Wiring previous-year banner sensors

[Cjmovie]

Hello,

We are trying to wire a banner reflective infrared sensor from previous years (among other, non-banner sensors). We are having some trouble getting the signal from these devices.

For instance, on the banner device: We have the device powered properly (the status light is green, and the orange NC/NO light correctly turns on and off as proper). I was able to find wiring information in posts from previous years on these forums.

However, when trying to get a signal from the device, we have a problem. Regardless of whether we use the NC or NO signal line, we are reading 0 on our digital input. On the analog input, we read voltages very close to 0 at all times. This happens with both sensor devices.

When tested with a multimeter, the voltage correctly switches between low and high (for one sensor, 0 and 12V, for the banner, 0 and 3ish). We only get this signal, however, when comparing the signal line to a positive line on the robot - not ground. When comparing to ground, we always receive a 0V signal.

I seem to recall that the digital sidecar compares all voltages to ground, which would be why we are having our trouble. However, I am unsure of how to wire these devices such that they will work properly - even the banner sensor, which was included in the KOP in previous years (legality of its use isn't a big deal, it's mainly being used for data collection at this point).

I have looked up also information on PNP vs NPN sensors, and although I understand the theory, am unable to grasp a way of forming the system to meet our needs.

Any information on the proper wiring of these devices would be greatly appreciated.

Thanks,
CJ

[Alan Anderson]

Originally Posted by Cjmovie

Regardless of whether we use the NC or NO signal line, we are reading 0 on our digital input.

It sounds like you have everything wired properly: 12 volts to the power (brown?) lead, ground to the ground (blue?) lead, and one of the signal lines (black? white?) to the digital input. You do have it to the SIG pin of a DIGITAL I/O on the Digital Sidecar, right?

The output of the Banner Sensor is essentially a switch that is either open or closed to ground. You won't read a voltage on it unless you supply one yourself, or if you connect it to something that supplies a pullup. The Digital Sidecar inputs have such pullups, so things should be working for you (they do for us -- we are using a Banner sensor as well).

The pullup on the Digital Sidecar won't work if there is no power coming to it. Double-check that the DS has all three power LEDs lit: 12v, 5v, and 6v.

[Cjmovie]

Originally Posted by Alan Anderson

It sounds like you have everything wired properly: 12 volts to the power (brown?) lead, ground to the ground (blue?) lead, and one of the signal lines (black? white?) to the digital input. You do have it to the SIG pin of a DIGITAL I/O on the Digital Sidecar, right?

The output of the Banner Sensor is essentially a switch that is either open or closed to ground. You won't read a voltage on it unless you supply one yourself, or if you connect it to something that supplies a pullup. The Digital Sidecar inputs have such pullups, so things should be working for you (they do for us -- we are using a Banner sensor as well).

The pullup on the Digital Sidecar won't work if there is no power coming to it. Double-check that the DS has all three power LEDs lit: 12v, 5v, and 6v.

All lights on the digital sidecar are active - we're able to run other sensors (e.g., limit switches) off of it well enough, as well as servos etc. It's really a confounding problem - I can't figure out anywhere that something is wrong, but at the same time, it's as if nothing must be right :-)
We also tried connected the line up to the analog bumper to get a voltage reading in, like I said - it was also powered, and operational (as we have a battery voltage reading on the driver station).

We had the NC/NO hooked up to the various colors of a PWM line, however, the ends of these went to the proper points on the digital sidecar / analog breakout. We had both NC/NO wired on one wire with it sideways on the line of signal inputs (perpendicular to the normal method) so that we could test both at once - nil.

I assume it would be smart at this point to try swapping the digital sidecars out?

[Mr. Lim]

We're using the QS18 this year.

Brown = +12V on the power distribution through a 20A
Blue = GND on the power distribution

Black = Do not connect
White = Run this wire, and ONLY this wire to the digital sidecar, any Digital INPUT, to the SIGNAL pin - where the white wire on a PWM cable would go.

