RE: CAN-H and Vbatt Short

Hi

You had several questions.

A node goes error passive for either transmit or receive errors.  Each
transmit error increases the error counter by 8,  each subsequent
successful transmission decreases the transmit error counter by 1.  So one
transmitted error requires 8 good messages. To go Error Passive it takes
only 16 messages to reach 128 and 32 to reach bus off.

But if the messages are always in error all the time the node won't go bus
off because it doesn't know there isn't a receiver out there.  It just
knows the message it put on the transmit pin doesn't match what it read
back and will transmit forever.  This will appear like the node never
returns to Error Active.

Now a more specific question about shorting the driver to the =12V rail.
I can't see anything specific in the data sheet that says that the TJA1040

device is supposed to be able to work error free with the CAN_H connected
to the rail.

My own experience with CAN_H connected to 36V suggests that although some
messages might get through the system becomes incredibly unreliable.

To verify this I'd suggest you return to first principles and try an
experiment.

Connect two transceivers together between 10m cable with a termination on
each end by the transceiver.  Set Tx Level on one so that the bus is
passive and connect a signal generator to the other.  Create a 500kHz
square and inject that into the other transmitter.

Examine the waveform on the bus and at the Rx pin of each transceiver and
become familiar with their shape.   if you have a programmable function
generator you could even create other bit sequences to learn what the bus
looks like with 5 or 6 bits in a row.

Now.  Connect CAN_H to +12.  Compare the signals.  Both on the CAN bus and
on the Rx Pins.  How are they different?

Since a passive bus condition is has a differential receiver threshold
voltage of worst case 0.9V (page7) and you have 2.5V on CAN_L and 12V on
CAN_H my guess using simple math is your receiver thinks it sees a
dominant.  It really may never see a clear recessive.

The point I guess is not to connect 12V to the CAN_ line.  I can tell you
from experience 36V doesn't go well.  Our solution was to remove each node
one by one until the 12V went away.  The offending node was opened up and
we found a sharply trimmed lead from an LED poking into the CAN_H line.
In another case we had to pull apart the moulded cable at a location our
TDR said the error might be to find the short circuit in the harness.

The worst situation was the defective cable that coupled the larger 75V
power supply to the CAN devices that were rated at only 42V.  Those 50
devices failed and the CAN driver had to be replaced.

CAN system are robust if the signals look good on the scope.

John Dammeyer

"ELS! The Solution"
Automation Artisans Inc.
http://www.autoartisans.com/ELS/
Ph. 1 250 544 4950


-----Original Message-----
From: [email protected]
[mailto:[email protected]] On Behalf Of
PDE-Elect - Software Team 1
Sent: Tuesday, December 18, 2012 1:52 AM
To: 'CANLIST'
Subject: Re: [CANLIST] CAN-H and Vbatt Short



Hi Kees,

I am using TJA1040 CAN transceiver, And speed of the bus is 500kbps,
terminated by 120ohms. When i do this CAN-H and Vbat Short communication
is happening in proper. But some error frames are generating during short.
and this error frames leading to "Error Passive" state.

Hi John,

2 nodes are connected in the bus but my device has only Tx messages. No Rx
mesages from partner node.

This case is mentioned in the document "ISO 11898-2".

But first of all i want know whether this error will happen OR should n't
happen. If it happen pls give me some reference..

Regards,

BoopathiRaja.G




"Kees Zagers" <[email protected]>
Sent by: [email protected]


18 - Dec - 2012 01:22 PM


Please respond to
[email protected]



To
<[email protected]>

cc

Subject
Re: [CANLIST] CAN-H and Vbatt Short






Sorry forgot one thing.

I presumed you have a CAN high speed transceiver chip. If you have a low
speed transceiver chip and further low speed network and did the
termination in the right way, the network should work with the shortend
CAN Hi to battery.

Kees

----- Original Message -----
From: <[email protected]>
To: <[email protected]>
Sent: Tuesday, December 18, 2012 8:40 AM
Subject: Re: [CANLIST] CAN-H and Vbatt Short


> Hello BoopathiRaja.G,
>
> The CAN controller can have 3 states:
>
> Error Active (normal, Rx cnt < 128; Tx cnt < 128)
> Error Passive (Rx cnt > 127 or (Tx cnt > 127 and Tx cnt < 256)
> Bus Off (Tx cnt > 255)
>
> So your node is not in the Error Passive mode, but in the Bus Off mode.
> It will not recover from that, until a hard reset takes place.
>
> It is logical that it comes directly into the "Bus off" state when you
> shorten the CAN-Hi to the battery. It will try to send a pattern of 0's
> and 1's, but it will read back all 0's. CAN-Hi minus CAN-Lo will allways

> be more than 2 Volts in difference, so everytime it does a retry of the
> first message an error occurs; 32 retries and the controller goes
> bus-off.
>
> Hope this clarifies your situation.
>
> Kees Zagers
>
> PDE-Elect - Software Team 1 schreef op 18.12.2012 07:01:
> > Hi Techies,
> >
> > I am facing one issue in my CAN-physical layer design.
> >
> > I am using Fujitsu micro. I have a CAN messages only for Tx from my
> > micro. No Rx messages for my system.
> >
> > When i am shorting CAN-H line and Battery voltage(11V to 14 V) line,
> > my controller generates the "Error passive" error.
> >
> > Why this is happening.? And my design validation team doesn't agree
> > this point.
> >
> > And then Error Passive doen't recover back to Error Active State even
> > if i removed the Short.
> >
> >  Pls help me out Is there any protection circuit vailable to avoid
> > "Error passive " State. Or If it happens in protocol suggest me some
> > documents for reference..
> >
> > Regards,
> >
> >  BoopathiRaja.G
>
> --
> Archives and useful links: http://groups.yahoo.com/group/CANbus
> Subscribe and unsubscribe at www.vector.com/canlist/
> Report any problems to <[email protected]>