PDA

View Full Version : optical sensors acting strangely


rosebud
01-23-2003, 05:52 PM
i have two optical sensors hooked up correctly to the digital inputs. one, however, only receives power when the kill switch is down. as soon as it is released the sensor's green status light flickers back to on. i tried switching the terminal on the fuse panel, but it didnt change anything. i even switched to a charged battery. why would a switch only receive power when the robot is not sending any output data?

Jnadke
01-23-2003, 06:01 PM
Did you try using a different optical sensor to test if it defective?

Do you have it wired properly?

Brown goes to the Fuse panel +
Blue goes to the Fuse panel ground -
You hook up the black wire for 0 (no sense) and 1 (sense).
White wire just leave off... or hook up instead of black if you want the reverse.

rosebud
01-23-2003, 06:49 PM
i know i have i wired properly because it used to work perfectly. now when i download one program the green light blinks, but once it sights white it is strong and steady. the other program doesnt have the sensor even receive power when the kill switch is not pressed, which is wierd because the optical sensor is wired so it should receive power no matter which program is running.

Jnadke
01-23-2003, 07:58 PM
I'm having a little bit of trouble following what you mean...

The sensor should always emit light no matter what when the robot is turned on... is it emitting any light?


Did you try connecting a different sensor to test if the other one was defective or broken?

Lloyd Burns
01-25-2003, 12:58 AM
There are two lights on the sensor - looking at the end nearest to the sensitvity control, the "12V-power-on" lamp is on the right when the big lens is down. On the left is the light which lights up when the sensor has found an object (sees the line). [In fact, the lenses look the same direstion as the mian lens, but you can see these lights through the yellow plastic of the sensor body.]

The "I got it" light will flicker when the gain is set too high: it will sense the line, but then will not turn off - the RC can't always decide if the sensor says yes or no, so it might or might not get reported in a debug window, or (if it is wired as one of the first three switches (rc_sw1-3), on the OI panel.

Turn the adjustment CCW (while the sensor is flickering) until it just stops flickering. Then verify that the sensor does still "see" the line.

-=-=-=-=

BTW - like variable ball size, and carpet direction/friction that FIRST has traditionally overlooked, and which have become issues which make or break a robot, I think the line, which FIRST says it is not going to maintain, may become so a big nasty that teams may abandon line following.

rosebud
01-25-2003, 08:31 PM
i apologize. i wasnt very clear. the green power led is what is off when the robot is running and the kill switch not pressed. when the kill switch is pressed down, the power comes back on and the sensor operates normally (the red light showing the sensor has seen or not seen). that is why it is soo confusing. it has power on a non-random basis (every time the kill switch is pressed) but the kill switch being pressed or not should have no influence on whether the sensor has power. if this is still confusing, please email me. thanks
m_n_cheese@hotmail.com

RebAl
01-25-2003, 09:05 PM
If by "kill switch" you mean disable then the sensors will turn off
also this years sensors have a yellow light to indicate the on state

rosebud
01-30-2003, 10:33 AM
never mind, never mind. it was just a bad crimp and i guess when the kill switch was pressed everything else stopped drawing current and the circuit could close. sorry about that. i thought i had tested all of the wiring.

Tom C.
02-01-2003, 11:55 AM
Originally posted by rosebud
i know i have i wired properly because it used to work perfectly. now when i download one program the green light blinks, but once it sights white it is strong and steady.

We're having a similar problem with our optical sensors, and maybe someone can help out.

One of the sensors worked fine, the second would always trigger regardless. A short was detected in the wiring and we re-wired it. Now, on the offending sensor, the green "I got it" light flickers constantly, and none of the sensors trigger switch LEDs on the operator interface anymore.

I tried adjusting the gain control on the sensor to no avail. I suspect we may have blown the sensor due to the short, but why are there no LEDs on the OI anymore? I will try swapping the sensors on the wiring.

rosebud
02-02-2003, 01:05 AM
try using a meter to see if the sensor will conduct electricity.

Lloyd Burns
02-02-2003, 11:29 PM
The switch 2 and 3 LEDs were my constant friends while I debugged - yes, they should blink with the sensor output state.

If you shorted the +12V to the signal line, you may have

a) damaged the sensor - but you can try the other output (if you used white, try black, or vice versa. The first line after the serin could then be

rc_sw# = 1 - rx_sw# ' # is the number of the input your using

or

sensor3 = 1 - rc_sw1 ' if you were using inout 1, but damged input 1 and now have the sensor on three, and you refuse to change you program .

This will invert the signal so you don't have the change anything you've programmed.

b) damaged the RC input. Check that switching (cross two wires, don't use the Banner sensor) that input makes a change in a

DEBUG cls, bin8 rc_swA ,cr ' or rc_swB for higher number inputs
If you only damaged one input, use another (you aren't using all 16, are you ? 8-b If you blew the entire bet, decide if you can get it repaired in time.