Which Micro Controller to use?

[Salgat]

Why do some people prefer the PIC to, for example the AVR's line of micro controllers? What are the advantages of the micro controller that you chose to use over others? I know that an extremely general answer is "You get what fits your needs", but I'd like a more specific answer than that.

 

[blueroomelectronics]

And some prefer AVR to PIC, I like PIC as I've been using them for years. Lots of source code out there, I also like the tools (MPLAB).

 

[donniedj]

To find out read about them, sample them, and design with them. Go to the manufacture's website and get a feel for what they are all about and the history behind them. Microchip produces not just a huge variety of controllers but all types of analog devices. And with each new addition to the controller family within a bus size the prices get lower and capability gets higher.

Its not just 1 vs. 2, for there are many microcontroller producers available. Put the free resource of the manufacture above this forum or any forum.

 

[3v0]

....I know that an extremely general answer is "You get what fits your needs", but I'd like a more specific answer than that.

The uC is less important then the tools and support for it. You can learn to program anything given enough guidance. The flip side is also true.

Currently my favorite "first chip" would have to be a member of the PIC 18F family. Free tools including C compiler. Inexpensive fast programmer/debugger support via PicKit2 or JuneBug.. Inexpensive.

I am using the 18F1330 because it has an 8MHz internal osc that can be used with the PLL for a 32MHz clock.

 

[bananasiong]

IMO PIC is more resourceful, this is commonly used in my college, compared to AVR. But to be honest, some AVR is faster compared to PIC with the same oscillation frequency.

 

[Salgat]

The one thing I've really been pondering is how the PIC uses in some cases one accumulator while AVR's 8 bit micros have 32 registers connected to the ALU. I've done some work with the 68HC11s which have only 2 accumulators and I really love having more registers than I know what to do with. Are PICs often used due to their popularity more than the actual architecture? Also, does anyone know a good PIC for me to check out. I want to look up some datasheets on a good one and see its abilities and performance, I don't have much knowledge on them.

 

[bananasiong]

I used to use 68HC11 too, I'm not sure about AVR, but if you have the background of 68HC11 like me, you can understand the PIC16F easily. For me, one accumulator is enough. Besides, for PIC18F, you don't even need to use a working register for moving the data from SFR to GPR or vice versa. Some bank switching can also be eliminated with 18F.

 

[simrantogether]

Well...

Well

AVR is the hot industry microcontroller... It is new and it is the only 8 bit microcontroller to have JTAG Support...

While PIC has enough Internet resources , AVR has good industry reputation.

It's new design is its advantage.

Go for pic if you are beginner... AVR if want to work in Industry...

Regards,

Simran..

 

[Nigel Goodwin]

Well

AVR is the hot industry microcontroller... It is new and it is the only 8 bit microcontroller to have JTAG Support...

While PIC has enough Internet resources , AVR has good industry reputation.

It's new design is its advantage.

Go for pic if you are beginner... AVR if want to work in Industry...

I hardly think someone from a country where 8051 is still comonplace is able to suggest what 'industry' uses

PIC outsells AVR by a MASSIVE margin in industry, and the AVR is only 'new' by comparison to the 8051 - although it is newer than the original PIC devices.

You would be amazed how much stuff you find that uses a PIC, although I have seen an occasional AVR in some products.

 

[simrantogether]

Ya..

I hardly think someone from a country where 8051 is still comonplace is able to suggest what 'industry' uses

PIC outsells AVR by a MASSIVE margin in industry, and the AVR is only 'new' by comparison to the 8051 - although it is newer than the original PIC devices.

You would be amazed how much stuff you find that uses a PIC, although I have seen an occasional AVR in some products.

The reason why it is more sold in industry is they are cheap and ..

The industry they are sold mainly is education industry...

Even i am also having lots of new pic controllers with me... which are lying waste...

Does pic user know what is JTAG?

People just stick to PIC have they ever listened MSP430... or simply ARM...


EDIT:

There is a suggestion:

Our forum's main members should update themselves to learn 32 bit microcontrollers as well...

Regards,

Simran..

 

[blueroomelectronics]

TI is a great chip maker but can and may abandon anything at any time remember the TI99/18, awful chip to work with. There is virtually no MSP430 support from TI unlike Microchip which has an active forum. Even the MSP book is considered awful. Not a good chip to start with.
https://www.amazon.com/Embedded-Systems-Design-MSP430-Technology/dp/075067623X
As for the 8051, nice chip 20yrs ago. I still have an 8052-BASIC somewhere.

