Searching \ for '[EE] CAN questions' in subject line. ()
Make payments with PayPal - it's fast, free and secure! Help us get a faster server
FAQ page: www.piclist.com/techref/index.htm?key=can+questions
Search entire site for: 'CAN questions'.

Exact match. Not showing close matches.
PICList Thread
'[EE] CAN questions'
2007\02\20@124855 by Sean Breheny

face picon face
Hi all,

I have some questions about CAN pricing that I hope someone here can
answer. All of these apply to a commercial product.

1) If I implement CAN in software (i.e., do not buy any CAN hardware
except a CAN transceiver like Microchip's MCP2551), so I have to pay
Bosch or anyone else royalties?

2) If I use a PIC or other microcontroller which includes a CAN
peripheral, do I have to pay Bosch royalties?

3) If I implement CAN in an FPGA (I do not buy any CAN hardware except
a transceiver), do I have to pay Bosch royalties?

Thanks,

Sean

2007\02\20@134539 by Vasile Surducan

face picon face
On 2/20/07, Sean Breheny <spam_OUTshb7TakeThisOuTspamcornell.edu> wrote:
> Hi all,
>
> I have some questions about CAN pricing that I hope someone here can
> answer. All of these apply to a commercial product.
>
> 1) If I implement CAN in software (i.e., do not buy any CAN hardware
> except a CAN transceiver like Microchip's MCP2551), so I have to pay
> Bosch or anyone else royalties?
>
> 2) If I use a PIC or other microcontroller which includes a CAN
> peripheral, do I have to pay Bosch royalties?
>
> 3) If I implement CAN in an FPGA (I do not buy any CAN hardware except
> a transceiver), do I have to pay Bosch royalties?
>

Does every TV manufacture pay something to Philips for the I2C standard ?

Maybe here is the answer:
http://www.can-cia.org/can/

2007\02\20@134655 by Gerhard Fiedler

picon face
Sean Breheny wrote:

> Hi all,
>
> I have some questions about CAN pricing that I hope someone here can
> answer. All of these apply to a commercial product.
>
> 1) If I implement CAN in software (i.e., do not buy any CAN hardware
> except a CAN transceiver like Microchip's MCP2551), so I have to pay
> Bosch or anyone else royalties?
>
> 2) If I use a PIC or other microcontroller which includes a CAN
> peripheral, do I have to pay Bosch royalties?
>
> 3) If I implement CAN in an FPGA (I do not buy any CAN hardware except
> a transceiver), do I have to pay Bosch royalties?

This is what I found:
<http://www.semiconductors.bosch.de/en/20/can/2-license.asp>:

------------------------------------------
The CAN Protocol is developed by Robert Bosch GmbH and protected by
patents. Additionally to the CAN IP modules offered by Bosch, a CAN
Protocol License is required. The CAN Protocol License is also required for
self-developed CAN modules, or for CAN modules purchased from another
vendor. Bosch is licensing the CAN protocol as follows:

1. CAN Protocol License for ASIC-manufacturers and FPGA mass programming
Performance of Bosch:

   * Delivery of the CAN Protocol Specification together with
comprehensive explanations
   * Delivery of a functional C Reference CAN Model
   * Granting the right to use the CAN Know-How and the CAN patents for
the manufacture of ASICs or the programming of FPGAs with CAN and their
worldwide sale

2. CAN Protocol License for FPGA Programming
Performance of Bosch:

   * Granting the right to use the CAN patents for programming of FPGAs
with CAN and their worldwide sale

For License Issues please contact our Sales staff by using the E-Mail form
sheet under "Contact"
------------------------------------------

Not sure how relevant it is that the CAN bus protocol is an ISO
specification. It probably is not relevant for patented ideas, but it may
be for the rest that's not patented.

Gerhard

2007\02\21@043717 by Alan B. Pearce

face picon face
>Does every TV manufacture pay something to Philips for the I2C standard ?

Manufacturers of I2C chips pay the license, and Philips then take it that
using a licensed I2C chip in a circuit pays the licence for that circuit.

2007\02\21@045157 by YAP

picon face
On 2/20/07, Sean Breheny <.....shb7KILLspamspam@spam@cornell.edu> wrote:
> Hi all,
>
> I have some questions about CAN pricing that I hope someone here can
> answer. All of these apply to a commercial product.
>
> 1) If I implement CAN in software (i.e., do not buy any CAN hardware
> except a CAN transceiver like Microchip's MCP2551), so I have to pay
> Bosch or anyone else royalties?

