Internet PLC Forum

Glad to help.

Please remember that the use of the ladder logic COUNTERS do come with some limitations:

• The COUNTER resets to an special value of "-1".  This happens as a result of PLC reset, PLC power up and the use of the [RSCtr] Special function.
• The maximum value that the counter can reach is determined by the Set Value in the I/O table.
• The maximum Set Value is 9999 decimal.
• The COUNTER will down count from 0 to the Set Value.  There are no negative numbers other than the special "-1" value.
• The contact associated with the COUNTER has some interesting behaviors in regards to when the contact is active.

You may find that using custom functions to increment, decrement and reset a simple variable may give you more flexibility than using one of the ladder logic counters.

Gary D

I, too,  worry about the need to change the TIMER SV often.  My concern is with how the SV are stored in the PLC.  I am worried that these are stored in an EEPROM and there is a danger that you may wear out the EEPROM by rewriting these values over and over.

Another alternative is to not use the ladder logic TIMER mechanism and do the work with custom functions.

The following shows how you can get the same basic behavior without the use of the ladder logic TIMER mechanism.  This approach may fit better with your HMI and it skirts around my concern with EEPROM, for the most part.

The following custom function, InitPumpCycle is invoked once with the 1st. Scan contact. This function sets up the initial values that define the pump ON/OFF times in 0.5 second units.

' InitPumpCycle - Set up Initial values for variable duty pump mechanism
'

' The next 2 varaibles hold the On/Off SetValues to be used to load
'  the CycleCntr variable.  
'
' The OnTimeSV and OffTimeSV variables are initialized on power up. These
' varaibles can be changed by your HMI code and the changed variables will
' be loaded into CycleCntr by the CheckPump function at the correct time
'
OnTimeSV = 10
OffTimeSV = 5???'

' This variable is used as a down counter for the current phase of the pump
'
CycleCntr = 0

[/color]

This custom function, CheckPump is run every 0.5 seconds using "Clk:0.5s" special contact. RUN is an INPUT that is used to determine when the PUMP is to be cycled ON/OFF or just left off.  You will notice that only a single variable, CycleCntr is used for both the ON and OFF timing.  


' CheckPump - Custom function to handle variable duty cyle PUMP OUTPUT
'
IF TestIO(RUN) = 1
 ' Pump is to be cycled
'
???CycleCntr = CycleCntr-1???' Decrement
???IF (CycleCntr <= 0)'  and check counter

' CycleCntr has hit 0
'   Change ouput state and
'   reload CycleCntr with the correct value
'
???ToggleIO PUMP

???IF (TestIO (PUMP) = 1)
???CycleCntr = OnTimeSV
???ELSE
???CycleCntr = OffTimeSV
???ENDIF
???ENDIF
ELSE
' Pump is to be stopped
'
???CycleCntr = 0???
???CLRIO PUMP
ENDIF
[/color]

Please note that the example takes advantage of the #Define to assign a locations in the DM[] array to the variables named OnTimeSV, OffTimeSV and CycleCntr.

#???#Define Name???Value
1???OnTimeSV???DM[1]
2???OffTimeSV???DM[2]
3???CycleCntr???DM[3]
[/color]

If you'd like a runnable version of this test code, just email me. In my user profile is my email address.  I can't put my email address in the body of this post because the YaBB software considers my email address to be obscene.

Gary D.

Hi,

I've have a couple of questions.

When you say that the "output does not cycle on and off in a precise manner", how big an error in cycle time are you seeing?

How much of an error in cycle time can your application tolerate?

Can you give me a hint of the code that your are using?  

I'm a little concerned about your reference to "PV" or present value.  

If you are using timers, you may really want to be changing the "SV" or set value.  The SV is the value that the TIMER is loaded with when it is made active. If you change a TIMER's PV this may change the time period for the current cycle, but the TIMER will reload with the SV on the next usage.  

Look at the SetTimerSV and the GetTimerSV (n) statements change or examine the SV for a given timer.

