IPS Zero: 30A Fuse blows

[Rol]

Hi,

I opened and my IPS ZERO (IPS141 ZERO-340WH) to change the wheel.

Now, when I switch on the device, the wheel starts to shake and the 30A fuse blows. I changed the fuse inside the controller many times, tried to reset it with the IPSUtil but without success.

Now I left the wires to the motor disconnected (to protect the fuse) and one can see on the app, that the controller tries to give high enegery to the motor. After some seconds the device starts to beep and the LEDs are blinking in red. The App shows the error message 0020: Idleing Protection.

Tony Lee wrote me, that the IPS stopped the production of the IPS ZERO and no spare parts are available.

Any Ideas what went wrong, oder how to repair it?

Greetings from Germany

Roland

Edited June 23, 2018 by Rol

[Keith]

55 minutes ago, Rol said:

I opened and my IPS ZERO (IPS141 ZERO-340WH) to change the wheel.

 Now, when I switch on the device, the wheel starts to shake and the 30A fuse blows.

There is a very big bit of what you did missing between these two lines surely?

You are saying you opened it up and now it does not work. Is that all you did? You did not disconnect a single wire?

What it sounds like you have done is either plugged in the sensor wires (the thin wires going to the motor) in the wrong order or the main wires. By the look of it, that may not be possible on this wheel as the thin wires are on a (polarised?) plug and main power wires colour coded  (but are you certain it was blue to blue, yellow to yellow etc it does not always follow!) 

If this is not the case then I think it is possible a wire may have broken when it was disturbed (or it might just be that plug not making proper contact?) it could be any one of those thin or thick motor wires being disconnected or swapped could, I think, cause those symptoms. 

Basically I am making the above assumptions,

• Firstly on the basis that the odds that an electrical component should pick just when you were working on it to die are VERY small (unless you handled the main board and did static electricity damage - unlikely?)
• Secondly, The thin wires go to magnetic sensors (hall-effect transistors) which detect the  position of the magnets in the wheel in order to decide which coils of the motor to energise, the control board pumps current down one set of coils (two of the three thick motor wires) until the sensors detect the movement and switch to the next set of coils (a different pair of the three thick wires) and  so on. If one of those thin wires were broken the motor would start to move and never see the movement so it would keep increasing power to the first coils. If the main wires got swapped it would try to run backwards and the sensors would get very confused, if one thick wire was broken the wheel cannot turn so it can over current the other two coils trying to get it to turn.

[The Fat Unicyclist]

The wheel wasn't accidentally put in the wrong way around?

[Rol]

47 minutes ago, Keith said:

There is a very big bit of what you did missing between these two lines surely?

You are saying you opened it up and now it does not work. Is that all you did? You did not disconnect a single wire?

I disconnected all the wires and plugs and assambled it in the right order - I hope so.

47 minutes ago, Keith said:

What it sounds like you have done is either plugged in the sensor wires (the thin wires going to the motor) in the wrong order or the main wires. By the look of it, that may not be possible on this wheel as the thin wires are on a (polarised?) plug and main power wires colour coded  (but are you certain it was blue to blue, yellow to yellow etc it does not always follow!) 

If this is not the case then I think it is possible a wire may have broken when it was disturbed (or it might just be that plug not making proper contact?) it could be any one of those thin or thick motor wires being disconnected or swapped could, I think, cause those symptoms. 

Basically I am making the above assumptions,

• Firstly on the basis that the odds that an electrical component should pick just when you were working on it to die are VERY small (unless you handled the main board and did static electricity damage - unlikely?)
• Secondly, The thin wires go to magnetic sensors (hall-effect transistors) which detect the  position of the magnets in the wheel in order to decide which coils of the motor to energise, the control board pumps current down one set of coils (two of the three thick motor wires) until the sensors detect the movement and switch to the next set of coils (a different pair of the three thick wires) and  so on. If one of those thin wires were broken the motor would start to move and never see the movement so it would keep increasing power to the first coils. If the main wires got swapped it would try to run backwards and the sensors would get very confused, if one thick wire was broken the wheel cannot turn so it can over current the other two coils trying to get it to turn.

All plugs are polarized and all the wires are color coded:

Unplugged. The thin wires are connected by a polarized plug. If it is pluged out, the App shows the message 0013: Holzer fault.

