Internet PLC Forum

I seem to like Support's approach to changing the changing to a string, and editing the string. Going to try using counters as indexes (I have 2 of those long numbers to handle) to the individual digits of each number, I should be able to change the discrete digits of the strings, and then change it back to the numbers for other programing to handle the switch closure incrementing. MID$ can be used to select the digit from the string.

Another approach would be to use support's approach, but use all numbers and don't get into strings. Has the advantage that you are changing the real number. But I thing you would have to break the number into individual DM[]'s for each digit and then load back to the operational number. As I'm writing, is getting to sound more difficult. :~)

Guess the next step is to start laying out storage locations, and try to implement some indexing scheme.

Thanks again for the replies.

Gary on your rotary encoder: Try operating on 2 digits as a time. How many steps per revolution? I'm guessing around 25, if so that would be a maximum of 2 turns to get to the nearest far  :~) number.  Kind of like that idea. I would physically be able to mount an encoder by the HMI.

This is with a T100MD + MD-HMI420

I would like to display a large number, maximum 999,999 on the display and am looking for a nice way to handle changing that number with the key pad. Since the DM[]'s only count to 65,xxx I need to break the number up to to DM[]'s

I have previously handled time and date by using a counter to keep track of which item is being edited. Then I use the KEY_1 to increment that counter. KEY_2 decrements the item, KEY_3 increments the item, and "Enter" saves the items. Some of these items are not real time, but date/time for an action, and are stored as DM[]'s. This works well.

For the large number, I plan to use 2 DM[]'s, one for the LSB 4 digits (thousands and less), and the other DM[] for 10 and 100 thousands. When the LSB one gets to 9,999, it would rollover to 0, and increment the other one.

For the changing or editing part:
1: Load all the pairs of digits into 3 counters , one for 10's and units, one for 100's and 1000's, and one for the 10,000's and 100,000's
2: A 4th counter (I have plenty unused available) would keep track of which counter is being changed. On the HMI, the edited digits would flash.
The actual editing routine uses subscripts, so you don't have lots of code, and handles under/overflow by incrementing the appropriate DM[] digit.
3: The 3 digit counters get loaded back into the 2 DM[]'s when "Enter" is pressed.

Looking for suggestions on an easier way. At this moment, just these thoughts are all that has been done.

Sorry for not getting back sooner, we had some heavy weather pass through, took down a couple of good sized Norway spruce, have been cleaning up.

Added a Cust Fun near the beginning of the ladder to determine the lapsed time for 1000 cycles through the ladder. Found 1783 maximum HS timer ticks of 0.01 seconds = 0.0178 seconds per cycle. This should easily grab my 2 second period for switch closures. This is in line with Gary's first comment. Hope my findings are correct. Here is the Cust Fun code:

'DM[3945] Is a counter for how many times we have been through the ladder loop
'First time here, start the timer at 6000
if DM[3945] = 0
TimerPV[1] = 6000

ENDIF

IF DM[3945] = 1000
'Timer 1 is set as a high speed timer with 01 second per tick, and will measure the period for the number of iterations
DM[3946] = 6000 - TimerPV[1]
'DM[3946] Is the actual number of time ticks elapsed, and DM[3947} is the maximum number of ticks
IF DM[3946] > DM[3947]
DM[3947] = DM[3946]
 DM[3945] = 0
ENDIF
ELSE
DM[3945] = DM[3945] + 1
ENDIF

I will save the code that Gary generated, probably will be a learning experience.

Thanks again.

Thanks for the reply.

I question whether the scan time is short enough to reliably catch the switch pulse. The program is 4200 code words long with many custom functions, some which do have brief waits. I use a counter to select a MD-HMI screen from a half dozen, using a function key to increment the counter. Clicking the function key doesn't reliably increment the counter.

I was thinking there was a method using a timer, but that likely would involve the scan time.

I was thinking, I might have to use a 555 timer to debounce, but if that's  a way to go, probably would just build up a PIC board. Probably got to do that anyway to interface a remote display. This system has no other involvement with the PLC. The PLC has plenty of unused I/O, and could handle this with minimum effect on other duties.

Good thought on the resistor to protect the switch. The switch does have a maximum current limit.

Just downloaded/installed 6.30 build 1. The standard Windows right click menu to copy/cut/paste is still missing, which goes back to 6.24. Is there a schedule to restore that functionality??  

For those of us that are maintainers rather than developers that's a pain, need to go back to old notes to see how do work.

Had previously gone back to 6.13 for the right click ability, but now see there is a dual core issue, and this  computer is a dual core P4. Are the dual core issues in 6.13? I have had unexplainable issues with monthly actions not being executed, with a series of "IF's" with "AND's"  and "OR's" being true when going to real time monitoring, and checking the variable values.

Do I need to go back to a single core computer running Windows ME?

Removed Trilogi V6.24, installed V6.3, no change. Installed V6.13 in another folder. V6.13 is OK, V6.3 is still, no right click menu or access to edit.

This is an Intel DP965LT motherboard and ATI 2400 HD graphics card. Both have latest drivers.

Sorry for the delay, was busy with other items.

Downgraded to Java2, 1.4.2_06 from the CD from 1.6.0_03-B05. There is no significant changes. Will keep 1.4... loaded for a few days if someone has some comments. fontIncrement   4 results in ladder rung font size a  bigger, about 10 point, but the rung comment, and custom function windows text is small and if changed is little, would like to get at least 2 sizes bigger. Fixed the Ladder comments by editing to shorter lines. The comments need a CR at the end so the cursor is on a blank line to prevent overlap with the rung below labels.

All the windows still open at the screen's left edge,under the left side always visible start bar. Since most applications don't need the 16:9 screen width, this is working well, where I wouldn't of considered with the more square proportions.