Conflicting Sequences

[rlrepoff]

Hello,

 

I am setting up a project that will measure the temperature and adjust appropriate flows of air through my system.  Right now, I have 2 different sequences that are controlling each the flow and the reading of the temperature and RH (using a SHT15 Temperature and RH sensor).  I have these connected to a LabJack.  

 

My flow is adjustable through a stepper motor, which is controlled by one of the sequences.  The SHT15 Temp and RH sensor is controlled by another sequence.  When I run the stepper motor sequence alone, the motor works at regular speeds.  When I run the stepper motor sequence with the SHT15 sequence, the stepper motor slows down drastically.  I do have other sequences in my system, but they do not adversely effect the stepper motor's speed.  I think it has to do with the way the data is collected from the SHT15 sensor via the LabJack.  The SHT15 Temperature and RH are both Test channels.  I have attached the two sequences below for reference.  Any help is greatly appreciated!

 

****************SHT15 Data Collection Sequence:********

 

private err;

global string strIn

private CO2PPM   

   

// Get reading and save them

while(1)

      

   // Line up requests

   AddRequest (ID, LJ_ioSHT_GET_READING, LJ_chSHT_TEMP, 0, 0, 0)

   AddRequest (ID, LJ_ioSHT_GET_READING, LJ_chSHT_RH, 0, 0, 0)

   

   // Execute requests and check for errors

   GoOne(ID);

   ErrorHandler(ID);

 

 

   // Get temperature reading

   err = GetResult (ID, LJ_ioSHT_GET_READING, LJ_chSHT_TEMP, @dblValue)

   eErrorHandler(err, LJ_chSHT_TEMP, LJ_ioSHT_GET_READING)

   Temperature.AddValue(dblValue-273.15)

   

   // Get humidity reading

   err = GetResult (ID, LJ_ioSHT_GET_READING, LJ_chSHT_RH, @dblValue)

   eErrorHandler(err, LJ_chSHT_RH, LJ_ioSHT_GET_READING)

   Humidity.AddValue(dblValue)

   

   catch() // catch any comm errors and ignore

   delay(.1)

  

endcatch

   

endwhile

 

 

 

********************STEPPER MOTOR SEQUENCE:**************************

global n=0

global string moving = ""

 

while(1)

   

   global DilutionCalc = 0

   global LowFlowCalc = 0

   global HighFlowCalc = 0

   

   

   LowFlowCalc = ((MicroAethFlow*((DilutionRatio-1)/DilutionRatio))-(MicroAethFlow*Tolerance/100))

   ?LowFlowCalc

   HighFlowCalc = ((MicroAethFlow*((DilutionRatio-1)/DilutionRatio))+(MicroAethFlow*Tolerance/100))

   ?HighFlowCalc

 

   

while ((mean(honeywellflow2[0,300]) > MicroAethFlow) || (mean(honeywellflow2[0,300]) > (HighFlowCalc)))

   //change the conditions to variables

   //need to turn counterclockwise to get flow reduced

   step_dir[0] = -1

   

   if(step_dir[0]==-1)

      while(n<15)

         for(private.i = 0, i <= 7, i++)

            for(private.j=0,j<4,j++)//loop to set up all digital channel output settings

            private.chanpass = testbit(motor_lookup,j)

            j=j+2

            AddRequest (0, LJ_ioPUT_DIGITAL_BIT, @j, @chanpass, 0, 0)//channel j to setting dictated in binary lookup values

            j=j-2

            endfor

         goOne(0) 

         delay(motorSpeed/1e6)//delay to allow for motor movement

         endfor

         n=n+1

      endwhile

      moving = "Moved CCW to lessen flow"

      ?moving

      n=0

      delay(60)

   endif

endwhile

 

//switch

  // case(step_dir[0]==1 || step_dir[0] ==-1)

  //  step_dir.addvalue(0)

  // case(step_dir[0]== 0 && step_dir[1]==1)

  //    step_dir.addvalue(-1)

  // case(step_dir[0]== 0 && step_dir[1]==-1)

  //    step_dir.addvalue(1)

//endcase

 

while(mean(Honeywellflow2[0,300]) < LowFlowCalc)

      //change conditions to variables

   step_dir[0]=1

 

   if(step_dir[0]==1)

      while(n<15)

         for(private.i = 7, i >= 0, i--)

            for(private.j=0,j<4,j++)//loop to set up all digital channel output settings

            private.chanpass = testbit(motor_lookup,j)

            j=j+2

            AddRequest (0, LJ_ioPUT_DIGITAL_BIT, @j, @chanpass, 0, 0)//channel j to setting dictated in binary lookup values

            j=j-2

            endfor

         goOne(0) 

         delay(motorSpeed/1e6)//delay to allow for motor movement

         endfor

         n=n+1

      endwhile

      n=0

   endif

   moving = "Moved CW to increase flow"

   ?moving

   delay(60)

endwhile

   

 

while(mean(HoneywellFlow2[0,300]) > 35 && mean(HoneywellFlow2[0,300]) < 50)

   step_dir = 0

   

   moving = "The flow is correct.  No movement neccesary."

   ?moving

   

   delay (120)

   

endwhile

 

endwhile

 

 

[AzeoTech]

The problem is that your SHT collection sequence runs at full bore.  There is no delay, except in the catch() which will only execute if there is an error.  This doesn't flat out hang DAQFactory because you are most likely on a multicore machine, and because there are small delays for the LabJack communications.  The problem is that 99.9% of the time spent in the loop is actually doing labjack comms, and when you have that running, you aren't leaving any space for the stepper motor commands to get through.  The pipe to the LabJack is only so big and you are filling it with the SHT sequence.  I don't know what the motorspeed variable is set to, so don't know how fast you are sending commands to the LabJack from that sequence.

 

So, add a delay to the SHT, maybe delay(1)?  Or maybe (0.1) if you really need it that fast.  That will likely fix it.  

 

BTW: it takes about 5ms (0.005 secs) to query the LabJack, i.e. in the GoOne() call.  During that time, any other GoOne()'s to the same device cannot execute and will wait.

 

BTW: Put 0's in front of your decimal places so the decimal doesn't get lost:  0.1 not .1, and don't forget to properly indent.

[AzeoTech]

I got some feedback from LabJack.  The 5ms time I mentioned is for just doing the query.  However, for the SHT15, it takes an additional 320ms to read the temp sensor and 80ms to read the humidity and the LabJack is fully blocked during that time, meaning your stepper motor control won't be able to communicate.  This means putting a delay() in the SHT sequence isn't going to fix the problem.

 

You have two choices here really:

1) switch to analog temperature and humidity readings, though you'd still have the issue of the 5ms, though it may not be a problem provided you add a delay 

2) buy a U3 or another LabJack and have one do the SHT and any other slow stuff, and the other do the stepper control