Hall sensor to replace potentiometer in DC motor controller project

Thread Starter

AbsoluteNoob

Joined Feb 23, 2021
2
Hi all. Apologies if this has been posted several times before. I won't confess to understand a huge amount about electronics so would appreciate some help please.

I'm building a battery powered 12v go kart for my children using the guts from an old golf trolley. The trolley has a 12v battery, and the motor controller receives input from a linear position sensor/soft pot https://www.spectrasymbol.com/wp-content/uploads/2016/12/SOFTPOT-DATA-SHEET-Rev-F3.pdf

The range of the motor controller input (please don't quote me on it) is between 0v and 4.8v.

I've purchased a cheap foot pedal from the jungle website as a bit of a punt, and when It arrived I realised it's actually based on a hall effect sensor. https://www.amazon.co.uk/gp/product/B07QBY88TJ/ref=ppx_yo_dt_b_asin_title_o00_s00?ie=UTF8&psc=1

Details of the sensor are:-

  • Input Voltage: 5V
  • Output: 0.8-4.2V
  • Output Rated Load Current: 10ma
  • Output Type: Voltage
  • Pedal Rotation Angle: 15 Degrees
  • Max Load Current: 10ma

The difficulty I'm faced with is that the motor control module requires absolute 0v from the signal wire, up to 5v for full power. With the hall sensor allowing 0.8v through the signal in its "resting" position, It's in effect sensing that the trolley would just set off when switched on, and this locks out the controller until the signal returns to 0v. I have understanding of the differences between the pot and the hall sensor, and I know I can overcome this issue by doing away with the hall sensor altogether, but I'd love work out a way of being able to use the foot pedal in its current guises, rather than having to botch a rotary pot into the equation somewhere.

If anyone has any ideas on how to get round this, I'd really appreciate the help. TIA.
 

KeithWalker

Joined Jul 10, 2017
1,783
You have supplied detailed information about both input devices but only the control input requirements for the motor controller. There may be a very simple solution but I can not make any suggestions without more information. Do you have a schematic diagram of the motor control unit?
 
it's always hard getting 0 and 5V. You can get close with a 5V supply.

In general, you should look for single supply OP amp that can run on 5V and will reach close to zero and close to +5.

You need to subtract 0.8 and then add some gain to get a 0-5V signal.

There is a chip that can generate a tiny negative supply, so that a +-5V powered OP-amp should work. You might power it with say -600mV and +5 so it can reach zero.
 

LowQCab

Joined Nov 6, 2012
538
First you have to quantify exactly what what you have to work with .......
It is very un-likely that your Motor Controller Input has to have "zero" Volts for zero output.
By the same token, it is likely to deliver full Power at something less than ~5V.
And, a very important point ....... What is the Motor-Control-Box Speed-Control-Input Impedance ?

It would be a huge bonus to have Schematics and Spec-Sheets for both Devices,
otherwise, it's a total "Crap-Shoot" as to
whether or not they will work together "as-is"
without causing the release of the dreaded "Blue-Smoke".

It would be helpful to know what the ACTUAL,
as-installed, and used, and measured,
Voltage Output of the original Analogue-Pot-Control was.

A Pot is "capable" of WAY more than the ~10ma maximum Current output of the Hall-Sensor,
and so, the Input-Impedance of your Motor-Controller could possibly be
too low for the Current Output capabilities of the Hall-Sensor.

A simple Amplifier Board could be built that
would guarantee compatibility between the two components.
Is this what you are asking for ??
.
.
.Golf Cart Throttle Flat .png
 

Thread Starter

AbsoluteNoob

Joined Feb 23, 2021
2
Thank you all for the kind replies.

I really did try to provide as much detail as I humanly could. Sadly the company who made the golf trolley (Powakaddy) have never supplied schematics for the controller. The trolley only sold for a few years and they are now obsolete. the controller can be found here https://www.electricgolftrolleyspares.com/powakaddy-touch-controller-pk6627-3653-p.asp which I know isn't much help.

Having rechecked the signal wire coming FROM the motor control, it in fact outputs at exactly 3.3v. The RETURN signal wire from the physical input control then needs to send back 0v for off graduating up to the full 3.3v for the motor to run at full speed.

Just to go back to what I mentioned regarding the lock out mechanism, this type of trolley uses a slide handle that you push in to make the trolley go. This then sends the trolley on its merry way, and as you walk you're holding the handle, so as your pace of walking changes, the trolley actually adjusts pace automatically. The shut off is in place in case the trolley is ever switched on with the handle depressed, preventing it from just taking off when switched on. The handle needs to be fully retracted so the control unit senses 0v on the return signal wire, before the lock out resets, and the trolley becomes active again.

