Created Date: April 21,2009

Created by: bcarlton

1. Do you know the final desired position?

2. Can you control the closing speed in finer increments than just the two speeds?

If your answers to #1 and #2 are 'yes ' than can you calculate a ramp to a lower speed based on the difference between the current position and the final desired position? I don't know anything about this process to suggest a method.

Created Date: April 21,2009

Created by: Philco

Hi bcarlton,

Yes I know the final position. It must because the PLC must know when to continue his process (melt injection...)

And yes I can control in finer increment but it's kind of useless. I want the fastest motion possible with consistence shot to shot. By now the precision I want to use is thousand of an inch. I understand I could put somekind of 2 or more cushion zone with each their own speed but I thought that PLC could be fast enough to catch everything quickly. After all, I only have less than 200 rungs.

Created Date: April 22,2009

Created by: AZRoger

Possible short cut

Philco,

I read the excerpt of monic code in your first post. It looks like the logic that is "watching " the die close is pretty small, only a small fraction of the 200 rungs you mentioned for the whole process. If during the close die part of the operation you can ignore most inputs , you might be able to reduce your scan time by skipping over the unneeded rungs. But it's a big if. This focuses your program on just one task until it is done - just like when you focus on a lock as you insert the key. You don't want to miss anything important while you focusing on that one task. :)

There are two ways that I know of to really skip code. One uses BLK and BEND from the RLL-Plus (Stage) tool set (Chapter 7 in the manual). The other uses GOTO and LBL (Chapter 5). Both of these allow almost direct flow of execution around unneeded (for the moment) rungs of ladder.

Disabling a BLK causes execution to flow from the BLK to the BEND "ignoring " everything in between. There are side effects from Stage Programming that might cause you trouble if you haven't planned for this ahead of time. I don't see nicknames for Stages so more study is probably needed before you jump on this one.

Enabling a GOTO causes execution to flow from the GOTO to the LBL really ignoring everything in between. This technique is easier to stick in to an existing program. But you still need to be careful. If some sort of jam could prevent the die closing from ever finishing, you will still need some other way (a timer?) to expand the focus of the program again, some sort of fail safe.

My understanding is that instructions skipped using these methods take 0 scan time. They are not executed. In fact, they are not even looked at by the CPU during the scan. This should reduce your scan time by quite a bit during this very focused part of the process. I'd appreciate it if someone from AD would confirm that I'm not missleading Philco. Thanks. :D

Roger

Created Date: April 21,2009

Created by: Philco

Hi,

I have a plastic machine that have a DL-205 in it (a 260cpu). One part of my program use to control a proportionnal hydraulic valve which control the clamping. I also have a linear potentiometer attached to the clamping so it gives to the PLC the position of the clamping.

I have program my ladder to make the clamping slowing his speed when it reach a position but it's not consistent. Sometimes, he passed out that cushion position a bit so clamping close a bit hard.

I have put something in the ladder to see if it was the PLC fault or an electrical problem and I saw that it's the PLC. I 've attached the program monic. You will see the V2732 out box, this one is use to record the position of the clamping when the cushion sig came on. I have see a good variability from shot to shot.

I must say that the clamping velocity is pretty fast but not that much for a PLC I think.

Do someone have a better idea to make it more consistent?

If you need informations to help me out, please ask!

Thanks!