Ninebot One e+ broken after 1 1/2 weeks. Any troubleshooters here?

[Cranium]

This thing didn't last long.  About 30 miles.  

I was riding about 10 minutes this evening and as I went down my driveway, it shut down.  I tumbled but wasn't going fast so didn't hurt.  But now the ninebot won't start up.  I don't know if it was related to the thermal shutdown or not.  I bought this directly from China so warranty support is out.

Symptoms and things I checked:

• Motor doesn't turn easily 
• Unplugging the motor allows it to turn easily
• Plugging the motor back in and turning on the unicycle on its side will start it up beeping (motor is disengaged).  As I straighten it up, it turns off at the point where it would engage the motor
• Battery voltage is fine at 59.24V
• I see no sign of magic smoke on the controller board

I believe this indicates it is a short somewhere across the windings but I'm not sure where to even start looking at what the issue is.  Anyone know the pinout for the motor?

Any suggestions to troubleshoot would be greatly appreciated.

 @esaj, have you worked on this before?

 

[Cranium]

Pics of the connector

[Viktiga]

Why wouldn't the unit be under warranty as long as it's genuine? I would contact Ninebot and check.

I have a thread here:

and here:

 

 

[Cranium]

1 hour ago, Viktiga said:

Why wouldn't the unit be under warranty as long as it's genuine? I would contact Ninebot and check.

Good suggestion.  Certainly wouldn't hurt to reach out.  I took pics of the serial number on the foot rest, on the box and a screenshot from my phone (it still will connect).  I also sent them the re-seller information order date and tracking information.  Hopefully I'll hear positive news but in the meantime would still like to isolate the issue here.

[Viktiga]

9 minutes ago, Cranium said:

Good suggestion.  Certainly wouldn't hurt to reach out.  I took pics of the serial number on the foot rest, on the box and a screenshot from my phone (it still will connect).  I also sent them the re-seller information order date and tracking information.  Hopefully I'll hear positive news but in the meantime would still like to isolate the issue here.

Sounds like good idea. Good luck.

[esaj]

4 hours ago, Cranium said:

This thing didn't last long.  About 30 miles.  

I was riding about 10 minutes this evening and as I went down my driveway, it shut down.  I tumbled but wasn't going fast so didn't hurt.  But now the ninebot won't start up.  I don't know if it was related to the thermal shutdown or not.  I bought this directly from China so warranty support is out.

Symptoms and things I checked:

• Motor doesn't turn easily 
• Unplugging the motor allows it to turn easily
• Plugging the motor back in and turning on the unicycle on its side will start it up beeping (motor is disengaged).  As I straighten it up, it turns off at the point where it would engage the motor
• Battery voltage is fine at 59.24V
• I see no sign of magic smoke on the controller board

I believe this indicates it is a short somewhere across the windings but I'm not sure where to even start looking at what the issue is.  Anyone know the pinout for the motor?

Any suggestions to troubleshoot would be greatly appreciated.

 @esaj, have you worked on this before?

Don't have a Ninebot, but the motor not turning easily when it's plugged in would suggest a shorted motor (based on my reading about 3-phased BLDC-motors lately), which probably happens at the drive half-bridges. My best guess is that one or more of the low-side mosfets have damaged and cause a short between a phase (or phases) and the ground. Could be wrong though. Can you measure the mosfets / bridge? If the problem's there, you probably need to replace the entire mainboard anyway, though.

As a somewhat, but not really related sidenote, last weekend I picked up my old electronics books and started to study electronics again. By the time I hit common emitter/collector/base-connections (basic amplifiers), I realised how badly I've forgotten everything, as I couldn't even at first tell where the amplified signal comes out (but in my defense, the pictures weren't drawn very clearly anyway), or what the voltages are at different parts of the circuits   Back to basics, better revisit calculating even more basic circuits...

Not that I probably will even need such a circuit as above, but I should (and probably did a long time ago) know this stuff  

[DangerousDick]

A short across two of the three wires, or even all three, will cause the motor to sieze up.  Might be worth measuring across the motor wires themselves to make sure the windings are all a similar resistance to each other - at least that would suggest the motor is OK?

Most likely the board somewhere - looks the same as the pic I took of mine yesterday.  Bit 'off topic' but can you tell me if it's an N10 or N20 model (first characters of the serial number)?  Also, just above the red wire, is there a 'version' number on the PCB?

Cheers, Richard.

Would a break in one of the three motor wires have a similar effect? 

