[etherlab-users] rtai_rtdm_dc example, "error: system_time_base less than system time" message

Mon Feb 8 22:28:06 CET 2016

Interesting, it shouldn’t matter which order the modules are in, but for a few exceptions.

Some of those exceptions being:

1) You can’t have more than two Passive Terminals grouped together (see Mounting of Passive Terminals in any of the Beckhoff documents)

2) Some modules such at the relay output module EL2612 don’t have power contacts, so need to be placed at the end of the terminal group

3) Make sure you don’t exceed the power capabilities of the EK1100 module.  This can provide 2000mA.  Each active terminal consumes some of that power.  This can be calculated by using the `ethercat slaves -v` command and checking each modules “Current consumption” value.  The EL1008 modules use 90mA each, so 17 of these would be fine, but it sounds like you have other modules which may use more (eg: an EL2008 uses 110mA, an EL5101 uses 130mA, an EL2612 uses 150mA etc).

Note: you can also use the command `ethercat slaves -v | grep “Current consumption”` for a quick summary.  The EK1100 module reports -2000mA because it is supplying rather than consuming.

If you have gone over the available current then you will need to insert an EL9410 power module (or similar).

Regards,
Graeme.

From: Bilko AS, Oguz Dilmac [mailto:odilmac at bilko-automation.com]
Sent: Saturday, 6 February 2016 1:42 a.m.
To: Graeme Foot; etherlab-users at etherlab.org
Subject: Re: [etherlab-users] rtai_rtdm_dc example, "error: system_time_base less than system time" message

Hi,

We solved the problem last week.
It turned out we have to place the same kind of bechoff IO modules next to each other.

In the group of modules which connected to a EK1100 we had 17 modules. In this group, we had four EL1008 one after the other; Then a few other modules; Then one EL1008; again afew more modules and finally two more EL1008. When we group all the EL1008 modules in a row, the problem disappeared.

Best regards,
Oguz.
26.1.2016 01:49 tarihinde Graeme Foot yazdı:
Hi,

The “system_time_base less than system time” is from the rtai_rtdm_dc example code.

In the example it looks like system_time_base is initialised to zero and adjusted by a maximum of +-1001 each cycle.  A value of -58905492552780 indicates that system_time_base is not being correctly initialised or is being corrupted with any other values.

As to the motor “knock” every now and then, this is due to the motors dc clock not being correctly synchronized with the masters cycle time.  It is due to the master either missing or sending two messages to the amps in one cycle every now and then.  Sort out your dc timing and the knock will go away.

Regards,
Graeme.

From: etherlab-users [mailto:etherlab-users-bounces at etherlab.org] On Behalf Of Bilko AS, Oguz Dilmac
Sent: Tuesday, 26 January 2016 4:37 a.m.
To: etherlab-users at etherlab.org<mailto:etherlab-users at etherlab.org>
Subject: [etherlab-users] rtai_rtdm_dc example, "error: system_time_base less than system time" message

Hi,

We have been using EtherCAT master 1.5.2 for about two years. We have successfully implemented 4 machines. Each has 5 Kollmorgen AKD servo drives and some Beckhoff IO modules.
Now we are implementing a new machine. On this machine we usually get the following messages:
...
"system2count() error: system_time_base less than system time (system_time_base: -58905492552780, time:680659495658"
"system2count() error: system_time_base less than system time (system_time_base: -58905492552858, time:680660495658"
"system2count() error: system_time_base less than system time (system_time_base: -58905492552936, time:680661495658"
"system2count() error: system_time_base less than system time (system_time_base: -58905492553014, time:680662495658"
...

We implemented our system based on the rtai_rtdm_dc example. Only main difference is, we use kernel modules with RTAI instead of LXRT user mod.
We are using "cyclic positon" (8) mode with distributed clock.
We choose the first DC capable device as the reference clock by calling ecrt_master_select_reference_clock(master, 0);
The first node on the bus is a AKD servo drive.

The error occurs from the startup and not disappear until we reboot the PC and the drives.
When the error occurs, it seems sometimes drives can't update the new position reference. Therefore, if it is moving, it tries to stop for one sample and continues to the new reference at the next sample.
This causes a loud "knock" sound on the machine.

We didn't get this message with previous 4 machines.
We tried different FW versions of the AKD drives. Also tried to change the bus node order of the drives. But nothing changed.
Do you know what can be the cause for this message? How can we fix the problem?

Best regards,
Oguz.
--

Oguz Dilmac

ARGE Bolumu

Bilko AS, R&D Department

====================================

Perpa Ticaret Merkezi B Blok Kat 13 Nr. 2568

TR-34384 Okmeydani Istanbul Turkey

Tel : +90 212 220 07 40  Fax :   +90 212 210 47 01

e-mail : odilmac at bilko-automation.com<mailto:odilmac at bilko-automation.com>

web site : http://www.bilko-automation.com

_______________________________________________

etherlab-users mailing list

etherlab-users at etherlab.org<mailto:etherlab-users at etherlab.org>

http://lists.etherlab.org/mailman/listinfo/etherlab-users

--

Oguz Dilmac

ARGE Bolumu

Bilko AS, R&D Department

====================================

Perpa Ticaret Merkezi B Blok Kat 13 Nr. 2568

TR-34384 Okmeydani Istanbul Turkey

Tel : +90 212 220 07 40  Fax :   +90 212 210 47 01

e-mail : odilmac at bilko-automation.com<mailto:odilmac at bilko-automation.com>

web site : http://www.bilko-automation.com

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.etherlab.org/pipermail/etherlab-users/attachments/20160208/a73f9dee/attachment-0003.htm>