Yes in this case I think you have to pay royalties.

>
> 2) If I use a PIC or other microcontroller which includes a CAN
> peripheral, do I have to pay Bosch royalties?

No need for that. Microchip done that for you.

>
> 3) If I implement CAN in an FPGA (I do not buy any CAN hardware except
> a transceiver), do I have to pay Bosch royalties?

Yes you have to pay royalties in this case.

/Ake

--
---
Ake Hedman (YAP - Yet Another Programmer)
eurosource, Brattbergavagen 17, 820 50 LOS, Sweden
Phone: (46) 657 413430 Cellular: (46) 73 0533 146
Company home: http://www.dofscandinavia.com
Kryddor/Te/Kaffe: http://www.brattberg.com
Personal homepage: http://www.dofscandinavia.com/akhe
Automated home: http://www.vscp.org

2007\02\21@045459 by Mark Rages

face picon face
On 2/21/07, Alan B. Pearce <A.B.PearcespamKILLspamrl.ac.uk> wrote:
> >Does every TV manufacture pay something to Philips for the I2C standard ?
>
> Manufacturers of I2C chips pay the license, and Philips then take it that
> using a licensed I2C chip in a circuit pays the licence for that circuit.
>

I believe the I2C patents have expired.

Regards,
Mark
markrages@gmail
--
You think that it is a secret, but it never has been one.
 - fortune cookie

2007\02\21@045726 by Wouter van Ooijen

face picon face
> >Does every TV manufacture pay something to Philips for the
> I2C standard ?
>
> Manufacturers of I2C chips pay the license, and Philips then
> take it that
> using a licensed I2C chip in a circuit pays the licence for
> that circuit.

Alternatively they use a prorietary 2-wire proctocol. Accidentally, that
protocol is 100% compatible with I2C.

Wouter van Ooijen

-- -------------------------------------------
Van Ooijen Technische Informatica: http://www.voti.nl
consultancy, development, PICmicro products
docent Hogeschool van Utrecht: http://www.voti.nl/hvu


2007\02\21@081330 by Gerhard Fiedler

picon face
YAP wrote:

> On 2/20/07, Sean Breheny <.....shb7KILLspamspam.....cornell.edu> wrote:
>> I have some questions about CAN pricing that I hope someone here can
>> answer. All of these apply to a commercial product.
>> [...]
>> 2) If I use a PIC or other microcontroller which includes a CAN
>> peripheral, do I have to pay Bosch royalties?
>
> No need for that. Microchip done that for you.

Do you have a source for this? Bosch says:

"The CAN Protocol License is also required [...] for CAN modules purchased
from another vendor."

My understanding is that the Microchip CAN module could be seen as a "CAN
module purchased from another vendor".

Another question could be whether it matters if the product is advertised
as using CAN or having a CAN interface, or just uses it internally.

Gerhard

2007\02\21@090828 by YAP

picon face
I'm certain I have seen something in the Microchip docs about that.
Question is where...

Anyway I don't know anyone that pay any royalties when they use
Micochip, Texas, Philips, ST etc chips.

Cheers
/Ake

On 2/21/07, Gerhard Fiedler <EraseMElistsspam_OUTspamTakeThisOuTconnectionbrazil.com> wrote:
{Quote hidden}

> -

2007\02\21@091429 by Wouter van Ooijen

face picon face
> Another question could be whether it matters if the product
> is advertised
> as using CAN or having a CAN interface, or just uses it internally.

For patented technology it is only the internald. For a TM it is only
the naming.

Wouter van Ooijen

-- -------------------------------------------
Van Ooijen Technische Informatica: http://www.voti.nl
consultancy, development, PICmicro products
docent Hogeschool van Utrecht: http://www.voti.nl/hvu


2007\02\22@214515 by Vitaliy

flavicon
face
Wouter van Ooijen wrote:
>> >Does every TV manufacture pay something to Philips for the
>> I2C standard ?
>>
>> Manufacturers of I2C chips pay the license, and Philips then
>> take it that
>> using a licensed I2C chip in a circuit pays the licence for
>> that circuit.
>
> Alternatively they use a prorietary 2-wire proctocol. Accidentally, that
> protocol is 100% compatible with I2C.

I'm curious what the official rules for that are. So if I were to take an
existing protocol, add a couple of things, but keep my new protocol 100%
backwards compatible, would that be illegal in most jurisdictions?