If you want me to measure anything from the board give me a shout, I've got my E+ with me at work today and it's sitting next to the desk just crying out for some attention

[Cranium]

5 hours ago, DangerousDick said:

A short across two of the three wires, or even all three, will cause the motor to sieze up.  Might be worth measuring across the motor wires themselves to make sure the windings are all a similar resistance to each other - at least that would suggest the motor is OK?

Most likely the board somewhere - looks the same as the pic I took of mine yesterday.  Bit 'off topic' but can you tell me if it's an N10 or N20 model (first characters of the serial number)?  Also, just above the red wire, is there a 'version' number on the PCB?

Cheers, Richard.

Would a break in one of the three motor wires have a similar effect? 

If you want me to measure anything from the board give me a shout, I've got my E+ with me at work today and it's sitting next to the desk just crying out for some attention

I was thinking the same thing for measuring the resistance across the coils in the motor first to see if there is a short.  To properly test them, I would need a megger but if it is an obvious short, I have a pretty good LCR meter that should work well enough.  But to do this, I need to find out the pinout of the motor connector.  If I can't find the pinout, I will have to take the next step and disassemble the motor which will be a real pain.

I have the N20 model running firmware V1.2.7.  Total mileage indicated is 51.6 miles.

I wish I was able to test today but am at work until tonight.  I will also try, as @esaj suggests, to test the mosfets for a short although this will be a bit hard to accurately test them installed.  But maybe I can get lucky and find a short between the source and drain and then replace just the one mosfet.  I have a vacuum desoldering station and digital soldering iron so this wouldn't be too hard to do.  I'm also concerned about the coating they put on the board to make it water resistant.  Anyone know what this is?  I would want to re-coat anything I have to remove.

 

[Chriull]

13 minutes ago, Cranium said:

I was thinking the same thing for measuring the resistance across the coils in the motor first to see if there is a short.  To properly test them, I would need a megger but if it is an obvious short, I have a pretty good LCR meter that should work well enough.  But to do this, I need to find out the pinout of the motor connector.  If I can't find the pinout, I will have to take the next step and disassemble the motor which will be a real pain.

The yellow, green and blue wire on the mainboard from your first picture should be the wires which feed the coils of the motor. So you could have a good chance, that these 3 wires go to the three "fat" pins on the connector...

The coils are connected like an Y - so you always can measure 2 coils at once in series.(Shown in this schematics: http://www.uaelec.com/English%20express%20release/BLDC.htm ) Additionally since there are more than three coils in the motor, there are always some in parallel. (Like shown here for 6 coils: http://eletechnic.com/technical_en.htm )

13 minutes ago, Cranium said:

I wish I was able to test today but am at work until tonight.  I will also try, as @esaj suggests, to test the mosfets for a short although this will be a bit hard to accurately test them installed.  But maybe I can get lucky and find a short between the source and drain and then replace just the one mosfet.

If one of them causes a shortcut, it should be possible to measure it between these yellow/green/blue wire and ground/Vcc (as seen in the first schematic i linked above). Turned off, they all should have a high resistance. Vcc should be the red wire from the battery.

Maybe you have to look to measure with the right polarity (The positive voltage from the Ohm meter to drain and ground to the source)

 

[Cranium]

9 hours ago, esaj said:

As a somewhat, but not really related sidenote, last weekend I picked up my old electronics books and started to study electronics again. By the time I hit common emitter/collector/base-connections (basic amplifiers), I realised how badly I've forgotten everything, as I couldn't even at first tell where the amplified signal comes out (but in my defense, the pictures weren't drawn very clearly anyway), or what the voltages are at different parts of the circuits   Back to basics, better revisit calculating even more basic circuits...

You are beyond my expertise.  I have learned what I have needed to get projects done.  I haven't had much formal electronics training.  It has always been just a hobby.  I know enough to recognize the 12V amp in the diagram with DC filters but not enough to be able to calculate the cutoff frequency for it or to build one myself without referring to someone else's design.  

I did want to say though that Finnish is an odd looking language.  määräävät  <-- Way too many ä in that word!  LOL

[esaj]

1 hour ago, Cranium said:

You are beyond my expertise.  I have learned what I have needed to get projects done.  I haven't had much formal electronics training.  It has always been just a hobby.  I know enough to recognize the 12V amp in the diagram with DC filters but not enough to be able to calculate the cutoff frequency for it or to build one myself without referring to someone else's design.  

