| Author |
Message |
GVS0001
New User
Joined: 07 Dec 2007
Posts: 2
|
| Dec 07, 2007 5:39 pm |
|
|
I like to communicate with a Grundfus pump using the GENIbus protocol. The protocol uses a CRC16 error correction. Although I have the specifications I do not manage to find out how the CRC is really calculated.
The specifications are:
Cyclic redundancy checking 16 bit CCITT, polynomial is 0x1021, Start Delimiter excluded.
Initialised to 0xFFFF, CRC value bit inverted after calculation.
High order byte transmitted first.
I have f.i. these two datagrams:
| Code: |
Start Delimiter 0x24
Length 0x10
Destination Address 0x01
Source Address 0x20
Class 2: Measured Data 0x02
Ack=0, Length=12 0x0C
Value example of i_rst INFO head 0x82
Value example of i_rst UNIT Index 0x3E
Value example of i_rst ZERO 0x00
Value example of i_rst RANGE 0x39
Value example of t_mo INFO head 0x82
Value example of t_mo UNIT Index 0x15
Value example of t_mo ZERO 0x00
Value example of t_mo RANGE 0x64
Value example of p_hi INFO head 0x82
Value example of p_hi UNIT Index 0x09
Value example of p_hi ZERO 0x00
Value example of p_hi RANGE 0xFA
CRC high 0x91
CRC low 0x0A |
second example:
| Code: |
Start Delimiter 0x27
Length 0x07
Destination Address 0x20
Source Address 0x01
Class 2: Measured Data 0x02
OS=3 (INFO), Length=3 0xC3
i_rst = ID 2 0x02
t_mo = ID 16 0x10
p_hi = ID 26 0x1A
CRC high 0x90
CRC low 0x1C |
I thought when I put these hexcodes behind each other (without the start byte) I would get a correct result from Lammert's calculator. But no I don't.
The hexstring I enter are:
| Code: | 100120020C823E003982150064820900FA
07200102C302101A |
Who knows what I am doing wrong? Or is it a special implementation of the CRC16CCITT?
Any help very much appreciated.
Geert |
|
GVS0001
New User
Joined: 07 Dec 2007
Posts: 2
|
| Dec 07, 2007 7:33 pm |
|
|
| After some more investigation I found out I have to invert the checksum the calculator gives for the 'CRC-CCITT (0xFFFF)' result. I then get the right checksum. Is this normal or am still doing something wrong? |
|
Guest
|
| Dec 07, 2007 10:52 pm |
|
|
Who can and will generate a lookup table for me and post it here?
I am not familiar with the C examples. I tried to compile the examples but they gave me the wrong result. |
|
regregex
Preferred Member
Joined: 30 Oct 2007
Posts: 119
Location: London, UK
|
| Dec 10, 2007 11:55 am |
|
|
I can tell you that the table you want begins 0x0000, 0x1021, 0x2042, 0x3063... and ends 0x1EF0. Plenty of examples online.
--G |
|
regregex
Preferred Member
Joined: 30 Oct 2007
Posts: 119
Location: London, UK
|
| Dec 11, 2007 10:09 pm |
|
|
| GVS0001 wrote: | | After some more investigation I found out I have to invert the checksum the calculator gives for the 'CRC-CCITT (0xFFFF)' result. I then get the right checksum. Is this normal |
Yes, as noted in the spec "CRC value bit inverted after calculation."
The algorithm corresponds to "CRC-16/I-CODE" (D64E). |
|
benj72
New User
Joined: 11 Feb 2009
Posts: 1
|
| Feb 11, 2009 11:37 pm |
|
|
I am a postgraduate student considering doing my major project on interfacing some Genibus pump controllers with a Zigbee mesh. Can anyone supply some specifications for the Genibus protocol?
cheers, Ben |
|
Raul
Guest
|
| Feb 18, 2009 12:57 am |
|
|
Hi Ben
Please, tell me what you need about Genibus to my email.
Cheers
rarranz at grundfos dot com |
|
FAQ
Search
Register
Log in