The pedal does actually function to a degree. If I simply connect the +3.3v and negative wires to the hall sensor, but do not connect the signal wire, this allows the motor controller to enter its resting state ready to accept the voltage from the physical control. If I then attach the signal wire from the foot pedal, the motor instantly starts at a very low speed, due to the fact the hall sensor is always outputting 0.8v. When I depress the pedal, the motor does in fact speed up in line with the pedal position, but then also doesn't reach the full 3.3v output that the control unit needs, as the hall sensor will only output 2.8v of the 3.3v that it receives.

On testing, the original pot outputs 0v at rest and then 3.3v at full power.

The interface LowQCab has posted I feel would almost certainly do the trick, presumably the output signal can be set to 0v at rest and then 3.3v at maximum output, and then these would output relative to the input from the hall effect sensor in the pedal. I hope I'm understanding that correctly? I also presume that this board would then need it's own power supply? I have to be honest though, I don't have the knowhow or equipment to build this interface so an off the shelf option would be a big help :)

Thanks again for all the input.
 

LowQCab

Joined Nov 6, 2012
538
The Circuit "as-is", will go from Ground to what ever the Maximum Supplied Voltage is, or
may be adjusted down so that the maximum output is around half of the Supply Voltage.

Now the only 2 questions are .........
Will the Hall-Sensor be Happy with 3.3V for a Power Supply, ( it will probably be fine ),
and, is the 3.3V Supply-Voltage severely Current-Limited,
it's probably not,
it is probably supplied by the Voltage Regulator for a Micro-Controller
that is built-into the Main-Motor-Controller.

So it's all good !!! ........

If I had known that the Voltage was 3.3V, I would have surmised that the Control-Input
has a very High-Impedance, meaning that, a ~50 Ohm Driving Impedance is ridiculous overkill,
but it won't hurt anything, and will tend to keep any outside electrical noise suppressed.

You can use just one 8-Pin DIP, (2- Op-Amps), instead of two 8-Pin DIPs, (4- Op-Amps),
just leave-out the bottom two Op-Amps, and the bottom two 200-Ohm Resistors in the Schematic.
The Op-Amps will operate just fine all the way down to 2.7V.
.
.
.
 

du00000001

Joined Nov 10, 2020
63
The least-effort (in terms of # of components) would be a small 8-Bit microcontroller with one ADC input and one DAC output:
connect the Hall's output to an ADC input, correct the offset and scale the result appropriately and send the result to the DAC.
A PIC16LF167x would come with 14 pins (although you won't utilize more than 4 of these). Add a cap and you're ready for any supply from maybe 2.5 V to 5.5 V.
This would be my personal favorite (has some more advantages) but requires a bit of software development.
 

RPLaJeunesse

Joined Jul 29, 2018
165
Perhaps the above circuit from LowQCab could be built more cost effectively with the MCP6004-I/P, a quad op-amp in DIP package selling in singles for $0.48? Why spend $12 when precision isn't called for?
 

LowQCab

Joined Nov 6, 2012
538
The LT-1632-IN-8 is simply one of my "go-to" "universal" Op-Amps.
It has a relatively stout Output Section at 35ma+,
2.7-V to 36-V Supply capability,
and Rail-to-Rail Inputs and Outputs,
and it's stable at Unity-Gain, and/or, with DC applications.
There are very few Chips that can make all these claims, all at the same time,
I know, I've spent an absolutely stupid amount of time doing searches,
and going over Spec-Sheets trying to find them.

All of which, make this Chip a pleasure to work with,
with very few compromises or "work-arounds" required.

This is not a "Mass-Produced" "Consumer-Commodity" Adapter-Box, so
~$12.oo, one time, is not a terrible price to pay
for no worries, an no extensive Circuit-Stability-Testing with a Scope.
It just works, AND IT'S NOT AN SMD-ONLY PART !!!! :) :) :)

( The MCP6004 is only rated for ~6-Volts, vs ~36-Volts for the LT-1632,
so it has zero margin for error, or immunity, from spikes/noise, on the Power-Supply-Rail.
It also supplies only 23-ma-Max Output Current,
( instead of ~35-ma @ 5-V, / ~75-ma @ 30-V, for the LT-1632 ),
and I was initially looking for a very comfortable ~100-ma Output capability,
before I "guessed" that it's probably only pushing an ADC Input on a Micro-Controller,
and probably not,
the possibly crude type of direct Power-Transistor-Drive that Golf-Carts are infamous for ).

There are very good reasons why the LT-1632 costs ~20-X more,
it's much closer to "the Ideal-Op-Amp" model than ~90% of the offerings out there.

Not counting Special-Purpose Op-Amps,
I only work with 4-Part-Numbers,
that, between the four,
work great for almost any application you could imagine.
.
.
.
 
Last edited:
Top