Well, nowadays I can at best mostly just twiddle with the numbers a bit using "ideal" components, but that's about it... I did have at least some (more or less) formal training in electronics/electricity, but that was mostly sometime between 2000 and 2002, if memory serves, plus half of it was just AC-stuff. And since I never actually had to use what I did learn anywhere, the little I knew has been forgotten a long time ago, so at times this feels like starting from zero again   Calculating some values from an existing circuit is one thing (if I can even get those right ), knowing how to design one is a whole another matter, and something beyond my grasp, at least in most cases. Probably I'll be also referring a lot to other peoples' designs and books, once I get all the electronics stuff I ordered, and burn at least some components due to mistakes  

 

Quote

I did want to say though that Finnish is an odd looking language.  määräävät  <-- Way too many ä in that word!  LOL

Yeah, I think the closest "relatives" of Finnish are something like Estonian and Hungarian... Germanic languages are way different. On top of lots of scandinavian letters (ö, ä), we also have funny rules for word compounds (if that's even the right term), that can lead to very long words, like "lentokonesuihkuturbiiniapumekaanikkoaliupseerioppilas"   (which translates to something like "airplane jet turbine assistant mechanic non-commissioned officer student", although I don't know if it's an actual title in use anywhere ).

EDIT: And let's not forget the inflections (or is it conjugations?):  http://languagecatalyst.com/wordpress2014/wp-content/uploads/2014/04/Finnish-comic.png 

[OliverH]

Firmware 1.2.7? Upgrade to the lastest. It was reported that the 1.2.7 can kill the mosfets as well as it have sudden power cuts. Maybe it's to late and you need a new control board.

[Cranium]

21 minutes ago, OliverH said:

Firmware 1.2.7? Upgrade to the lastest. It was reported that the 1.2.7 can kill the mosfets as well as it have sudden power cuts. Maybe it's to late and you need a new control board.

1.2.7 is the latest firmware. Are you mistaking 1.2.6 with 1.2.7?  The previous version had known issues but I haven't heard of anything with the latest version I am on. 

[OliverH]

You're right. We still have 1.2.2 or so installed. Have a look in shorts of the 3 phases to the motor.

[FORWARD california]

Hi Cranium,

We have contacted Ninebot inc. with info about the issue you are having with your new Ninebot One model E+. We will do our best to help you get the issue resolved and get your Bot repaired.  If you dont mind me asking who did you purchase the Bot from?

[Cranium]

32 minutes ago, FORWARD california said:

Hi Cranium,

We have contacted Ninebot inc. with info about the issue you are having with your new Ninebot One model E+. We will do our best to help you get the issue resolved and get your Bot repaired.  If you dont mind me asking who did you purchase the Bot from?

Wow!  Thank you so much for stepping in and helping!    

Being a distributor for Ninebot, I'm hoping you will have better luck than I can get with a response.  

Order information
Order Date: October 07, 2015
Delivery Date: November 12, 2015

My AliExpress Order Number: 69764074976285. 
Store Name: Star Stores
Contact Name: Mr. Xing Zhou
Street Address: Bantian
City: Shenzhen, China 418000
Contact URL:  http://www.aliexpress.com/store/contactinfo/1920547.html

Here are a couple of additional screenshots that may help as well:

Again, I greatly appreciate any assistance you can provide.  Especially since I didn't buy this from you.  

 

[FORWARD california]

Ninebot inc. just requested the serial # of your Bot to proceed. 

 

You beat me..lol. I will get this info over to Ninebot and keep you updated.

[Cranium]

5 minutes ago, FORWARD california said:

Ninebot inc. just requested the serial # of your Bot to proceed. 

 

You beat me..lol. I will get this info over to Ninebot and keep you updated.

Sweet!  :)

[FORWARD california]

Hi Cranium,

Ninebot inc. says that the issue is the control board.  We can offer you 2 options: 1. We can have a new control board mailed to you and you do the repair or 2. You can mail us your Bot and we can repair it and ship it back to you.  Let me know what works best for you. We want to get you back on your Bot asap.

- Team FORWARD cali

[Cranium]

3 minutes ago, FORWARD california said:

Hi Cranium,

Ninebot inc. says that the issue is the control board.  We can offer you 2 options: 1. We can have a new control board mailed to you and you do the repair or 2. You can mail us your Bot and we can repair it and ship it back to you.  Let me know what works best for you. We want to get you back on your Bot asap.

- Team FORWARD cali

That is great news!  I would be more than willing to replace the board myself.  

I would also be more than willing to pay for some expedited shipping if this is available.  I want to get back to riding ASAP!  

Thank you very much!  Going to PM you to take this offline to handle details...

