Reversing MiniPro Steering sensor

[xJlaIIax]

Donald, Thank you very much!

[xJlaIIax]

Donald, I want to place the potentiometers(4,7KR and 10KR) and the switch(Reverse and Normal) outside of the PCB.
All connections are to be made on the connectors JST PA series.
I made changes to the schema.
Please check the schema I made.
All aright?

If all is well then I will start making the PCB.

 

Photo:

Edited July 14, 2017 by xJlaIIax

[donaldduck8]

Yes, the schematic looks OK.

So that there is no interference, the cable to the potis + switch make as short as possible.
If longer than 10cm, please use shielded cable.
The best thing would be, after installation, to test the circuit with an oscilloscope.
A stable DC voltage (without ripple voltage) must be visible on OP pin3 + 6.

The PCB layout shows the photo of JST-XH.
http://www.jst-mfg.com/product/detail_e.php?series=277
These have a pitch of 2.5mm

The right JST-PH have 2.0 mm pitch.
http://www.jst-mfg.com/product/detail_e.php?series=189

It is probably at X1 or X2 the plug from the mainboard directly contacted?
I can not see how the signals are connected in your PCB layout. I hope it's right

For the sensor board, you will probably use an adapter cable with 2 jst connectors at both ends?
Also pay attention to the correct connection!

 

 

Is this really connected?
X1 goes to the mainboard connector and X2 to the sensor board?

 

If so, then the JST plug has to have this pinning!

If x2 has the same pinning as X1, black and yellow in the wiring harness can be interchanged.

 

 

Great work xJlaIIax!

Edited July 14, 2017 by donaldduck8

[xJlaIIax]

7 hours ago, donaldduck8 said:

So that there is no interference, the cable to the potis + switch make as short as possible.
If longer than 10cm, please use shielded cable.
The best thing would be, after installation, to test the circuit with an oscilloscope.
A stable DC voltage (without ripple voltage) must be visible on OP pin3 + 6.

Ok. I will use a cable that will be as short as possible.

 

7 hours ago, donaldduck8 said:

The PCB layout shows the photo of JST-XH.
http://www.jst-mfg.com/product/detail_e.php?series=277
These have a pitch of 2.5mm

The right JST-PH have 2.0 mm pitch.
http://www.jst-mfg.com/product/detail_e.php?series=189

You correctly noticed that on the 3D model shows the connectors JST-XH (2.5 mm pitch).

But in production I'll use the connectors of the JST-PA (2.5 mm pitch).

7 hours ago, donaldduck8 said:

Is this really connected?
X1 goes to the mainboard connector and X2 to the sensor board?

Yes it is.
To connected to the mainboard will use the original cable from the ninebot mini:

I'll make the cable for connecting to the sensor board.

It will have two JST-PA connectors on both sides:

In this case, should I place contacts on the X2 connector in reverse order?

So how in the photo is this correct?:

7 hours ago, donaldduck8 said:

Great work xJlaIIax!

Thank you! 

Edited July 14, 2017 by xJlaIIax

[xJlaIIax]

9 minutes ago, xJlaIIax said:

You correctly noticed that on the 3D model shows the connectors JST-XH (2.5 mm pitch).

But in production I'll use the connectors of the JST-PA (2.5 mm pitch).

But in production I'll use the connectors of the JST-PA (2.0 mm pitch).*

Edited July 14, 2017 by xJlaIIax

[RooEUC]

@xJlaIIax  Why are you spending all this time and effort when the reverse idea has been proven to not work?  Riding the unit backwards does not work.  There are dangerous sensor issues when riding on even a slight lateral slope and when hitting top speed.

Edited July 15, 2017 by RooMiniPro

[donaldduck8]

15 hours ago, xJlaIIax said:

So how in the photo is this correct?:

Yes that's right

 

The two JST connectors must be connected in this way

 

 

Donald

[donaldduck8]

3 hours ago, RooMiniPro said:

Riding the unit backwards does not work.  There are dangerous sensor issues when riding on even a slight lateral slope and when hitting top speed.

I would not say that
He was often referred to the sensor problems.
Let him make his experience with this circuit.

The high speed sensitivity is improved by the gain poti.
The top speed problem can be avoided if you do not drive faster than 25Km/h.
A GPS app could output a warning tone at 23Km/h.

 

3 hours ago, RooMiniPro said:

when riding on even a slight lateral slope

 

I want to clarify this problem with pictures:
On straight road 0 degree slope, Normal mode
Ninebot goes straight ahead

 

On straight road 10 degrees slope to the left, normal mode
Ninebot goes to the left
You have to correct the steering bar to the right

 

On straight road 10 degrees slope to the left, reverse mode
Ninebot goes to the right (crazy)!
You have to correct the steering bar to the left

 

This problem persists.
Probably you may get used to this driving behavior

 

 

Donald

 

 

Edited July 15, 2017 by donaldduck8

[xJlaIIax]

Donald, Do I understand correctly that is connected to the input of hall sensor(Sensor Board) and output connects to the Mainboard?

[donaldduck8]

Do it like the photo!

