AD 250-1 block behaviour

Hi all,

I am struggling all day to understand the strange behaviour I observe, hope that someone here can help.

I have a rotating table with two sets of clamps on both sides.

Stage 1 waits for "Clamp 1" button, stage 2 clamps c1 and waits for "Clamp 2", stage 3 clamps c2, sends "Rotate table" signal and waits for "Rotation complete" response from servo, then jumps back to stage 1. Rotation and control take place in a separate block (which is enabled in stage 3), and that's where the trouble resides - second "Rotation complete" signal never comes on (while during the first rotation it does exactly what I want it to). I am just simulating at this moment, so the block is empty except for a timer that sets the "Rotation complete" bit 1 second after the block is enabled.

I am monitoring all the relevant coils and can see only one changing - the bit enabling servo block toggles constantly, at a rate that doesn't correspond to timer values. When I look at the contacts on the rung that enables the BCALL, I see no change at all - the block should be on until timer sets the "Rotation complete", which in turn causes the stage to go back to Stage 1.

What am I doing wrong?

Kind of hard to guess without seeing the actual program. Any chance you can post it?

In DirectSOFT, pulldown File-->Export-->Program to create a text file. Then use 'Attach Files' here to attach that .txt file to your post.



-Eric

Here it is. It's rather large, I simplified it in the description above. S300 is where I get stuck.

Thank you!

1. Had to delete two unassigned nicknames near the bottom of the file (they start Ub).

2. The only rung which turns on C24 is at rung 78. The one shot (rising edge) tests which are between this and the end of the program will work as expected. The ones at rungs 55 and 70 will not fire due to the way one-shots work in DL.

bernie_carlton said:

1. Had to delete two unassigned nicknames near the bottom of the file (they start Ub).

Click to expand...

I just renamed S110 to S110a...
-----------------------------------------------------------------------------
Glad to see you stopped in Bernie! I'm not a fan of stage programming (Can't port to other brands), so programs like these make my head spin...

I will add my 2¢ anyway...

I would try to avoid using transitional contacts. If I need a one-shot, I only use PD outputs. Those WILL turn on for a complete scan.

I also notice places where you have identical output conditions on separate rungs. For example, rung 55 and 56 perform the same function, but trigger on different conditions. I would rather see an OR condition on the left side of ONE rung.

STRPD C24
ORPD C26
RST C100
JMP S1

Although, I would create one-shots for C24 and C26 and skip the transitional contacts.

STR C24
PD Cx
STR C26
PD Cy
STR Cx
OR Cy
RST C100
JMP S1



-Eric

Just a quick note concerning the Block Call (BCALL). It's not like a subroutine where the current code stops while the called code executes. A Block Call merely enables a section of stages, specifically enabling the first one within the block. The block executes when the scan gets to it in the normal manner.

Thank you for the advice!

Bernie, I moved the stage that looks for change in c24 to the bottom of the program (to quickly check if that's the reason), but the problem persists. Thank you for the info - that's very close to how I imagined it would work, but definitely need to be extra careful with these.

Eric, thanks for the tip - I will OR the rungs. However, I don't see how I can use PD, they are output coils, I'm looking for a transition of an input ('servo ready' signal). Could you explain?

It's my first PLC program, I love the process so far and really appreciate any critique/suggestions. Thank you for your time.

However, I don't see how I can use PD, they are output coils, I'm looking for a transition of an input ('servo ready' signal). Could you explain?

Click to expand...

Like this:
PLC 250(-1)
// Rung 1
// Address 0
STR X0
PD C0
// Rung 2
// Address 2
STR C0
OUT Y20
// Rung 3
// Address 4
NOP

Ok, I got rid of all the transitional contacts and rearranged the whole program such that the flow is consistent. I still have the same problem! My brain is boiling.

I want the program to cycle through s1-s2-s3-s300-(servo block)-s1-... etc, but every time I try to flip either a table or a frame twice it gets stuck at the servo block and toggles on and off there forever.

Please help!

Yiou are still using a transitional of C24 in the program you posted on rung 59 which is before the rung which sets it (68). As I mentioned before, this won't work.

Yes, but these stages are at the next level - first need to figure out the basic flow (s1-s2-s3-s300), rung 59 is not checked at this point...

I can't follow the program very well. Does ' flip either a table or a frame twice ' involve any other stages other than S1, S2, S3, S300?

Yes, flipping is done in separate blocks (c500-502), where I only simulate a 'servo ready' signal, also there is an 'always active' s1000 stage and intermediate stages for clamping adjustments. I guess I should have taken out the stages irrelevant to this question, sorry about it. And thank you for helping!

The word 'flip' occurs in the ddocumentation in stages S410 and S411. The first of these stages contains the transition of C24 that I referred to. If this has nothing to do with it then please provide more complete information. I can only go by the words you use.

I would suggest getting rid of ALL transitional contacts unless it is referring to an input (X). Keep no others.

Here's a stripped down version of the program. I am building something like this - http://www.stryver.com/robots_09.jpg - so there is definitely a lot of flipping going on..