[JFP9]

Two big thumbs up to FORWARD CA for stepping up! 

thanks for being one of the good guys.

Cheers,

Jim

[Cranium]

I received my new board already.  Only ONE WEEK to get the replacement board!  Impressive!  I now have to wait for an unlock code from Ninebot to activate the board.  The new board is on firmware version 1.2.2 and I plan on upgrading to version 1.3.0 once it is unlocked.

There are some differences between the boards.  Here is a list of differences:

• Original board is version 1.31.  Replacement board is version 1.3
• Original board is on a Green PCB.  Replacement board is on a Black PCB
• Original board has 100N8F6, 80V, 100A, 176W MOSFETS.  Replacement board has 75NF75 75V, 80A, 300W MOSFETS
  • Less voltage and amperage on the replacement but higher power dissipation capabilities.  And both are running fairly close to the voltage limits.  
• Original board has a N20 serial number.  Replacement board has a N10 serial number.
• Replacement board has a socket fitted PCB in the center (not sure what the chipsets are for).  Original board has these surface mounted on the main PCB.
• Original board was sprayed with a waterproof coating (they didn't do a very good job on the lower MOSFETS attached to the back of the PCB).  Replacement board looks like it was dipped in a water proofing and has better coverage and looks better

Finally, in the 3 phase driver, it is interesting to note that there are 12 MOSFETS.  There are 6 on the front that you see in the picture and there are another 6 in parallel to them mounted to the backside of the PCB.

 

Version 1.31 on the left and Version 1.3 on the right

The other 6 MOSFETS on the back side of the PCB

[esaj]

Didn't know that Ninebot also has 12 mosfets (so far I've known that at least most IPS's do). I don't think the voltage should ever get to 75V (considering the max battery voltage is something like 63-63.75V, as Ninebots use 15S-batteries)? There can be very short-lived inductive spikes of several hundred volts from the motor, but there should be something like transient voltage suppressor diodes somewhere (inside the motor?) that eat these... I actually ordered 25 of those 75NF75's from Aliexpress a couple of weeks ago for 6.10€ (6.50 USD) with free shipping, so not that expensive, maybe you could get some and try replacing the mosfets in your old board to see if it could still work.

Pure guess, but the green PCB on the replacement could be the gyro, although it's at a bit funny angle, but based on purely size, it looks similar as the MPU6050's I ordered in break-out boards:

The blue board on the bottom right of both mainboards looks like Bluetooth serial module (here attached to a break-out board, but if you compare the "lower" part):

 

[Cranium]

1 hour ago, esaj said:

maybe you could get some and try replacing the mosfets in your old board to see if it could still work.

This was my thought exactly.  The issue I'm going to have with that is that the MOSFETs are attached using thermal adhesive.  I'm not sure what the best way to get them off the heatsink other than prying them off.  

 

1 hour ago, esaj said:

the green PCB on the replacement could be the gyro

Good thinking.  You are probably right.  I was going to start scrapping off the protective film over the ICs to try to start identifying them.  Hopefully they haven't scraped off the marks on the other ICs.  I think it would be good to have all the components mapped out.  For the new MOSFETs, I had to use a stereo microscope to be able to see the numbers on them.  

[esaj]

2 minutes ago, Cranium said:

This was my thought exactly.  The issue I'm going to have with that is that the MOSFETs are attached using thermal adhesive.  I'm not sure what the best way to get them off the heatsink other than prying them off.  

Are you sure it's thermal adhesive? Could be just thermal pads like these: http://www.aliexpress.com/item/10mm-10mm-1mm-100-pcs-Thermal-Pad-GPU-CPU-Heatsink-Cooling-Conductive-Silicone-Pad-high-quality/32259809537.html  

The Firewheel mainboard had similar stuff like that between the heatsink and the surface-mounted mosfets (these: http://cdn-reichelt.de/documents/datenblatt/A100/DS_IRF7759.pdf), it wasn't adhesive, but just held in place by the bolt of the heatsink going through the PCB (might not be very efficient cooling, I think). As it looks like the Ninebot mosfets are also screwed to the heatsink, I doubt they'd need to use adhesive (but they still could have). Replacing the old thermal stuff with something like those pads (it's also available as bigger mats that can be cut) could work, as long as it conducts heat good but not electricity. I originally thought I'd replace the pad in the Firewheel with just thermal paste, but learned that at least in most non-SMD mosfets the metallic back-plate part is connected to one of the pins, so I could have connected some pins of the mosfets (gates?) together...  And once I took the heatsink off, I could see getting the paste "right" wouldn't have been easy anyway.