Hello everyone!

I have one strange problem with s7-1200 1212 DC/DC/DC

Machine is equipped with 3 phase motor and frequency converter for material adding option

So basically im running 3 phase motor with VFD and reading angular position with 100 PPR encoder

Everything is working, HSC counts, scan times for HSC inputs are lowest (20uS), I set up reference finding procedure and that works

I've done previously bending machine with very similar ladder and had no problems with it (2 machines!)

So problem is this...When I take ID1000 where is stored current HSC counting and do anything mathematically with it like converting to real, multiplying, dividing... whole PLC goes crazy

I mean, setting marker without any permission to do that, outputs turns on whenever they want, oscilalting outputs, whole myrad of events

It took 2 days for me to figure that this caused PLC to malfunction

I replaced VFD, PLC, I/O, tried different power supply, with grounding, without grounding, checked power supply and every single input with scope...everything seems ok

But when i try to manipulate with encoder measure...disaster

This wouldn't be problem if i don't need to read and input linear measure...I must multiply current count with encoder resolution to get linear measure and so on

As i said before, i did this on previous 2 machines and no such problems whatsoever, same PLC with same I/0s

Please help i don't have any more ideas

Tnx

Tnx for reply!

I downloaded to PLC againg and it says its all up to date so configuration is checked

Also there is excactly 100 pulses per encoder revolution, i checked now

Can i ask why is that important? I haven't anywhere defined how many pulses encoder has

Could it be some kind of arithmetic overflow because of 1212 CPU so something could be adjusted to ease little bit?

I tried again now to multiply ID1000 with -1 and as soon as i download that block to cpu markers that have no permission are set

UNIOR86

So problem is this...When I take ID1000 where is stored current HSC counting and do anything mathematically with it like converting to real, multiplying, dividing... whole PLC goes crazy

I mean, setting marker without any permission to do that, outputs turns on whenever they want, oscilalting outputs, whole myrad of events

It took 2 days for me to figure that this caused PLC to malfunction

 Can you show us the code you describe in the above quote? Are you saying you are doing mathematical operations on ID1000 itself?

Kal.A

UNIOR86

So problem is this...When I take ID1000 where is stored current HSC counting and do anything mathematically with it like converting to real, multiplying, dividing... whole PLC goes crazy

I mean, setting marker without any permission to do that, outputs turns on whenever they want, oscilalting outputs, whole myrad of events

It took 2 days for me to figure that this caused PLC to malfunction

 Can you show us the code you describe in the above quote? Are you saying you are doing mathematical operations on ID1000 itself?

The last reply by Kal.A inspires another question ->
Is %ID1000 declared as a DINT (as it should be) or DWORD (default by TIA)?

Yes ID1000 is DINT which then i multiply by -1 for other direction, then DINT to REAL conversion and finally multiplication with 0.05 for correct linear measure, or at least i would like to because PLC malfunctions as soon as i insert any of those blocks

This is demonstration of what happens when any of those blocks are downladed to CPU:

[snip]

You can see that M0.0 marker is set by EMG pushbutton but only M0.0 in Networks 1 and 2 are set

In other networks are not but its the same marker

Sometimes few outputs oscillate, turn on and off randomly, different markers are set or reset without any permission

I installed update 4 yesterday and same thing is present

UNIOR86

[snip]
https://drive.google.com/drive/folders/ ....
[snip]

You may find some of us do not evaluate off-site sources.
Sincerely: Good luck finding your solution.

UNIOR86

Yes ID1000 is DINT which then i multiply by -1 for other direction, then DINT to REAL conversion and finally multiplication with 0.05 for correct linear measure, or at least i would like to because PLC malfunctions as soon as i insert any of those blocks

This is demonstration of what happens when any of those blocks are downladed to CPU:

[snip]

You can see that M0.0 marker is set by EMG pushbutton but only M0.0 in Networks 1 and 2 are set

In other networks are not but its the same marker

Sometimes few outputs oscillate, turn on and off randomly, different markers are set or reset without any permission

I installed update 4 yesterday and same thing is present

 I would like to see the calculation part of ID1000 to understand what you mean.

ID000 := ID1000 * -1  would not work as ID1000 is constantly updated.

MD100 := ID1000 *  -1 would be OK

But

MD0 := ID1000 *-1  is a problem because it creates a memory overlap between MD0 and M0.0 to M0.7, M1.0 to M1.7,  M2.0 to M2.7, 3.0 to M3.7

NOTE: I tired to add attachment to show how to view an overlap of memory but it looks like attachments are not possible as I tried several file types and sizes.

Hello.

I am confused how the behaviour of %M0.0 and others relates to the original question. For this situation I am wondering if the special bits are activated in the hardware configuration which now overlaps with these bits.

Might be a good idea to move the bits to base byte 50 (like %M50.0) which would also cover what Kal.A responded to.

@UNIOR86 - It would be to your own advantage to discover how to use global DBs for memory storage. More advantage if you split your work into manageable parts, coded in FBs and FCs.

BR