If it is plugged in the app is able to show the speed, when I rotate the wheel to left and right (both directions)

Here you can see a videos what happens when all wires are connected an I switch it on:

https://1drv.ms/v/s!AlC8lKs4xvyNiMxu3X6OQF_z_5qhNw

At first is stays still. When I turn the device a little bit it starts to vibrate. Hence, I switched it off.

 

 

 

[Rol]

16 minutes ago, The Fat Unicyclist said:

The wheel wasn't accidentally put in the wrong way around?

I am not sure if I understood your question correctly. But I calibtated the upright position using the App.

[Chriull]

. @Rol is there a "fault" at the yellow motor cable insulation in your first photo, just before the three motor wires go into the black insulation, or is this just some artefact/something else?

Maybe @IPS Malta has some experience with such cases? Or also he (or other resellers) still could have some zero motherboards (not reserved for their customers) in stock...

I'd personally would swap two of the motor wires - there should be quite some chance that it will just work, but also that this totally wrecks the board (which is maybe already more dead than alive?). If you decide to risk it and try this, do it somewhere outside without burnable stuff around ( in the state the wheel is now you should not do anything with it where fire could lead to harm...)

Edited June 23, 2018 by Chriull

[Chriull]

50 minutes ago, The Fat Unicyclist said:

The wheel wasn't accidentally put in the wrong way around?

 

31 minutes ago, Rol said:

I am not sure if I understood your question correctly. But I calibtated the upright position using the App.

I assume @The Fat Unicyclist meant if you flipped the wheel by 180 degree? But imho that's not possible witht he motor wires coming out on one side of the axle?

[Keith]

1 hour ago, Rol said:

At first is stays still. When I turn the device a little bit it starts to vibrate. Hence, I switched it off.

The wheel does turn completely freely when powered down and with the motor cables plugged in doesn’t it? If it is very stiff it would indicate a failed transistor on the main board, which I stilll think unlikely.

I still feel that this looks like a disconnected power wire, I would agree that your above post does seem to exclude the sensor wires being faulty. If you put a multimeter on ohms across every combination of the three motor wires (the three thick wires disconnected from the main board) you should read pretty much the same resistance across all combinations. It will be wired something like the below picture with the blue yellow and green wires being points 1,2 and 3 below and the resistance should be very low. Don’t worry about the star or delta bit, I don’t know which yours is wired (probably star as torque is higher) but, as you can see the resistance should be the same across every combination regardless.

unfortunately, none of us know, and I certainly cannot find on the internet anywhere, the IPS error codes. As @Chriullhas said @IPS Malta has a lot of expertise, he is a really nice guy, and may even be in China at the moment so if he sees this post he may well be able to identify the error code?

Edited June 23, 2018 by Keith

[Chriull]

41 minutes ago, Keith said:

I still feel that this looks like a disconnected power wire

Seems very sound - would fit the picture of the described problems. And also could happen quite easily while tire changing?!

[Rol]

23 hours ago, Keith said:

The wheel does turn completely freely when powered down and with the motor cables plugged in doesn’t it? If it is very stiff it would indicate a failed transistor on the main board, which I stilll think unlikely.

I still feel that this looks like a disconnected power wire, I would agree that your above post does seem to exclude the sensor wires being faulty. If you put a multimeter on ohms across every combination of the three motor wires (the three thick wires disconnected from the main board) you should read pretty much the same resistance across all combinations.

 

@Keith The electrical resistances across each combination of the thin wires of the motor (of course separate from the thick wires from the motherboard) are about 0.3 ohms. The peak voltage measured by the multimeter is about 56 V in yellow-green and yellow-blue. Green-Blue shows a signal, but the sampling time of my multimeter is too slow. I do not have an oscilloscope.

I found an article about the electrical design within this forum - I repaced the fuse several times ?

On 10/28/2015 at 10:49 AM, micro said:

Electrical design:

There is a 30 amps fuse at the input, it is rated 32 volts. The fuse will blow only when there is something wrong on the board (burnt FET or voltage suppressor in this case). This fuse serves only to contain the fire when something bad happens. You should never try to replace it.

 

On 6/23/2018 at 10:58 AM, The Fat Unicyclist said:

The wheel wasn't accidentally put in the wrong way around?

@The Fat Unicyclist: The mechanical situation is the same as before.

