also the both "coil" electrodes of the solid state relay (pins 3 & 4).

 

 

Step 1 - Open the right side of the wheel and disconnect the two XT60 connectors.
Step 2 - Open the left side of the wheel, and remove the plastic plate covering the electronic boards compartment.

Identification of highlighted wires: The yellow and green wires come from the 2-pin battery connector on the other side of the wheel (picture 1). The twisted cable (picture 2) brings the power supply to the audio card. It will be used to derive the supply voltage of this card.

Step 3 - Unplug the white 30-pin connector at the top of the motherboard. This is just recommended to avoid a risk of destruction of this by static electricity during the intervention. (ESD risk)
You will then see this: (picture 5). For the same reason, unplug the motherboard-side connector of the twisted cord highlighted in the blue circle (picture 2).

Step 4 - Cut the yellow wire that appears on the picture 1 towards the middle. You should have something like picture 3.
Step 5 - Disconnect the yellow/red/black 3-pin connector connected to the audio card, then untwist the wires on 2 inches and cut the red and black wires one inch from the connector. (picture 6)

Step 6 - Realization of the derivation on each of the red and black wires : Use a piece of black wire about 4 inch (I used a gray one on the photo). Solder the three black wires together. Insert a heat-shrink sleeve before solder the three black wires together. (picture 7)
Step 7 - Do the same with the red wire using a piece of red wire about 4 inch long. You then see something similar to the picture 8.
Step 8 - Glue the piece of Velcro tape to the static relay as shown in picture 9, and the other one on the plastic shell.

Step 9 - Solder the free end of the red wire that you have prepared on the terminal 3 of the relay, and do the same with the end of the black wire (mine is gray on the photo...) on the terminal 4. Remember to put on heat-shrink sleeves before soldering.
Step 10 - Now it is a matter of extending both ends of the wire cutted in step 4. If you have yellow wire use it otherwise any other color will do the trick ... Strip the end of the yellow wire connected to the white connector 30-pin, and solder a 4 inch long wire. Place two heat-shrink sleeves and then solder after shortening to the required length the free end of the wire to the relay terminal 2.
Step 11 - Do the same by connecting the other cutted yellow wire end from the rubber cable clamp to the relay terminal 1.
Step 12 - Place heat-shrink tubing and heat. (picture 10)
 
Don't do like me: don't forget to reconnect the 30-pin connector before power up the wheel ... OOOOOO
Reconnect both ends of the audio card cable as well.
Close the wheel without forgetting to reconnect the XT60 connectors of the battery.

First report of test…
After 50 hours, the average voltage drop is 0,025V
The voltage loss is  0,500mV/h. 50 hours is too short to obtain a better precision, but it is the value I expected !

This fix works fine.

[stephen]

Jeeez nice solution i think I'll just ride it daily 😊 

I don't know how people have all this knowledge hats off to you my freind 👍

[Ch.Eng.62]

Hello @FilalapatThanks for the solution. Probably I will make the mod because when I am on contract (around four months), I make my wife to check regularly the battery charge. It is difficult to find the CRYDOM CMX100D10 Solid relay. Can you give the specs to look for similar please 

[Ch.Eng.62]

@Filalapat Is this suitable https://www.amazon.com/Aideepen-G3MB-202P-DC-AC-Solid-Module/dp/B01JCPPBI4/ref=lp_6374820011_1_17?s=industrial&ie=UTF8&qid=1558391933&sr=1-17

[steven55tw]

@Filalapat Good job! Thanks for your information.

[Planemo]

Outstanding! A simple solution to a big problem for some. Thanks for sharing your skills!

Edit: Sorry to be a pain but I can't get the Crydom or Amazon relays in the UK. Would you be kind enough to select the correct relay from this page please? The filters really help but I have no idea where to start!

https://www.mouser.co.uk/Search/Refine?N=18383526

[Lefteris]

That is a great fix, difficulty level 9/10, but still a fix!

[Filalapat]

The main characteristics for this device are :

Coil voltage : 3 à 10V,  DC voltage
Output voltage : 100Vmax, DC voltage
Output max current : 10A

the manufacturer's datasheet is downloadable here :

http://www.crydom.com/en/products/catalog/cmx-series-dc-pcb-mount.pdf

This device is highly oversized here, max current is really in the mA range, output voltage don't exceed here 6V peak to peak... 

This device is available on farnell, rs components, ebay... in europe, but you should be able to find these devices worldwide.

@Ch.Eng.62
The G3MB-202P models seems to be made for AC mains voltage output. Probably they integrates a triac stage output, this will not be compatible with this application. 
Output stage should be made for DC voltages.

