RE: Non Powered Devices

I think you can also take it one step further and ask the question whether or not the CAN interface on the ESP32 meets specifications.  At one point in time Luca Stagnaro at ESTEC published his HurrinCANe  VHDL core code.  Unfortunately I never did get all of it and shortly after 2000 it was removed from publication.  I downloaded it mainly to see how it worked.  Apparently it had a few bugs and licensing became an issue as more and more people started using FPGAs.

Back then logic analyzers and scopes couldn't decode CAN messages so I had to write a special module for the logic analyzer I was using so we could verify the SuperCAN version we were developing matched the design specifications.  SuperCAN was what CANFD is now except that CANFD can't sent 1150 byte Ethernet packets and we could.   Anyway there are a lot of tiny details to correctly decode CAN messages at the bit level.

But nothing ever completely vanishes from the net so it's always possible that some variation of CAN 2.0B in VHDL exists somewhere.  But given the repeated similar mistakes people keep making with High Level Protocols (HLP) on CAN I'd not hold my breath that the ESP32 implementation works correctly.  

And it's those once every 10,000 messages that break a system or even potentially hurt someone that are impossible to find if the HLP is poorly designed.  Bring that down to the bit level and who knows.  The licensing per part is really so low that Expressive could afford it.  That they choose not to makes me wonder if their CAN works properly and it's too expensive to fix it.

John Dammeyer




> -----Original Message-----
> From: [email protected] [mailto:[email protected]] On
> Behalf Of Kees Zagers
> Sent: August-06-18 1:03 AM
> To: [email protected]
> Subject: Re: [CANLIST] Non Powered Devices
> 
> I think we all agree that the ESP32 is a nice low-cost controller, but
> let us go back to the first question about the licensing of the CAN
> interface. We see that Chinese chip builders like Allwinner (A20) and
> Espressif (ESP32) integrate CAN in their chips, however don't promote
> it. It is not mentioned in their basic features. Companies like Olimex
> detect such hidden features and make "open source" hardware for it, so
> it can be used. The world knows about the hidden feature, but does not
> know if the CAN interface is really licensed by the chip manufacturer.
> What about the smart device I sell to the automotiv industry in
> quantities? I'm still expecting that the chip manufacturer paid the
> licensing costs to Bosch by rule, but what if they did not?
> 
> Kees
> 
> 
> Op 6-8-2018 om 04:25 schreef Peter Lauer:
> > Bye the way the ESP32 has a wired Ethernet interface too...
> > We are using it for an in vehicle IO controller that gets the command via
> CAN but can interface to a smart phone for information.
> >
> > Peter
> >
> > -----Original Message-----
> > From: [email protected] [mailto:[email protected]] On
> Behalf Of John Dammeyer
> > Sent: Sunday, August 05, 2018 3:38 PM
> > To: 'CANLIST'
> > Subject: RE: [CANLIST] Non Powered Devices
> >
> > The idea of having a CAN driver on one of these little Wifi based units is
> really quite exciting although I must admit, I just switched my 3D printer
> Raspberry Pi over from WiFi to wired.  What happened over the space of 2
> months is the location at one end of the house where the printer was
> located suddenly had 8 extra WiFi devices visible (using inSSIDer3).  Four of
> those devices had a higher signal strength than my own network and the
> Octoprint essentially became useless.
> >
> > Running Ethernet cable not only solved the problem but the video camera
> watching the 3D printing updated much faster.  As did downloading print
> files.
> >
> > I realize we can add repeaters etc. but the reality is a wired system like
> Ethernet or CAN is still way better than WiFi.  So that then begs the question
> is really does it matter if the ESP32 has CAN?  Still the modules are so cheap
> compared to a Raspberry Pi that doesn't have CAN.
> >
> > John Dammeyer
> >
> >
> >> -----Original Message-----
> >> From: [email protected] [mailto:[email protected]] On
> >> Behalf Of Heinz-J�rgen Oertel
> >> Sent: August-05-18 1:13 PM
> >> To: CANLIST
> >> Cc: John Dammeyer; Kees Zagers
> >> Subject: Re: [CANLIST] Non Powered Devices
> >>
> >> Am Sonntag, 5. August 2018, 19:35:47 CEST schrieb Kees Zagers:
> >>> Sure it has a CAN-controller and here is a product that uses it,
> >>> complete with CAN transceiver and connector:
> >>>
> >>> https://www.olimex.com/Products/IoT/ESP32-EVB/open-source-
> >> hardware
> >> OK Kees,
> >> the evaluation board has a CAN transceiver which is connected to GPIO5
> and
> >> GPIO35. But I can not get any information about the CAN module insight
> the
> >> ESP32-WROOM32, Please help.
> >>
> >> Am Sonntag, 5. August 2018, 18:23:46 CEST schrieb John Dammeyer:
> >>> Don't know if this helps.
> >>>
> >>
> https://github.com/eerimoq/simba/tree/46778af636a4fe58773753b5ed9da8
> >> 3916cfaa
> >>> 06
> >>>
> >>> It appears to have CAN module.
> >> Checking the source code leads to two modules:
> >> src/drivers/can.c
> >> src/drivers/mcp2515.c
> >>
> >>
> >>
> >>
> >>
> >> --
> >> Archives and useful links: http://groups.yahoo.com/group/CANbus
> >> Subscribe and unsubscribe at www.vector.com/canlist/
> >> Report any problems to <[email protected]>
> > --
> > Archives and useful links: http://groups.yahoo.com/group/CANbus
> > Subscribe and unsubscribe at www.vector.com/canlist/
> > Report any problems to <[email protected]>
> > ?!???jwn???X???m?????.?????o???
> > n????&?jwn?????m??
> 
> /y?h??&???+-????Z?*k??^??hq???h?w???-???~??j?de==
> 
> 
> --
> Archives and useful links: http://groups.yahoo.com/group/CANbus
> Subscribe and unsubscribe at www.vector.com/canlist/
> Report any problems to <[email protected]>

--
Archives and useful links: http://groups.yahoo.com/group/CANbus
Subscribe and unsubscribe at www.vector.com/canlist/
Report any problems to <[email protected]>