View previous topic :: View next topic |
Author |
Message |
kender
Joined: 09 Aug 2004 Posts: 768 Location: Silicon Valley
|
CAN bus optical isolation |
Posted: Mon Aug 23, 2004 5:22 pm |
|
|
Colleagues,
I'm concieving a medical instrument with multiple PICs, which are talking to each-other through a CAN bus. I need to isolate one of the PICs. Is it possible to optically isolate a part of the CAN bus? As far as I understand CAN bus is differential, and doesn't have a dedicated TX and RX lines. Optical isolation, however is directional. Could anyone point me to the appnone or a schematic?
Nick |
|
|
Yashu
Joined: 08 Oct 2003 Posts: 26
|
Re: CAN bus optical isolation |
Posted: Tue Aug 24, 2004 8:51 pm |
|
|
kender wrote: | Colleagues,
I'm concieving a medical instrument with multiple PICs, which are talking to each-other through a CAN bus. I need to isolate one of the PICs. Is it possible to optically isolate a part of the CAN bus? As far as I understand CAN bus is differential, and doesn't have a dedicated TX and RX lines. Optical isolation, however is directional. Could anyone point me to the appnone or a schematic?
Nick |
idea.... mag isolation with NVE Corp IL710T between PIC and CAN transceiver and use DC-DC converter to power the transceiver |
|
|
treitmey
Joined: 23 Jan 2004 Posts: 1094 Location: Appleton,WI USA
|
|
|
kender
Joined: 09 Aug 2004 Posts: 768 Location: Silicon Valley
|
Re: CAN bus optical isolation |
Posted: Sun Dec 19, 2004 3:17 am |
|
|
Quote: |
idea.... mag isolation with NVE Corp IL710T between PIC and CAN transceiver and use DC-DC converter to power the transceiver
|
Checked the IL710T datasheet - it looks as directional as an opto-coupler.
Thanks for the idea though!
Nick[/quote] |
|
|
libor
Joined: 14 Dec 2004 Posts: 288 Location: Hungary
|
|
Posted: Sun Dec 19, 2004 4:31 am |
|
|
I remembered reading an article about galvanic isolation of the I2C bus, which is in this this regard has the same challenge having a bidirectional data line. The article shows and explains how to use two optocouplers to split the bidirectional dataflow and recombine it. This might be of help to you, if you can adapt it to the diffential signal lines somehow (if all else fails maybe with two CAN transceivers, one on each side)
Here is the article. |
|
|
libor
Joined: 14 Dec 2004 Posts: 288 Location: Hungary
|
|
Posted: Sun Dec 19, 2004 5:11 am |
|
|
A follow up:
You can't spare using CAN-bus transceivers on each side of the isolation border. These transceivers have splitted unidirectional data lines on their host side anyway, so there is no need for tricking this bidirectional thing into the circuit, just use one optocoupler on each direction per side. A total of four will be needed for one link. (you may have up to two or even four of them in one package)
This is also the way industrial CAN bus products are made, look at this Application Note at agilent.com: AN 1321- High Speed CMOS Optocoupler Applications in Industrial Field Bus Networks (search for "AN 1321" on their site) |
|
|
Sandman
Joined: 26 Jan 2004 Posts: 15 Location: Kiruna, Sweden
|
|
Posted: Sun Dec 19, 2004 11:46 am |
|
|
Do you have to isolate the CAN-transceiver from the bus? Why?
If not: you could use ordinary optical (like opto-couplers) or digital (like; ADUM1XXX from analog div.) isolators between the PIC and the transceiver. This method is used in space application where redundancy and security is a major design factor which it also is in the medical industry.
The CAN transceivers are usually very good when it comes to stay functional in unfriendly environments.
Last edited by Sandman on Sun Dec 19, 2004 12:22 pm; edited 1 time in total |
|
|
Yashu
Joined: 08 Oct 2003 Posts: 26
|
|
Posted: Sun Dec 19, 2004 12:01 pm |
|
|
libor wrote: | A follow up:
You can't spare using CAN-bus transceivers on each side of the isolation border. These transceivers have splitted unidirectional data lines on their host side anyway, so there is no need for tricking this bidirectional thing into the circuit, just use one optocoupler on each direction per side. A total of four will be needed for one link. (you may have up to two or even four of them in one package)
This is also the way industrial CAN bus products are made, look at this Application Note at agilent.com: AN 1321- High Speed CMOS Optocoupler Applications in Industrial Field Bus Networks (search for "AN 1321" on their site) |
thanks for elaborating on what I assumed was obvious. |
|
|
asmallri
Joined: 12 Aug 2004 Posts: 1634 Location: Perth, Australia
|
Re: CAN bus optical isolation |
Posted: Sun Dec 19, 2004 5:09 pm |
|
|
kender wrote: | Colleagues,
.... Optical isolation, however is directional....Nick |
For the record this is no longer the case. With current generation optical multiplexers and transceivers, bidirection comms can be implemented over a single strand of fibre. _________________ Regards, Andrew
http://www.brushelectronics.com/software
Home of Ethernet, SD card and Encrypted Serial Bootloaders for PICs!! |
|
|
kender
Joined: 09 Aug 2004 Posts: 768 Location: Silicon Valley
|
|
Posted: Tue May 03, 2005 11:02 am |
|
|
I see: the CAN-to-plain-serial chips are unavoidable. Although, CAN is becoming increasingly popular, and hopefully some day some company will make a CAN optoisolation chip, which would integrate all the necessary stuff.
Nick |
|
|
MikeW
Joined: 15 Sep 2003 Posts: 184 Location: Warrington UK
|
|
|
kender
Joined: 09 Aug 2004 Posts: 768 Location: Silicon Valley
|
|
|
kender
Joined: 09 Aug 2004 Posts: 768 Location: Silicon Valley
|
|
Posted: Thu May 05, 2005 1:51 am |
|
|
By the way:
Is there a standard connector for CAN?
Is there a free version of a CAN bus spec? |
|
|
PCM programmer
Joined: 06 Sep 2003 Posts: 21708
|
|
|
kender
Joined: 09 Aug 2004 Posts: 768 Location: Silicon Valley
|
|
Posted: Sat May 21, 2005 12:31 am |
|
|
PCM programmer,
I have requested the "CiA DR 303-1 V1.3: CANopen cabling and connector pin assignment" on the CiA web site, but they didn't send it to me. Could you send this document to me?
Nick
alexeev@stanford.edu |
|
|
|