Count Command skips counting intervals.

SteveWoodrough

OK guys, I'm missing something easy here.·
I'm using a QTI sensor as an encoder input, counting·white sections on a black servo wheel.· The servo wheel is mounted to an RC electric motor shaft.· The motor is controlled separately via a servo exerciser and a separate power supply.·THERE IS NO connection from the motor to the the·STAMP.· ·I have a 10K resistor across the the red and white QTI leads and the QTI senses the white and black areas very wheel.

Here is the problem:· Using the COUNT command the STAMP and QTI will read several time intervals with amazing precsion.· There are 4 black and 4 white sections on the wheel and I was measuring over 1000 COUNTS in 1000 ms.··The trouble is every so often I get a COUNT value of 0 for the entire period.· NOT 1 COUNT.· Then all works fine for a few more periods (seconds) and I get another 0?· What am I missing?·

Here is meat of the "code",·all 4 lines....

DO
· INPUT 0
· QTI=IN0·············· 'Bit VAR QTI=Input into pin 0
··COUNT QTI,1000, Pulses········ 'Count QTI inputs over 1 sec and store as Pulses
· DEBUG DEC Pulses, CR
LOOP

Here is a "sample" of the out put in COUNTS per 1000 ms:

200
201
200
0
200
0
200
201
200
200
200
0
200
200
0
200

I get the same "random" type result at 50 COUNTS per second all the way up to past 1000 COUNTS per second.· Works great then 0 data, etc.


Thanks...
Steve Woodrough

·

Mike Green

Read the description of the COUNT statement. For some reason, the COUNT is timing out. It's not seeing any pulses during the measurement period.

SteveWoodrough

I went back and RTFM'd the Syntax Manual. I changed the COUNT line to look like this:

COUNT QTI,(1000 */ $100), Pulses

Same result as before.....

Carl Hayes

I believe the trouble is in the statement

QTI = IN0

This reads the STATE of input pin 0. This will be either 0 or 1 as pin 0 is low or high.

When you execute this statement at the start of your loop, and pin 0 happens to be low, then QTI is set to 0 and your COUNT statement counts pin 0 as you intend.

When you execute this statement at the start of your loop, and pin 0 happens to be high, then QTI is set to 1 and your COUNT statement counts pin 1, which is not what you intend. Then you get a zero count.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
· -- Carl, nn5i@arrl.net

Tracy Allen

Carl is right about QTI=in0 returning either 0 or 1 as a pin number. If you always want it to be pin 0 then use the statement,
QTI PIN 0

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Tracy Allen
www.emesystems.com

Carl Hayes

Also, I'd take that INPUT statement out of the loop.· No need to redo that every time through the loop.

So your program could be thus, and run more efficiently·-- and incidentally you'd save the byte of variable space you now use for QTI.

QTI··· PIN·· · 0
········ INPUT 0·· ' Actually this is superfluous in any case, unless you earlier used pin 0 for output

········ DO
············ COUNT QTI,1000, Pulses········ 'Count QTI inputs over 1 sec and store as Pulses
············ DEBUG DEC Pulses, CR
········ LOOP

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
· -- Carl, nn5i@arrl.net

SteveWoodrough

Works Great! When I grow up I hope I'm half as samrt as yuo guys.
COUNT is an amazing command. I had the motor at FULL boogie, 30,000 RPM, and got a 2005 COUNTS per second for as long as I dared. The "encoder" is a servo wheel with a drilled out RC airplane wheel collar pressed into the spline. The QTI is reading 4 white painted sections on a black wheel. Way cool. Thanks again guys!