Has anyone used a CAN embedded controller?

[JW]

This is the embedded CAN BUSS controler brand im looking at. - http://www.microchip.com/stellent/idcplg?IdcService=SS_GET_PAGE&nodeId=1480&redirects=can


JW

[Yyrkoon]

No, I have not. Have some experience with embedded Ethernet though. Out of curiosity, what did you have in mind ?

[JW]

Ive built an external combustion burner that burns corn or wood pellets. Here is a video-
Jeremys Cornburner.mp4
http://www.youtube.com/watch?v=IKDSX3lMasw&list=UUfC39iNDbfNShcbK9gFvenw

This one you see in the video is totaly manual. I have applied for a patent on the automated version, will post some picture's here when I get sometime.

The unit is a hybrid convection burner that uses forced and free convection. It has an afterburner, preheated air injection, a fiberoptic control that controls fuel admission using an auger. Also there is a lambda sensor to control the cyntrifugal blower which attenuates combustion. Its kind of a tricky thing to automate. I have alot of automotive experience thats why I choose the CAN BUSS, it will be easy to make a graphic interface.

JW

[Yyrkoon]

JW, that sounds like a really cool, and fun project. Not to mention useful.

Are you looking for specific information concerning these different CAN capable MCU's ? Or are you wondering if you can use something else that is different  or better ?

I was just looking through their evaluation / development boards, and I have to say they are pricey . . .
http://www.microchipdirect.com/ProductDetails.aspx?Catalog=BuyMicrochip&Category=CAN%20or%20LIN&mid=13&treeid=6

[Tight Yorky]

Good Evening JW and Yyrkoon,

Apologies, I only visit this site periodically and hope my contributions are not too late to be of assistance.
I have read your post and viewed your 'Youtube' video. Very impressive.
Unfortunately, as per Yyrkoon's posting I haven't actually used the CAN bus features of the microchip MCUs. I did evaluate the use of CAN bus for some products but unfortunately it was concluded to be a 'square peg in a round hole'.
The industry my company 'sells to' is a proven technology, "pile it high-sell it cheap" market. However, this said, the CAN communications was not the issue but the fact that customers were using MODBUS, PROFIBUS and manufaturer de-facto standards. All our products are now PIC based, having originally started 6502 based (actually 6503) !
The advice I would give related to the Microchip CAN bus MCUs is to use the PIC24 or PIC33 units, as a minimum. The complexity of using CAN, in my opinion, requires the additional facilities (all free!) provided over the PIC18 by the 16 bit MCUs. A brave statement but... do not let the "advanced" marketing literature put you off. The 16 bit units are actually easier to develop than the 8 bit.
I do hope this guidance is of use and I look forward to your posts on your progression.
Best regards.
Tight Yorky.

[JW]

Yes, I was thinking about the 8 bit or 16 bit controller.

I know the buss is terminated at each end with a resistor. (ref high and low)

What im wanting in addition, is to use the control logic with a nice user graphic color LCD user interface(or something simular, neon, led, etc.

I had selectioned a micro contoller called "micro commander" again for the user interface, pwm contol etc. But this is gone now.

 the lambda sensor is also a big plus, I would like to use this to control the fan speed in a rich/high speed- lean/low speed for the blower fan speed. I know that O2 sensors have cross counts (10) per every 4 seconds or so, so the controller has to be able to handle that information.

JW

[Yyrkoon]

JW, not sure what you mean by cross counts. But I have read that O2 sensors vary in voltage, and resistance.  Depending on exhaust gas to O2 ratio, and intake O2.

What I was thinking though, was that you do not need CAN. An ADC can handle the O2 sensor, and a PWM can handle the auger(Not 100% positive). After that, you would just need SPI for an LCD screen. Then if you needed external communication you could use ethernet, USB, CAN, zigbee, wireless ethernet. etc.

Anyways, just my thoughts. I do not know exactly what you're doing.

http://www.ti.com/tool/eks-lm3s2965 

There is one such eval kit from TI using an ARM cortex M3. Has an LCD, in circuit debugger, and loads of software possibilities.  Including SYS/BIOS free from TI. Which is a preemptive RTOS. Atmel also has similar processors, and their C/C++ IDE is free, with lots of code examples, and libraries. However, I am not very familiar with Atmel so I can not really advise what to use from them.

http://www.ti.com/tool/tmdx28069usb

This one here is what I will be using soon, but I have been using another eval kit for the LM3S6965 Cortex M3 processor. The piccolo  ControllSTICK is what I would think  would be a perfect match for you. However, there is a lot to add externally, and I do not know how good you are with electronics etc. Whereas the lm3s2965 would probably work right out of the box for you.