 

links to find the devices...

https://www.ebay.fr/itm/1-X-CMX100D10-Relay-solid-state-Ucntrl-3A-10VDC-10A-0A-100VDC-THT-SIP/112616804660

https://fr.farnell.com/sensata-crydom/cmx100d10/relais-statique-10a-100v-sip/dp/1779773?st=cmx100d10

@Planemo
On your mouser webpage link :
choose manufacturer : sensata
choose control voltage range : 3VDC to 10VDC
choose load voltage rating : 100VDC
and apply filters

and maybe you see this :

mouser no : 558-CMX100D6
https://www.mouser.co.uk/ProductDetail/Crydom/CMX100D6?qs=sGAEpiMZZMsUriz2CNI3E8GqrC4g63wzREBBmyTRWmc%3D

mouser no : 558-CMX100D10
https://www.mouser.co.uk/ProductDetail/Crydom/CMX100D10?qs=sGAEpiMZZMsUriz2CNI3ExU7VxMWr6ZeCmO7dwd6xcU%3D
 

[Planemo]

Thanks Filalapat, perfect! Quite pricey at £30, surprised at that, thought they might be a bit cheaper but hey-ho.

This is a real easy mod guys, couple of wires to be cut and a bit of soldering. In fact I think it could be even easier - rather than cutting the wires to the Bluetooth module could you not just splice them? Just strip back a little of the sheath and solder on the news wire to them.

I'm going to do this at some point soon - I have not left my wheel for longer than 2 weeks so far but in winter it may sit for a few months so doing this just gives me peace of mind and I can forget about it

[Ch.Eng.62]

@Filalapat Did you charge the battery after the mod? I am asking because during the charging the wheel is switched off and, if the wheel is modded, there is not communication between the mainboard and battery. Is this a problem? 

[Rotciv]

29 minutes ago, Ch.Eng.62 said:

@Filalapat Did you charge the battery after the mod? I am asking because during the charging the wheel is switched off and, if the wheel is modded, there is not communication between the mainboard and battery. Is this a problem? 

This is what concerns me about this mod.

I know from experience that if the battery pack gets too hot, it will not charge. Without knowing exactly how the battery protection logic works with the BMS and main control board, any fault with the battery while charging might escalate into a bad situation.

Perhaps improve the mod by ORing system standby and charging. The two leads from the charging port connects directly to two of the four leads at each battery pack.

[Fremalea]

Thanks for the tip! But I am a novice in electronics so I will never dare to do that!

[Filalapat]

As I said in my upper post in this topic, digital data communication between the battery and the motherboard is only maintained in power-up state. No data transfer is required between the battery and the motherboard during charging, and as soon as charging starts, if the wheel was powered up it will turn off automatically.
Without applying the fix, any owner of Z can also verify that the charging process is OK when the connection of the 2-pin connector is broken: just unplug this connector and charge...
Here is my wheel charging at 6A, after a 40km ride. (with a 3-way of charge custom charge doctor)

Concerning the heating of the batteries, there is nothing in this patch which can cause an abnormal heating. There was no change in the charging circuit built into the wheel. Like any other type of wheel, it is the charger that regulates the charge and limits the current during most of the charging process, and then limits the voltage at the end of this process. (CC/CV regulation)
Regarding the fast charge, I will not personally charge beyond 6A. The following post has already shown that problems have appeared for currents of about 7A.

 

[Ch.Eng.62]

Thanks for the clarification. 

I am owner of Z10, but I am not at home. Will be at home after 5 or 6 days. My reliever is already onboard and my contract is about to finish. 

I know that for  the previous Ninebot wheels, the charger is connected directly to the battery, but didn't know how it is in the Z10. 

In the previous models there was not communication cables at all. 

[boat dog]

Good Evening Mr F

excellent idea and very clearly laid out instructions, thank you very much.  This could be achieved using a locking catch waterproof switch? Located somewhere discrete so as not to be accidentally activated?

As you are clearly knowledgeable, I would like to ask you about my z10 batterie experiences;  I just bought a bust z10, stupidly yes as it looks likely an expensive door stop and I cant afford it, but it is basically new and such a lovely chunky design. Anyway I have opened it up to find the usual burnt connections (x2) on the powerboard, however as it all still worked (except for balancing of course) I thought I would fully charge it at the very least. I read many entries on this super forum about charging and left it overnight, so some 8 hours after the charger block light went green, in order to allow for cell balancing.

In the morning the wheel was off, no charging bars and the charger block light was but a dim dim red, hmm looked fishy so as the wheel was open I decided to take direct voltage readings from both batteries...only to see over 70v each on each, yikes. I immediately removed the pack and placed it in plain sight in case it decided to ignite and over the day (today) I have been checking the voltages.  the readings have been all over the place from a healthy 58.4 to still 65-70v, this is mad for a 14s pack, and would surely would indicate serious issues with the BMS?  I am going to find another meter just in case, but this one has been very reliable.

I wonder if the z10 issues are all down to a BMS fault?  The high voltages would at the very least test the system and components, 8-12v over surely is significant and this would quickly damage the cells....or am I getting this wrong?

I presume that the operating system needs to get info from the BMS, the yellow/green wires, so I cannot just fit 2 x proven 14s BMS boards?  Likewise I would imagine hacking the ninebot code to program my own BMS would be a nightmare if even possible?

 

Any thoughts appreciatted

james