 

 

Donald

 

 

[RooEUC]

Well it is certainly an interesting learning exercise to do all this work.  If anything, you will gain more experience in electronics and the workings of the Minipro too.  But I think at the end you will realise why the MiniPro is not practical or safe to ride backwards.  Good luck though!

[xJlaIIax]

Your photo was made differently:

Then it should be connected like this:

 

Correctly?

Edited July 17, 2017 by xJlaIIax

[donaldduck8]

Oh man, I am so far from the X1 of the mainboard and X2 the sensor is connected.
You had your previous pictures so marked.
Now the whole thing looks different.

Now you have first uniquely labeled X1 + X2 in the circuit diagram and PCB + pinning.

Now, however, is still a mistake.
Pin 1 + 3 of X1 in the circuit diagram does not match the pinning of the PCB
Pin 1 + 3 of X2 in the circuit diagram does not match the pinning of the PCB
How is it right?
 

Donald

[FreeRide]

On 6/20/2017 at 9:35 PM, RooMiniPro said:

I appreciate you trying this out and all, but to be fair you just did a very long winded and dragged out version of what we already did weeks ago and what we explained in this very thread in much more to-the-point posts - ie. reverse the magnet polarities and discover that the steering doesn't work properly as the unit steers to one side all the time.

Sorry you could not appreciate the benefits of the scientific approach or the information that was revealed by the approach without me spelling it out.  There was also a change in the way I did the mechanical modification compared to previous attempts that I did not disclose as I was waiting for testing to complete to be sure the even with the steering corrected that it was viable.  

[Wilsonintexas]

On ‎6‎/‎4‎/‎2017 at 3:48 PM, FreeRide said:

Good News!

Step Two: Measure the sensor when idle.  Stock sensor and control shaft measured 2.5xxV as expected based on measurements done on the German forum.  

Step Three: Change Control Shaft.  Wow the screws for the steering bar/Quick-Connect were some tight... they used Red Loctite on these ones, Blue is used everywhere else so far.  The shaft sold by More4Mini unfortunately is the Chinese version and it is cheap metal and it doesn't use a screw to retain the magnet assembly it uses epoxy (bas***ds).  The shaft is nicely lubed on the original, you'll have to do that yourself on the new one.  They use a good grease, but I don't know what it is.

Step Four: Measure the idle sensor voltage with new control shaft.  Also measures 2.5xxV, so far so good.  no calibration needed with the new shaft.

Step Five: Reverse the magnetic field.  Of course the adhesive on the Cheap control-shaft went where you don't want it.  It took significant effort and two small hand wounds to free one of the magnets. I slightly chipped the magnet, and some damage to the housing. It would have been easier to do the quality component.  The Chinese assembly also does not label or mark the poles on the magnets so do that before disassembly.  Re-assemble again.  It still works there is no slight turning of the MiniPro when idle. So far so good.

Step Six: Measure the control voltage with the reversed shaft.  The voltage measures 2.5xxV / 2.6xxV that is it is on the very high side like 2.59x, and sometimes 2.6xx so it has moved further away from center this was somewhat expected because I damaged one magnet and the housing so I could not get exact placement.  The good news as reported in Step Five, this change of roughly 0.04-0.05V appears to be within tolerance.  I did not have to switch off Automatic Steering Sensitivity in the app, and the steering control seems to operate exactly as before but in reverse. I may try tweaking it before final assembly.

Step Seven: Reassemble completely suitable for short road test.  Pending

Step Eight: First Road Test. Pending

Do you know the full volt range for the sensor, and is the forward and reverse sensor the same voltage?

I have a thread on converting to a joy stick I starter, I do not want to hijact this thread, but since you had some info on sensor volts, I thought that I would ask.

 

[donaldduck8]

20 hours ago, Wilsonintexas said:

Do you know the full volt range for the sensor

Full Rage is from 1,95V to 3,17V
If the Sensor in idle (middle Position) you have 2,5V

 

20 hours ago, Wilsonintexas said:

and is the forward and reverse sensor the same voltage?

yes, but the direction of the voltage is reversed

 

20 hours ago, Wilsonintexas said:

I have a thread on converting to a joy stick

What do you want to do with the Joy Stick?

 

Donald

[Wilsonintexas]

I do not want to hijack this thread,  so I started another one for my project:

I am not sure how to cut and paste a link to the thread, sorry

 

[jojo33]

@RooMiniPro

@xJlaIIax

&

On 10/5/2017 at 4:27 PM, donaldduck8 said:

Donald

here's how Ninebot makes use of the potential of reverse gear with the Mini (24kmh)

 

[xaveus]

Hi, I saw Jojo33 mention the footpad sensors being left and right-sided, does anyone know if this is 100% correct?   I'm not entirely sure what the footpad sensors do, other than detect there's a rider standing with both feet on the foot-pads, which overrides the 'remote control' or other features where a rider shouldn't be aboard.  I know that putting weight on left or right pad has no effect (neither does lifting one foot).

The custom transporter I'm building doesn't need the complication of the foot-pad sensors, I was going to wire them as if they were 'on' all the time.