[JW]

Yyrkoon. thanks for your input.

The idea here is a "work in progress"

There are two diameters with the burner, currently im thinking of automating that size burner dia in the video.

I just got enuff metal shells to build two more burners of the smaller size[9'25in dia]. I recycle freon contaners to build the burner units.

Im also working on a design that uses propane cylinders.

Any additional questions or suggestions are welcome...

JW 

[Yyrkoon]

Err, I should have said that the lm3s2965 would require less external components to work out of the box.

For the ADC, you'd have  to buffer the input. As the input voltage range is 0V-3V I believe. For the PWM, you would need a MOSFET driver, and MOSFET to drive the auger. That is, if I understand what it is. Sounds like it may be a motor and screw-gear type thing with an optical position sensor.

But anyhow here is something I just found that looks fairly cool.  http://www.mikroe.com/eng/products/view/586/mikromedia-for-dspic33/

I was reading on another forums where several people suggested using the dsPIC33 to another user for automotive purposes. When I ran across that kit. For the development process it may help speed you along. Supposedly, this MCU has very few to no hardware bugs, and can be / is reliable. Where some of the other PIC MCUs can be, or are buggy in hardware. But after skimming through the datasheet, it does seem to offer PWM ADC, and CAN ( up to two buses ). Plus, a lot more.

[JW]

the lambda sensor is also a big plus, I would like to use this to control the fan speed in a rich/high speed- lean/low speed for the blower fan speed. I know that O2 sensors have cross counts (10) per every 4 seconds or so, so the controller has to be able to handle that information.

For the ADC, you'd have to buffer the input. As the input voltage range is 0V-3V I believe.

Yes, some place between .3v and .8v for the heated O2 sensors I plan to use. With OBD2 the cross counts are measured and this is computed in the "short term/long term fuel trim"

For the PWM, you would need a MOSFET driver,

Yes for the blower fan speed setting

and MOSFET to drive the auger. That is, if I understand what it is. Sounds like it may be a motor and screw-gear type thing with an optical position sensor.

No, the fuel auger uses a timer, that is triggered by an (IR) optical sensor that is observing the underside of the solid fuel grate, if the fuel grate has lots of hot coals, its time to add more solid fuel. So once the grate is full with red hot coals, fresh solid fuel (pellets) are added to the top. If the auger runs for 10 seconds it moves X amount of fuel into the combustion chamber. if the auger runs 6X the time that much more solid fuel is added, when this happens a stirring device is added to the mix, which "agitates the fuel pile, and the red hot coals go out, this agitator is time'd after the auger. Then the optical sensor waits to see red hot coals again and the cycle is repeated.

JW

[Tight Yorky]

JW/Yyrkoon,
I have attached some info relating to the PIC33 that may be of assistance. From the details posted it appears a single PIC could satisfy your control. I took a PIC33 out of stock yesterday and just spent an hour setting it up on a breadboard and 'cutting' down an existing program to run on it.
Attached is.. a photo of the breadboard. The PIC project file and PIC 'c' file zipped. Sometimes the biggest hurdle is to get the MCU up and running. This simple application configures everything to get the CPU 'spinning' and produce the "Hello world" flashing LED. The original application used the serial port. I have retained the configuration of the COM pin in the code so you can see how the PIC33 provides its peripherals. It uses the internal clock so does not need a crystal.
If you decide to use a PIC, buy a PICkit 3 (£20). This unit can debug as well as program, and supply power to the chip for testing. All the data sheets are on Microchip's website, free.
I hope the above provides assistance.
Also I have attached a sheet that assists in setting out your design. Note that this sheet is not for the PIC33 but for a PIC that we do not use anymore, but can be modified to be of use.
Regards.
TY

[Yyrkoon]

Tight Yorky, did they finally fix /change the firmware in the pickit v3 ? I've readon the EEVBlog forums that when they came out with the pickit v3 that they changed ( for the worse ) how the pickit operated.

