Hello.

I work with  ending an a old LOGO! 7 project (wit a 10" HMI).

I start with larger version, which contains traditional blocks and UDF blocks and what tied together totally more than 250 blocks. Because it worked in a simulator but not actually in the inside controller (the machine did not work properly), I started to change the program, because the only purpose is to get it to work. In the meantime, I also updated the software to version 8.3

Now, it's a month with working wit it. I've seen how things are going well in the simulator and the program crashes totally on a real machine; or anomalies (machine movements) occur that are not present in the program. I have seen, in PC monitor real time connection, how connections between UDFs  inside the program are going into resonance /blinking on a working time.

Today's situation is that the machine like works..Totally used 228 bloks. But, if issuing multiple work orders in a row (without pause), then controller go to madness (it works, but totally wrong counter numbers and etc).

Hence the questions:
* It is possible, that controller stay in box 5 year and it is not longer really good?
* It is possible, when the program crashes totally in first time (about the max 1 min work in crash condition), then it ruined something inside the controller and as a result, it will not work longer error free condition?

Thanks in advance..

erku

But, if issuing multiple work orders in a row (without pause), then controller go to madness (it works, but totally wrong counter numbers and etc).

Hello.

This sounds like an overlap of processing or so called timing glitches. In simulation it is more difficult to produce triggers (work orders) without pause timing. The pauses are long enough for the processing cycle to complete. Also the trigger is cleanly on and cleanly off.

I suggest you build in some counters and timers for the purpose of understanding the influence of multiple work orders that comes through without any pauses.

Below and attached is one timing glitch that shows that the memory bit block delays the signal with one scan time. The attachment is an animation.

Best health. 

Thanks for the answers.

Yes, I agree that the program is too big and I started to reduce it, as much as possible. Some of questions stay, about the software and work principles.

Remained question about signal transmission  (file in attachment). Signal transmit of one UDF to other UDF.
Inside attachment is shown two blocks, B004 and B003 (And-Edge, Wiping relay). Of one these, And-Edge go pulse Q1 out to another UDF always (no errors detected at all) but if change Wiping relay to And-Edge, then the pulses do not go (not at all, or incorrectly- sometimes yes some no). I tried to use different variations before Q1 but the result was still the same.
Question stay, why can not use 2  And-Edge before Q1 output in one UDF?
Maybe there's a rule for that?