These are sinking outputs, and the DSC has pull-up resistors in them. It's a common ground from the digital input to the power distribution board, and so this works without too much of a problem.

Good luck!

[Alan Anderson]

Originally Posted by Cjmovie

We had the NC/NO hooked up to the various colors of a PWM line, however, the ends of these went to the proper points on the digital sidecar / analog breakout. We had both NC/NO wired on one wire with it sideways on the line of signal inputs (perpendicular to the normal method) so that we could test both at once - nil.

I can't tell what you mean here. The sensor signals don't go to a PWM port; they go to a DIGITAL I/O port. I can't think of a good reason for any signal to be connected to both a Digital Sidecar and an Analog Breakout. In any case, if you connected both black and white outputs of the Banner sensor to the same wire, you'd always see the result at ground.

[Cjmovie]

Originally Posted by Alan Anderson

I can't tell what you mean here. The sensor signals don't go to a PWM port; they go to a DIGITAL I/O port. I can't think of a good reason for any signal to be connected to both a Digital Sidecar and an Analog Breakout. In any case, if you connected both black and white outputs of the Banner sensor to the same wire, you'd always see the result at ground.

I meant that we cannibalized a PWM signal cable, and then connected through that to the lines on the digital sidecar input.

We only temporarily connected the signal line to the analog breakout to see if there was any slight voltage fluctuation when the sensor switched between states.

We have tried both the NC/NO wires as the signal line. I meant that we plugged the PWM cable female end into the digital input sideways, so that all three wires on the line were going to one of 3 digital input signal lines on the digital sidecar (not ground or power). This was to save killing even more cables to test both the NC and NO lines at the same time.

[Mike Copioli]

Please do exactly what Shawn suggested above. We have the same Banner sensor and we wired it the same way Shawn did (except we used the black wire). It doesn't matter what wire (white or black) you use for signal, but only use one. You must, however, get your 12v from the PDP adn a 20 amp breaker or it will not work. You can not get power from the digital sidecar for your banner sensor as it can only deliver 5 volts to the digital IO.

[Cjmovie]

Originally Posted by Mike Copioli

Please do exactly what Shawn suggested above. We have the same Banner sensor and we wired it the same way Shawn did (except we used the black wire). It doesn't matter what wire (white or black) you use for signal, but only use one. You must, however, get your 12v from the PDP adn a 20 amp breaker or it will not work. You can not get power from the digital sidecar for your banner sensor as it can only deliver 5 volts to the digital IO.

That's how we currently have it wired. The sensor is active as well, showing correct status indications on its LEDs. Also, like I said, if checked with a voltmeter it sends the correct signals. However, if I hook this same line up to a SIG line on the digital sidecar and declare it as an input in code, it does not change in value when the sensor switches. The same code detects switching properly with a microswitch.

We have tried two banner sensors.

[Alan Anderson]

Originally Posted by Cjmovie

When tested with a multimeter, the voltage correctly switches between low and high (for one sensor, 0 and 12V, for the banner, 0 and 3ish). We only get this signal, however, when comparing the signal line to a positive line on the robot - not ground. When comparing to ground, we always receive a 0V signal.

Originally Posted by Cjmovie

...like I said, if checked with a voltmeter it sends the correct signals.

I finally figured out what you're doing with the voltmeter check. The signal line from the sensor doesn't supply a voltage; it is an NPN transistor switching the line to ground. If you put the positive lead of a multimeter on a voltage source and the negative lead on the sensor signal line, you should read that voltage when the signal is active and see an open circuit when the signal is inactive.

I'm a little concerned by the "3ish" you measured, however. I don't see any way for the Banner sensor we're talking about to do that (unless it is broken).

Here's a simple three-step diagnostic procedure that will help close in on the source of the problem. With the sensor disconnected from the DIGITAL I/O SIGNAL pin, measure the voltage on the pin. You should see +5 volts (the pin has a 10k pullup resistor). Connect the Normally Open sensor output wire. You should still see +5 volts. Activate the sensor. You should now see 0 volts. Does this match your results?