I have more questions, but this should get the dialogue going.


Gary D.

Your in luck.  I was bored and wrote you a brute force CRC16-CCITT algorithm in TBASIC.  

You can copy and and paste this algorithm into your test code.

Here is the guts of the algorithm, "CRC16_CCITT":

// CRC16_CCITT - custom function to compute CRC16-CCITT CRC
//
// This is a brute force bit at a time algorithm.
//
// On input: Byte8 holds the 8 bit data to be accumulated into the CRC
//           CRC is the 16 bit CRC
//
// On exit: CRC is updated
//          Byte8 is all cleared


Byte8 = Byte8 * 256???// shift incoming data 8 bits to the left
// this aligns the most significant bit of the
// incoming data with the 16-bit CRC value to
// slightly simplify the algrorithm.

FOR i = 1 to 8
???j = Byte8 ^ CRC & &H8000???// check to see if feedback multiplier will be needed
???LSHIFT CRC,1???

???if (j)
???CRC = CRC ^ &H1021???// feedback multiplier for CRC16-CCIT
???ENDIF
???LSHIFT Byte8,1
NEXT


Here is the test harness:


// TestString - test CRC16-CCIT algorithm
//

CRC = &Hffff???// Initialze CRC accumulator to all ones

Byte8 = &h03???: CALL CRC16_CCITT
Byte8 = &h0a : CALL CRC16_CCITT
Byte8 = &h81 : CALL CRC16_CCITT
Byte8 = &h00 : CALL CRC16_CCITT
Byte8 = &h00 : CALL CRC16_CCITT
Byte8 = &h00 : CALL CRC16_CCITT
Byte8 = &hff : CALL CRC16_CCITT
Byte8 = &hff : CALL CRC16_CCITT

// 03 0A 81 00 00 00 FF FF  Expected CRC 86 8A


A couple of notes: the variables CRC and Byte8 are actually DM[1] and DM[2] respectively.  I am using the #Define feature that is available in iTRiLOGI 6.43 to create code that is marginally readable.

Good luck,

Gary D

I'd suggest that you use two or more DM[] locations for each stepper  channel.

As an example use:
DM[1] for control for stepper #1
DM[2] for position for stepper #1
DM[3] for control for stepper #2
DM[4] for position for stepper #2
   .
   .
   .
DM[13] for control for stepper #7
DM[14] for position for stepper #7

The master Plc updates the position info DM[2], first, then writes a non-zero command to the control register, DM[1].

The slave Plc detects the non-zero value and executes the appropriate stepper command.  When the command completes, the slave Plc clears the control value (sets it to 0).

The master Plc polls the control register in the slave Plc and must not change values until it "sees" that the slave is ready to accept the next command for that stepper channel.

You will need to use modbus commands to read and write to slave Plc DM registers.

Gary d

The F1616BA has a chunk of non-volatile FRAM.  You can initialize this non-volatile memory using the EEPROM manager that is part of the I-Trilogi program.

On the first scan you can copy the non-volatile values into the DM[]. Look up the load_eep and save_eep tbasic functions to see how to access this true non-volatile memory.

I have used this on several of my projects. It allows me to customize a system without having to modify the PLC program.  I can change some of the tuning constants in the non-volatile memory and then reboot the PLC so that it can use the new values.  This is a very useful technique for storing user data.


Gary d

OK

1. The code you posted is not the code that fails.

2. Your code has been modified to look like the following to allow a 20 character string and does not work:

If z < 10 and Len(z$) < 20
  Z$ = z$+str$(z)
Endif

3. Your code still has bugs the crash the simulator.

Lacking ESP, I don't have a clue how to help.  

Best of luck,

Gary d

TIMER present values can be directly accessed using the following notation:

timerPV[n] = 10       ' n is the timer number 1..255
                             ' set timer n to the value 10

A = timerPV[n]        ' read current value of timer n into variable A

gary d