Last I heard about it was that Dave from EEVBlog did a video on the changes, and that the CEO or someone high up in the chain of command at microchip sent him a response promising to try and do something about it. However last time I was on microchip's webpages, I saw they were selling the pickit v2 again . . .

[Tight Yorky]

Evening Yyrkoon and JW,
I have just seen your reply while waiting for a film to start on the TV.
Thanks for the pointer to the EEVBlog website (did a search on google). I wasn't aware of this site. I searched for the PICkit3 issues and found a few I haven't experienced.
I personally have not experienced any major problems. But there are known quirks. One of these is when swapping PIC device types it is necessary to configure the PICkit3 as a programmer first before using it as a debugger. There are others!
One thing I am aware of (but not fully knowledgable about) is that the PICkit 3 updates it firmware from the Microchip programming software. That is, it is programmed itself first for the device type before interfacing to the PIC to be programmed.  I assume this allows it to be capable of being updated via Website downloads to allow support of all the devices.
I have listed the versions from the "About" box on the menu of the programming software...
PICKIT3.dll  V1.0.1.68
IDE   V8.84.00.00
SIM30 V4.30.00.00
MPLABC30 V3.30c
None of the company software is allowed to automatically update. That includes Windows! When products are dispatched into the big wide world, you have to be confident that they work! The above versions may be out of date with Microchip but they work
As a guide for JW, I suggested this PIC33 as it is a standard through hole type (not surface mount) and has useful features for his burner. The chip has analogue output as well as analogue inputs, 5v compatible pins and open drain outputs (easy to drive relays). If you do decide to have a go, post a note on the forum, and I'm sure there will be many willing to assist.
Best regards.
TY.

[JW]

Ok, so to get this project started I have to collect some used freon tanks.


This is a bottom up picture just before the afterburner assy. The grid is stainless steel with high nickle content. As a matter of maintenance this grid must be replaced at intervals. Some have speculated this component may have a caytlizing effect on the combustion.



Materials unuff to make at least two cornburners-

Isometric veiw of the first design



Top down look into the afterburner assy



The combustion chamber door





The reason a R12 cylinder is to give some clearance to the combustion chamber as you can see here.

JW

[Yyrkoon]

Sorry guys, I have been busy doing tons of research for another project, and reviewing code for the same project. Wish I could stay focused on  single aspect . . . lol but not the way it is working out so far.

However,  while looking around at various uC's I ran across the TI MSP430 line. Apparently they've been around for years, but I had not noticed them yet. The MSP430 looks interesting, and has a few very inexpensive eval kits. Might be something to look at. The only thing I personally do not like about them is the code size lock on code composer studio for the free options for the IDE. There is also the IAR IDE that comes with it, but I have always felt if you lock yourself into a brand use the brands IDE / compiler to do the job. That, or use the gcc port for the device, and go totally free. Which personally I am not into ( Too many potential headaches / complications ).  Anyway . . .

Tight Yorky, thanks for the info, and yeah, stuff like that bothers me about considering PIC for a project. More than likely though my bias comes from being familiar with TI uC's from the start. Well having dealt with TI's parts first at any rate. Lots of people like them, and use them so I suppose if I were more familiar with them I'd be fine with PIC too. At this stage in my own personal project, I CBF'd . . . one thing at a time sort of situation. With too many "things" on the fire already.

JW, looks like you've been busy

[JW]

Getting ready to proces these cylinders at some point soon.

Here is another link on this topic-
http://www.fieldlines.com/index.php/topic,143752.msg968071.html#msg968071



Ive got to rebuild the transmission on my 2004 F150 4wd first, I may post some pictures as im rebuilding the trans, we'll see. But I will definitly posts some pics as I start to process the cylinders, im also going to be working with propane tanks.

JW

[Mary B]

For quick and dirty proof of concept why not go with an Arduino http://www.arduino.cc/ they have a bunch of models, some will do output to displays.

[Yyrkoon]

So JW, hows the project coming along ?

Been burried up to my eyeballs in documentation again, but this time I am actually getting somewhere. Been writing code for a for a solder reflow oven controller ( toaster oven conversion ). We finally settled on using the MSP430 launchpad as a platform for this project, then I wil get time to play with other projects