RE: Non Powered Devices

Yes,  I miss all the discussions.  Most lists I belong to now rarely have
any postings.  My question probably requires some in depth research.  My
products use the TI drivers designed by the late Steve Corrigan but recently
our systems are also using more and more CANopen products from others.
Since our system is battery operated shutting down non-essential units is an
attractive option for my client.  And I don't know what drivers are inside
many of the sealed products.
 
Many years ago I did an experiment to determine if CAN could work with only
two wires. I wired up two drivers on a breadboard and powered each one with
a 9V battery and 5V regulator.  So the grounds were totally isolated.  I
optically coupled the TTL I/O and then sent messages.  CAN does work with
just two wires but it's not very noise immune that way.  Steve confirmed
that when I showed him the results.  So my experiments, confirmed by the
chip designer, demonstrated that although it works, CAN, if you want robust
signalling _must_ have the ground wire too.  Yet now many web sites have
postings implying the ground isn't needed.
 
I guess a similar experiment needs to be done with a number of different CAN
devices.  Run signals on the bus and remove power from each device and
compare not only the signals but how much bleed through current shows up at
the 5V (or 3.3V) power supply.  So many projects.  So little time.
 
Thanks
John Dammeyer
 
From: [email protected] [mailto:[email protected]] On Behalf
Of Kees Zagers
Sent: July-04-18 5:24 AM
To: [email protected]
Subject: Re: [CANLIST] Non Powered Devices
 
Hi John,
Waking up the list again :-)
As far as I know this problem only occured at the old 250 driver. Since the
251 all CAN transceivers should be well protected and also don't influence
the bus if they are not powered. I mostly use NXP and Microchip transceivers
and did not notice any disturbance with non-powered devices connected to the
bus. They behave as very high impedance inputs.
Kees
 
Op 4-7-2018 om 02:19 schreef John Dammeyer:
Way back at the beginning it was pretty well a rule that all nodes on a CAN
bus should be powered.  With Steve Corrigan from TI no longer with us I
don't have access to a CAN driver designer to ask that question.   And there
are way more drivers out there than there were 25 years ago.
 
What's the general rule on that now?   Will an unpowered CAN driver
adversely affect the CAN bus signals and/or are there some that can even be
damaged?
 
Thanks
John Dammeyer
 
 
 
 
"ELS! Nothing else works as well for your Lathe"
Automation Artisans Inc.
www dot autoartisans dot com