As for 32bit chips they are useful if you need that kind of processing power, not nessary for someone just starting out. Any 8bit micro can flash an LED (even the 4bit Intel 4004 could do that)

 

[3v0]

Does pic user know what is JTAG?
Simran..

From BNET.com

Microchip Technology Introduces $99 In-Circuit Debugger
Business Wire, Nov 6, 2000

Business Editors/High-Tech Writers

CHANDLER, Ariz.--(BUSINESS WIRE)--Nov. 6, 2000

Microchip Technology Inc. (Nasdaq:MCHP) Monday introduced a low-cost version of its MPLAB(R) In-Circuit Debugger (ICD), a powerful development tool for the Flash PIC16F87X microcontroller (MCU) family.

At $99, the MPLAB-ICD brings high performance debugging capability for any embedded designer.

With the Pickit2 that $99 has come down to $33 and the interface update speed has beed dramaticaly improved with the move to USB.

In regrads to run control these units work about as well as a multi thousand dollar processor emulator.

IMO PIC is more resourceful, this is commonly used in my college, compared to AVR. But to be honest, some AVR is faster compared to PIC with the same oscillation frequency.

One clock cycle per instruction vrs multiple clock cycles per execution. It just means that one requires a faster clock. Also means that using the clock as an indication of thru-put is folly. Benchmarks....

 

[3v0]

Allow Me To Mount My Soapbox

You wil be more valuable as an embedded programmer when you have worked with chips from several vendors. In large production runs the chip is often picked for you. For products with small runs code is often leveraged from a previous product, again there is no processor choice on the part of the programmer/engineer.

After a few years/decades of working with uC's they all sort of look the same. With a few execptions, given the choice I choose to use the chip/family that I used last or most frequently. In general switching chips and toolsets has more downside then any advantage of picking X over Y.

**Silicon is much less expensive then a programmer/engineers time. That is why industry in general uses higher level languages.

You have to understand the processor and all the hardware connected to it. You have to be good at reading and understanding data sheets.

My advice is to stop the debate about processor brands. Etch some boards write some code. ASM till you understand the processor well then move up to a higher level language.

EDIT: When you sit next to the guy interviewing you for that first job he will care less about what uC's you use and more about your ability to think (solve problems) on your feet. That ability can be improved with by doing original designs rather then copying. Write a I2C driver from the spec. When you are up to it do somthing that has not been done. But do not expect the interviewer to be impressed by any of what you have done. He will want to know if you can think..

---------

** In large production runs the cost of the chip is more of a concern. However a higher level language should reduce the time to market enough to choose it over ASM. Maybe not for a toaster, but it is with more complex items like IPODS, cell phones etc.

 

[bananasiong]

Hopefully this is not a dumb question..
I'm a student, interested in microcontroller and electronics, at the same time, I'm working partime in the industry (design department). But my job is not relevant to any microcontroller or any discrete components, they are talking about nano meter. I found that what I've learned is not applicable in my job.
I can see many of my senior here also the same, work and learned are totally different story. Maybe this is because most of the vacancy of embedded engineer here require working experience for a few years.
My lecturer told me that if I want to do the relevant job, then do teaching.. lol I cannot teach

 

[futz]

After a few years/decades of working with uC's they all sort of look the same.

Boy is that ever true!

 

[3v0]

I found that what I've learned is not applicable in my job....
I can see many of my senior here also the same, work and learned are totally different story.

What you learn is not so important as your demonstartion that you can learn.
If the company you are with is willing to take you on after graduation and there is a path to the type of work you want to do. Then I would go with that.

If not you can come closer to what you need to know in the right masters program.

Grades are important too.

 

[bananasiong]

What I'm doing now is not my ideal job, really. I have a few friends that have done the internship in different department in the same company of mine, but none of them worked about embedded systems.
Due to the financial problem, I think most probably I will stay there after my graduation.
As for the grades, hopefully I can get first class..

 

[3v0]

What I'm doing now is not my ideal job, really. I have a few friends that have done the internship in different department in the same company of mine, but none of them worked about embedded systems.
Due to the financial problem, I think most probably I will stay there after my graduation.
As for the grades, hopefully I can get first class..

Some companies have programs where they will send you back to school for an advanced degree. You would have to show enough talent/promise to make them think it is worth doing.

 

[bananasiong]

Not sure whether they need that high qualification in that department, but a degree owner with good result will get a job there with satisfied salary for sure.
I'm thinking of doing partime advance degree.