ETG1020 CommentFormV1i1i0D 2015-12-01

ETG.
1020 EtherCAT Protocol Enhancements

Please return form to [email protected]
ETG
member
Clause/
Paragraph
Subclause Figure/ Table
Type of
comment
Date
Document
18.11.2014
17.11.2015
01.12.2015
ETG1020_v1i1i0_S_D_ProtocolEnhancements.pdf
COMMENTS
Proposed change
OBSERVATIONS OF ETG
on each comment submitted
(General/
Technical/Editorial)
OMRON
Code0x006, what does Slave needs

BOOT-INIT transition means
Table1
I think it can be deleted because Slave

needs BOOT-INIT transition is only
specific action for Semi devices but the
code may also be used by other slaves
which may have different recovering
behavior, e.g. re-download, power cycle
Proposal:
Meaning:
Delete Slave needs BOOT-INIT
transition
Adapt:
SII/EEPROM information does not match
Firmware
Description:
if the slave checks if the SII/EEPROM
content matches the firmware, e.g.
process data description or revision
number, and detects a mismatch
Init -> PreOp, final state ErrInit
OMRON
Table1
ETG
Table 1
OMRON
Table 60
ETG
10
Code0x002E, Cycle time to small
Cycle time too small
Code 0x0070 Description wrong:

Detected Configured Module Ident List
(0xF030) and Configured Detected
Module Ident list (0xF050) does not
match
some Footnote numbers 2 are strange
(too big in size and wrong position)
correct
There are additional objects defined in

ETG.5001 that need this procedure
Add objects
0xF01x, 0xF02x and 0xF03x to the list of
objects
Correct
Accepted by TAB e-mail voting

10.12.2014
Proposal: accepted
10.12.2014
Proposal: accepted
10.12.2014
Proposal: accepted
10.12.2014
Proposal: accepted
10.12.2014
ETG
member
Clause/
Paragraph
Type of
comment
COMMENTS
Proposed change
OBSERVATIONS OF ETG
(General/
OMRON
OMRON
OMRON
11.1
11.2
11.2
What does 0-terminator mean?

Does it apply for ALL single entry with
flexible length? (OCTET_STRING,
UNICODE_STRING, ARRAY_OF_*,
VISIBLE_STRING, Byte stream,
international language string, etc)
Item 6
Item 7
That means some objects are allowed

not to support CompleteAccess. How
does a master/tool know which object
does not support CompleteAccess? By
ESI Object/Flags@SdoAccess?
I believe a tool/master should use
Object/Flags@SdoAccess
what should a master/tool use to identify

whether an object supports
CompleteAccess or not? The specific
DataTypes of entries in an object or ESI
Object/Flags@SdoAccess?
Proposal:
Entries with flexible length, as specified
in table 93, may have a shorter length
than given as the maximum size of the
entry.
For entries with data type
VISIBLE_STRING the following applies:
Write access
- Master can send a VISIBLE_STRING
without 0-terminator:
slave shall add the 0-terminator if the
written VISIBLE_STRING is shorter than
the maximum entry size
- Master can also write
VISIBLE_STRING with 0-terminator
Read access:
- If the master reads a VISIBLE_STRING
the slave should return the
VISIBLE_STRING with its actual length.
- Bytes following the 0-terminator shall
be zero.
10.12.2014
Proposal:
Add new:
Even when the slave supports complete
access there still may be objects that
cannot be accessed via complete
access. For descriptions about complete
access in ESI, see ETG2000.
If
Object/Flags@SdoAccess=CompleteAcc
ess in ESI, the device shall support
complete access to the object.
10.12.2014
Proposal:
See above
10.12.2014
ETG
member
Clause/
Paragraph
Type of
comment
COMMENTS
Proposed change
OBSERVATIONS OF ETG
(General/
OMRON
11.2
OMRON
11.2
OMRON
11.3
OMRON
13.1
Item 10
, also some other objects may not

support CompleteAccess. E.g. very big
objects with thousands bytes, supporting
CompleteAccess will require extra RAM
cost.
I believe ETG should allow other objects
not mentioned explicitly here not support
CompleteAccess. For those objects
without supporting CompleteAccess, it is
required to indicate in ESI by
Object/Flags@SdoAccess=SubIndexAc
cess (default value).
concerns about item 11, 12 of Subclause
11.2.
My assumption is the 2 rules are for
some special objects such as 0x10F3.
My concern is if the 2 rules apply for all
objects, the usability would be very bad
for a tool: it would receive data packages
without knowing how to decode them.
E
Note
typo in 11.3 title: SDI -> SDO
NOTE: A state machine for Diagnosis

Handling is under preparation
When the state machine is supposed to
be available?
OMRON
13.2
Table 27
SI 3
Writing = 0:
(support optional) the slave will clear all
messages
Proposal:
See above
10.12.2014
My proposal: add NOTE:

-- this applies only for 0x1F03, (and other
known ETG-defined objects)
-- this shall not be used for other objects.
Proposal:
The concerns are reasonable. But it is up
to the slave vendor if a complete access
makes sense for an object or not. An
EtherCAT master need to rely on the
structure information provided by the ESI
(or slave application).
10.12.2014
Proposal: accepted
10.12.2014
Proposal: the diagnosis state machine
handling is implicitly described by the
existing diagnosis chapters.
Delete Note
10.12.2014
Proposal:
Add
, i.e. resetting SI2, SI3, SI4 and Si5 Bit 5
10.12.2014
ETG
member
Clause/
Paragraph
Type of
comment
COMMENTS
Proposed change
OBSERVATIONS OF ETG
(General/
OMRON
Beckhoff
13.2
13.3.
ETG
14.3
OMRON
14.4.
OMRON
19.2.1.1.6
SEW
23
Table 27
SI 5
If any bit is changeable by write, SI_5 is

writable. For bits not changeable by write
of Bit0, Bit3 and Bit4, the slave shall
check with its respective default bit
values. If all are matched, the
slave responds positively
Table 28
Table 32
Proposal:
There can be additional data types for

DiagMessage parameter
Bit 0-11 = Data type Index of the data

type of parameter 1
Bit 0-11 = Data type Index of the data

type of parameter 1
0x0001: BOOLEAN; ESI specifier: %c
0x0002: INTEGER8; ESI specifier: %d
0x0005: UNSIGNED8; ESI specifier: %u
0x0006: UNSIGNED16;
ESI specifier: %u
0x0007: UNSIGNED32;
ESI specifier: %u
Data type IDs:

0x0001: BOOLEAN
0x0002: INTEGER8
0x0003: INTEGER16
0x0004: INTEGER32
0x0005: UNSIGNED8
0x0006: UNSIGNED16
0x0007: UNSIGNED32
0x0008 : REAL32
0x0011 : REAL64Ae
0x0015 : INTEGER64
0x001B : UNSIGNED64
Add:
When writing SI5, the readonly bits
should match the current values. Bit5
shall be "dont care".
The slave should send an abort with
0x6090030 Value exceeded in case the
readonly bits differ from current values.
10.12.2014
Proposal: accept
10.12.2014
The corresponding text parameters and

formatting are specified in the ETG.2000.
Typo
Change to TxPDO Parameter
Proposal: accept
Name: RxPDO Parameter
10.12.2014
the last extension is allowed to have size Please delete "end of entry" from the last Proposal: accept
of 240 according to Rule3.
2 examples
There is no need for "end of entry" in the
last 2 examples.
30.11.2015
renumber items starting from 1, not
Proposal: accepted
current 5
10.12.2014
Data types are already defined in
Add note that this chapter enhances the
Proposal: accepted
ETG.1000.6
data type definition of ETG.1000.6
10.12.2014
ETG
member
Clause/
Paragraph
Type of
comment
COMMENTS
Proposed change
OBSERVATIONS OF ETG
(General/
OMRON
23
Table 93
I have a concern about explanation in

Table93 for VISUAL_STRING. It may
conflict with the definition in ETG2000
and our discussions.
OMRON
23
Table 92, 93
94
OMRON
23
Table 93
Why do you put GUID for some

add a GUID for the other data types as
DataTypes, not for others in Table92, 93, well
94?
What are the GUID is supposed to be
used?
Table 93 contains strange texts, e.g
{18071995-????????????-xxxx}, in
explanations about Index 0x0009,
0x000A and 0x000B
OMRON
23
Table 93
2. Can GUID be used in ESI? If yes, it

will cause a compatible problem:
almost all tool can not support it.
To avoid the trouble, I would like to
propose adding a NOTE: using GUID in
ESI is reserved.
Please clarify
I propose to delete ByteSize = xxx+1 in

table 93 and adapt description for xxxx
as follows:
VISIBLE_STRING
with xxxx equals n (in HEX)
OCTET_STRING
with xxxx equals n+1 (in HEX)
UNICODE_STRING
with xxxx equals n+1 (in HEX)
10.12.2014
Proposal: accepted
10.12.2014
Proposal:
GUID is missing in the last cell of the
table heading.
xxxx is described in the corresponding
cell, i.e. "with xxxx (in HEX) is length of
data."
10.12.2014
Proposal: reject
There might be objects/entries, that are
of type GUID. E.g. we already defined
some in the EAP specification.
Therefore a definition of a GUID data
type is useful and should be allowed as
any other BaseDataType.
For the configuration tools: this is a
feature enhancement, which needs an
enhancement in the configuration tools.
However, in the moment there are no
standardizes objects in the EtherCAT
specs, hence it might not be necessary
to be supported any time soon in the
tools.
Agreed by TAB e-mail voting 30.11.2015

ETG1020 CommentFormV1i1i0D 2015-12-01

Uploaded by

Copyright:

Available Formats

ETG1020 CommentFormV1i1i0D 2015-12-01

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

ETG1020 CommentFormV1i1i0D 2015-12-01

Uploaded by

Copyright:

Available Formats

ETG.

1020 EtherCAT Protocol Enhancements

Code0x006, what does Slave needs

I think it can be deleted because Slave

Code0x002E, Cycle time to small

Cycle time too small

Code 0x0070 Description wrong:

There are additional objects defined in

Accepted by TAB e-mail voting

What does 0-terminator mean?

That means some objects are allowed

what should a master/tool use to identify

, also some other objects may not

typo in 11.3 title: SDI -> SDO

NOTE: A state machine for Diagnosis

My proposal: add NOTE:

If any bit is changeable by write, SI_5 is

There can be additional data types for

Bit 0-11 = Data type Index of the data

Bit 0-11 = Data type Index of the data

Data type IDs:

The corresponding text parameters and

I have a concern about explanation in

Why do you put GUID for some

2. Can GUID be used in ESI? If yes, it

I propose to delete ByteSize = xxx+1 in

You might also like