For me, it feels like the software of the main board is not able to get the complete set of sensor data and/or it is not able to react in the right way due to problems with the electric. Any ideas for the next step?

Edited June 24, 2018 by Rol

[Keith]

26 minutes ago, Rol said:

 

@Keith The electrical restiances across each combination of the THIN wires of the motor (of course disconnected from the thick wires from the main board) are about 0.3 Ohms. The of the peak voltage measured by the multimeter is about 56V across yellow-green, and yellow-blue. Green-blue shows some signal, but the sampling time of my multimeter is to slow. I do not have an oscilloscope. 

I found an article about the electrical design within this forum - I repaced the fuse several times ?

For me, it feels like the software of the main board is not able to get the complete set of sensor data and/or it is not able to react in the right way due to problems with the electric. Any ideas for the next step?

I said to check across the THICK wires not thin. The thin wires go to transistors you aren’t likely to see much there. Was that just a typing error?

whilst the article suggests the fuse will blow only on a board fault, I am not absolutely certain this is correct? A disconnect in just one of the three power wires will also cause over current - I’ve seen this a number of times in model aircraft BLDC motors which are much the same technology. It is possible the IPS might have a safe overcurrent shut down algorithm so would only blow the fuse on a board fault, but on an older wheel like this it seems unlikely they would have protected against a fault that shouldn’t happen.

You seem to be sure  you have proved you have a board fault, if so you have already said you can do nothing about it if you cannot obtain a spare - unless of course you feel you can diagnose back to a possibly faulty FET or two and replace them yourself - not the easiest thing to do and not something we will be able to help with much unless you can physically see burnt transistors. Even then it could have cascade failed other components on the board as well.

I still think it would be very odd that the electrics should fail at the same time as just changing a tyre, unless, of course you did actually power it up with one power wire disconnected which may well have blown one of more FET’s at the same time as the fuse.

You would have to actually spin the wheel at an even speed to measure the output across the power wires, the board would have to be seeing the wheel passing through every set of magnets (assuming sensors worked properly)  to evenly send signal down all three coils (one after the other) if the wheel is stationary it will be driving only one or two coils 

Edited June 24, 2018 by Keith

[Rol]

29 minutes ago, Keith said:

I said to check across the THICK wires not thin. The thin wires go to transistors you aren’t likely to see much there. Was that just a typing error

@Keith No, I thought it was a typing error of you, because you have send pictures of the electrical scheme of the motor - Sorry.

The electrical resistance across all combinations of thick wires (from the motherboard) is nearly infinite.

The device was fully functional before I changed his tire. I am not sure about the situation when the tire is changed. But every wire does not have to be connected for mechanical reasons.

I do not have the equipment at home to measure pulsed current or voltage, just a multimeter.

It's difficult for me to follow your suggestions, because when I turn on the device, I have a few seconds before the device switches to the safe mode and does nothing more. It reacts as if it is not connected to the ground and switches to error mode.

Do you need to measure while the device is on?

I do not want to switch to combinations of the engine's color coded cables, as I'm pretty sure the combination is right (colors fit).

By the way, what does "FET" mean? I could ask someone from my university to check the boards components. I'm a mechanical engineer and a computer scientist - almost nothing to do with electrical components.

I think it's bad that the device worked before and now unfortunately not anymore.

[IPS Malta]

Hi guys, I just returned to China from some long busy travelling. I will get in touch with IPS to have more knowledge on the error code while also reading deeper in the thread as I just skimmed through. Will get back to you shortly.

[Chriull]

On 6/24/2018 at 12:25 PM, Rol said:

 

@Keith The electrical resistances across each combination of the thin wires of the motor (of course separate from the thick wires from the motherboard) are about 0.3 ohms.

 

By thin wires you meant this three wires (yellow, green blue) going into the motor? So this 0.3 ohms in every combination would be fine and show no broken cable..

On 6/24/2018 at 12:25 PM, Rol said:

The peak voltage measured by the multimeter is about 56 V in yellow-green and yellow-blue. Green-Blue shows a signal, but the sampling time of my multimeter is too slow. I do not have an oscilloscope.

This also seems to be fine - the normal commutation sequence is voltage between 2 wire pairs and none at the other.

On 6/24/2018 at 12:25 PM, Rol said:

 

I found an article about the electrical design within this forum - I repaced the fuse several times ?

 

@The Fat Unicyclist: The mechanical situation is the same as before.

