Seeking powerful controller for OBD-II Application

Thread Starter

DrDaniel

Joined Dec 11, 2014
3
Hello all,

The company I am working at is looking to implement an embedded controller to work with the ELM327 to communicate with the OBD-II port for light duty vehicles.
The system will need to run a massive amount of routines, and constantly be cycling through data requests and running the proper functions to read in the requested data properly. From there it will need to forward the data to wifi or bluetooth transmitters to send the data to external servers or processors.

So the question at hand is, what is a good microcontroller to select that satisfies the following:
1) Large amount of programmable flash (Currently using ATMega328, 32k is not enough)
2) UART and SPI connecvtivity
3) Easy to program, preferably in C or C++
4) Fairly priced

I am not too aware of the market out there currently, aside from PIC and arduino controllers. Any suggestions for a good embedded controller you yourself would prefer would be appreciated.

Thanks,
-Daniel
 

MaxHeadRoom

Joined Jul 18, 2013
28,686
I seem to have seen something on the Picmicro site using the 18F series that posses a 'ECAN' module that I believe is aimed at the CAN-bus systems.
May be worth a look on the Picmicro site.
Max.
 

Papabravo

Joined Feb 24, 2006
21,225
I seem to have seen something on the Picmicro site using the 18F series that posses a 'ECAN' module that I believe is aimed at the CAN-bus systems.
May be worth a look on the Picmicro site.
Max.
Unfortuneately CAN and OBD-II are different animals.
I would be looking at the ARM family from Atmel or NXP, or possibly the Freescale(formerly Motorola) offerings.

Can you elaborate on your definition of fairly priced? Do you understand the bathtub curve pricing model of semiconductors?
 

Thread Starter

DrDaniel

Joined Dec 11, 2014
3
Unfortuneately CAN and OBD-II are different animals.
I would be looking at the ARM family from Atmel or NXP, or possibly the Freescale(formerly Motorola) offerings.

Can you elaborate on your definition of fairly priced? Do you understand the bathtub curve pricing model of semiconductors?
I recall examining the bathtub curve when I was doing a study on design for test. I guess for fairly priced I wouldn't be able to name an exact number, but i was hoping to find an embedded controller that had sufficient communication link capabilities (CAN, UART, SPI...) and physical memory storage, but in contrast wasn't too expensive because it was jacked up with loads of GPIO's, ADC's, PWMS, etc. Really just looking for a controller that's primary focus is communication connectivity.
 

Papabravo

Joined Feb 24, 2006
21,225
I recall examining the bathtub curve when I was doing a study on design for test. I guess for fairly priced I wouldn't be able to name an exact number, but i was hoping to find an embedded controller that had sufficient communication link capabilities (CAN, UART, SPI...) and physical memory storage, but in contrast wasn't too expensive because it was jacked up with loads of GPIO's, ADC's, PWMS, etc. Really just looking for a controller that's primary focus is communication connectivity.
I think it is a mistake to think that there is a cost for the silicon allocated to a variety of functions. If you look at product design and fabrication costs you want to go for a package which has a smaller number of pins. As we go from DIP, to SOIC, to QFP, to BGA there is a definite price impact. In my estimation you want to pick the newest parts that are trending down the initial part of the bathtub. You want your project to hit production at the first knee to maximize the life cycle of the product. You want to EOL your product when it starts climbing back up the backside of the bath tub.
 
Top