Vitaliy

2007\02\22@223517 by John Dammeyer

flavicon
face
Hi Vitaliy
> I'm curious what the official rules for that are. So if I
> were to take an
> existing protocol, add a couple of things, but keep my new
> protocol 100%
> backwards compatible, would that be illegal in most jurisdictions?
>
> Vitaliy


Bosch has done an odd thing with the CAN licensing.  When you buy into
the CAN license you get source code for doing your own VHDL gate array
plus test suites etc. for verifying that you conform to the standard.
All well and good.  You can include CAN into your custom product with
minor licensing costs.

But,  and it's a huge 'BUT' any changes you make to the protocol become
the property of BOSCH and are automatically shared with all other
license holders.  So if you designed an upgrade to the CAN protocol to
have larger block sizes for example, that protocol becomes the property
of Bosch.  You can't license the enhancements to anyone else because
anyone who's making CAN stuff automatically gets access to it in their
existing license.

What that has done of course is stifled any upgrades or improvements to
CAN and it's stayed at 2.0B now for almost 20 years.  Not bad for a
protocol and as a result that stability has it installed in millions of
products world wide including outside the world in space.

So stability in exchange for no upgrades.  It will be interesting to see
what happens when the CAN2.0B patent runs out.  

I worked on the design of a SUPERCAN system that was 100% compatible
with existing CAN messages on the bus at the same time but could run
5000 metres at 2.6mbps with messages up to 4K in size.  We ran real time
MPEG4 video over power line modems with it.  Alas the tech bubble burst
and so did the company.  And because of the way the NDA was done I can't
use the information.  Sigh....

My CANRF product would have also benefited from a few tiny upgrades to
the existing CAN protocol.  It worked with MCP2510s and SJA1000s etc.
but it could have worked better with a few changes.  Longer ACK bit
windows.  No Error Flags.

John Dammeyer



2007\02\23@030934 by Wouter van Ooijen

face picon face
> I'm curious what the official rules for that are. So if I
> were to take an
> existing protocol, add a couple of things, but keep my new
> protocol 100%
> backwards compatible, would that be illegal in most jurisdictions?

The law knows no protocols. It knows trademarks, copyright, patents,
specific contracts (non-disclosure), etc. So the situation would depend
on how the  protocol is protected, how 'specific' (different from
everything else) it is, and exactly what you copied. You might for
instnace end up with a protocol that is fee, but a description that is
not. Or one that can't be called by its original name. Or a protocol
that is free but a creator who has violated a contract.

Wouter van Ooijen

-- -------------------------------------------
Van Ooijen Technische Informatica: http://www.voti.nl
consultancy, development, PICmicro products
docent Hogeschool van Utrecht: http://www.voti.nl/hvu


2007\02\23@041440 by Alan B. Pearce

face picon face
>What that has done of course is stifled any upgrades or
>improvements to CAN and it's stayed at 2.0B now for almost
>20 years.  Not bad for a protocol and as a result that
>stability has it installed in millions of products
>world wide including outside the world in space.

I can vouch for that. Smart-1 (that got pranged into the moon last Sept)
used 2 CAN busses that I am aware of, one for system bus and one for payload
bus. The instrument I am currently working on also uses CAN as its payload
bus, as it is a development of the instrument we had on Smart-1.

2007\02\23@062900 by Gerhard Fiedler

picon face
Vitaliy wrote:

>> Alternatively they use a prorietary 2-wire proctocol. Accidentally, that
>> protocol is 100% compatible with I2C.
>
> I'm curious what the official rules for that are. So if I were to take an
> existing protocol, add a couple of things, but keep my new protocol 100%
> backwards compatible, would that be illegal in most jurisdictions?

As Wouter says, it depends :)

If, for example, the protocol or parts of it are protected by patents, it
would be illegal in most if not all jurisdictions where these patents are
valid. For a patented technology, it doesn't matter how you found out about
it: clean room approach or straight copy, it's all covered by the patent. I
think Bosch has patents on the CAN protocol.

If the protocol's name is protected by trademark, you may be able to do
legally what you said. You wouldn't be able to say that it is "<original
name>(tm)", but probably that it is "<original name>(tm) compatible".

Gerhard

2007\02\23@122105 by John Dammeyer

flavicon
face
It's really a business decision.  If you make a protocol that swims like
a duck, looks like a duck and quacks like a duck Bosch may not come
after you until you've been successful enough to justify the legal costs
based on their lost income.

OTOH,  if it were a certain OS manufacturer, they'd lose money just to
punish you for even trying.

John Dammeyer


{Quote hidden}

> --

More... (looser matching)
- Last day of these posts
- In 2007 , 2008 only
- Today
- New search...