For me, it feels like the software of the main board is not able to get the complete set of sensor data and/or it is not able to react in the right way due to problems with the electric. Any ideas for the next step?

 

On 6/24/2018 at 1:51 PM, Rol said:

The electrical resistance across all combinations of thick wires (from the motherboard) is nearly infinite.

By thick wires you mean the three wires (red, green, blue) coming from the motor and having the connector to the same colored a bit thinner wires to the motor?

If so nearly infinite resistance is fine (if the cable is not broken...)

On 6/24/2018 at 1:51 PM, Rol said:

The device was fully functional before I changed his tire. I am not sure about the situation when the tire is changed. But every wire does not have to be connected for mechanical reasons.

So motor wiring could be ok and next which could be broken while the tire change are the thinner hal sensor wires (yellow,red, green, yellow and blue from the bigger white connector) or their connectors messed up?

On 6/24/2018 at 1:51 PM, Rol said:

It's difficult for me to follow your suggestions, because when I turn on the device, I have a few seconds before the device switches to the safe mode and does nothing more. It reacts as if it is not connected to the ground and switches to error mode.

Do you need to measure while the device is on?

No. But i think you already done the measurements @Keith suggested - imo the misunderstanding was with "thin" wires again - imo he understood the hal sensor wires by "thin" and you meant the a bit thinner (compared to the same cables coming from the motherboard) cables to the motor coils?

On 6/24/2018 at 1:51 PM, Rol said:

I do not want to switch to combinations of the engine's color coded cables, as I'm pretty sure the combination is right (colors fit).

Best thing to do now seems to wait for @IPS Malta's advice.

On 6/24/2018 at 1:51 PM, Rol said:

By the way, what does "FET" mean?

Field effect transistor-

 

[IPS Malta]

@Rol I reached IPS and the person assisting us just came back to the office. you can test both the motor and the board to find out where the problem is. The fuse you know how to check and you know it is getting broken. For the sake of others, I will write it anyway.

Fuse test: Disconnect all the board wires completely from LED lights, motor and battery, open the control board box and locate the fuse. The fuse resistance should be less than 1Ω

MosFET Test: You will need to test 12 Mosfets as follows (At this point the document they sent to me is a bit misleading so I am quoting as they say in their document with my suggestions in brackets)

• common or ground (-ve black probe) on the middle leg and +ve (red) on the left and right leg. record each resistance for the 12 fets i.e. 24 tests (due to polarity they request I think you need to find the voltage drop of each using the diode setup)
• Connect the +ve (red) to the left leg and -ve (black) to the right leg and record the last 12 resistances (again due to polarity they request I think you need to find the voltage drop of each using the diode setup)
• All recorded resistances should be similar and all high values of several Ohms (or I believe you should find no shorts)

MosFET Diagram with PCB Behind FET for spoon feeding orientation

|----------|------ Right

|  FET  |------ Middle

|______|------ Left

Motor Test: With the motor disconnected, set the multimeter to Diode mode (this is a bit misleading for several reasons here too so will take it the same as the above)

• connect the +ve (Red) probe to the black wire and connect the -ve (black) probe to the green yellow and blue lines (do them for the 6/7 as I am not sure it's the Haltzer sensors wires or motor cables and about that red cable in pictures). You should read a voltage drop of around 0.7v their picture shows 0.602 on the hall sensor connection only not power cables
• Connect the -ve (Black) probe to the same black wire and connect the +ve (red) probe to the red green blue lines (again I don't know if its the power or hall lines, if the above is hall sensors then this might be the power lines but that's my assumption on the power lines not being tested)  You should get a 0 of no resistance on the multimeter at this point.
• this is my point. While at it check to have same resistance on yellow-green, green-blue, blue to yellow I presume it should be.

Once you do all the tests above kindly report the findings and results here before any conclusions as I have some assumptions and I don't have a Zero to check on ATM. Start off with the motor test... that's the easiest with no need to opening the control board

Edited July 12, 2018 by IPS Malta
Not working properly in writing

[IPS Malta]

@Rol did you manage to get some investigation done on your problem?

[MeaganArky]

Hi... For some reason manufacturers use the same wire colors as a house breaker panel. With all the wires connected correctly on the battery, it may have a dead short somewhere. Not fun to find. With all fuses pulled for anything that is normally drawing current.