# Review/improve anti collision circuit using hall effect sensors

I'am designing an anti collision circuit with hall effect sensors.

Mechanical principle: it's basically two squares, one smaller than the other. The largest square hits an obstacle making the magnets move past the sensors, and the movement shall stop.

The hall effect sensors goes low then the magnet is near.

Electrical Principle: My design idea is to wire the four sensors parallel with a resistor on each output pin.

In normal state the resistance would be small and a low voltage drop will appear. But then one magnet is out off range from the hall effect sensor, there will be three parallel resistors which will give a larger voltage drop.

I'am using a window comparator(good old LM324) to check the voltage and turn on a solid state relay, which turn on a power relay(controlling the movements).

The idea is that the 10k resistor should be mounted at the sensor, making it possible to detect short circuited wires(GND/10V to output)

Schematic

Questions/worries

1. The difference between the two states is only 0.3V. I would like to raise this to a coupled of volts. How would you do it? (I tried with some diodes and LEDs but did not go as expected)
2. Is the circuit design to sensitive ? how would you have done it?

A few comments as I scan through the schematic:

• Use ground symbols. These are instantly recognisable and avoids having to trace wires and would eliminate about 30% of them.
• Put V+ at the top. Put the socket at the right.
• What's a turn-on relay? If it's a transistor then show that in the symbol.

It's not clear why you added R1 to the circuit. It dilutes the changes as R6, 7, 8 and 9 are switched.

The following table shows calculations show the percentage of V+ (10 V in your case, I think) at the junction of R6, 7, 8 and 9 for the various combinations.

                    5k + 39k pull-up      10k pull-up
1 x 10k -> 10k           18%                  50%
2 x 10k ->  5k           10.4%                33%
3 x 10k ->  3k3           7%                  25%
4 x 10k ->  2k5           5.4%                20%


You can see that decreasing R2 to 10k and shorting out R1 will result in much higher differences.

simulate this circuit – Schematic created using CircuitLab

Figure 1. A cleaner solution?

• If the sensors are analog then simplify the comparitor circuit as shown here. Most have open-collector outputs that can be paralleled in an 'OR' configuration.
• If the sensors are digital then just wire the sensors in 'OR' configuration. Use diodes if required to prevent them fighting to pull high / low simultaneously.
• Thanks for your reply and your time, I really appreciate it. Your schematic improve suggestions is on point. I use R1 to make sure that there is a voltages drop. Your alternative solution is quite clever, thanks for sharing. Commented Jul 28, 2017 at 19:56
• R1 reduces the voltage drop. If you require further clarification then add some more details to your question - particularly part numbers for the Hall sensors. Commented Jul 28, 2017 at 20:00