which microcontroller

[ahmed quraishi]

anyone please suggest me a highly reliable microcontroller for my design
my design is a water level controller which must be highly reliable. I am going to manufacture this in large scale. after the design completion and prototype testing.
microcontroller requirements:
< must be highly reliable (not a hobby controller like pics and avr. this is a product not a college project.)
< six analog inputs
< availlability of chip (in india)

 

[Wade_Hassler]

Excluding "hobby controllers" is a an unnecessary and perhaps costly restriction.
If you want to manufacture on a large scale, it would be best to keep an open mind,
Besides, the controller is only part of what you are doing

How about more details?
You might be deluges with suggestions from people who think they know better than you do, but, at this point, specification-changes are still free.

 

[MaxHeadRoom78]

< must be highly reliable (not a hobby controller like pics and avr. this is a product not a college project.)

How come you rate Microchip as 'Hobby Controller'?
Max.

 

[Nigel Goodwin]

< must be highly reliable (not a hobby controller like pics and avr. this is a product not a college project.)

There's no such thing as a 'hobby controller', and certainly the PIC is EXTREMELY reliable and used in endless professional and commercial projects, it's amazing what you find them in.

The PIC sounds ideal for your device, as it's reliable and requires very few support components (which further increases reliability of the end product).

 

[kubeek]

The reliability of your design will depend very little on the microcontroller you choose, and very heavily on how you design the rest of the circuit.

 

[DerStrom8]

PICs and AVRs are widely used in millions of professional products all over the world. They are most certainly not limited to "hobby" use.

 

[ronsimpson]

Because you don't like PIC:
PIC10F220, 10F222, 10F320 OR 10F322
PIC12F510
PIC16F506
I think the first one we payed $0.25 (US) for it in large volume.

 

[Misterbenn]

As everyone else has said already PICs and AVRs while they are popular in the hobby community they are used in bulk on professional industrial project world wide.

Another important point is that PICs and AVRs come with the right type and number of inputs for you, i.e six analogue inputs. The more specialized processors come with more like 40 multiplexed analogue inputs which is overkill for your application. For reference we're using MPC565CZP56 microprocessor in a current aerospace application.
In reality the best way to ensure reasonable system reliability is to make sure your component stress and tolerance is optimized. by:

• ensure the input voltage to ICs doesnt exceed 80% of the maximum
• ensure the junction temperature of components doesn't exceed 80% of maximum
• ensure the working voltage across capacitors is <50% of the rating and pulsed voltage is <90% of rating
• ensure the power dissipatio of resistors doesn't exceed 80% of rating
• calculate all the above for nominal and worst case conditions (maximum supply voltage including tolerance, maximum ambient temperature, worst case combinations of resistors capacitors [think resistor dividers!]

If this was an aerospace DO178C DAL A system, where failure is catastrophic, you would implement two measurement / control channels with redundant sensors and on board error checking. So in the event of a fault (processor failure, sensor failure) you can switch to the other channel. This means you can use components with a reliability lower than your system reliability target but then take credit for compound reliability of two independent systems. But on this approach you would need two power supplies, two processors, two communication/output interfaces, two measurement systems / sensors.

An alternative to the above is to use an FPGA, where you can internally create multiple instances of the same function so provide computational redundancy. I prefer this method because FPGAs are classed as a hardware implementation not software so in DAL A systems you can avoid the costly business of code verification.

 

[DirtyLude]

To shorten what is said above^. If you are manufacturing this in "large scale", hire an EE who knows what they are doing to design this for you . The design, much more than the microcontroller will determine the reliability.

'Reliable" is a relative term. There are design restrictions and part certification for critical systems, like aerospace mentioned above and I know automotive systems and controllers. "Reliable" depends on your application and you can over engineer a part until the money runs out if you want to to get that failure rate as close to 0 as possible.

In short, based on how basic your question is and how critical you believe this design needs to be, someone else needs to do it.

 

[Joe G]

I've seen pic16F628's in smoke detectors, VERY mass produced, and a life saving device.

 

[zahwi]

If you are making it for NASA then relays are more reliable than PIC.

 

[Nigel Goodwin]

If you are making it for NASA then relays are more reliable than PIC.

Sorry, utter rubbish - relays are particularly unreliable - hence the growth of 'sold state relays'.

However, there are VERY strict requirements for micro-controllers/processors used in space, and for the software running on them.

 

[zahwi]

Sorry, utter rubbish - relays are particularly unreliable - hence the growth of 'sold state relays'.

However, there are VERY strict requirements for micro-controllers/processors used in space, and for the software running on them.

Nigel
Good of you to reassure the OP that if he is making his product for NASA then ETO is the right place for him to get advice.

I guess the name was changed to 'sold state relays' because they are always sold out.

 

[DerStrom8]

Nigel
Good of you to reassure the OP that if he is making his product for NASA then ETO is the right place for him to get advice.

zahwi, there is no mention of this project being for NASA, and Nigel never said, or even hinted, that ETO is the right place to get information about a project for NASA. Your tone is coming across as a bit disrespectful, I suggest you be a little more polite.

Matt

 

[OBW0549]

anyone please suggest me a highly reliable microcontroller for my design

I'm not aware of any particular microcontroller that stands out as significantly more reliable than the rest; you'd probably do best by looking at the manufacturers' published reliability data. And as others have already pointed out, the overall reliability of your design will depend on many factors in addition to the reliability of the μC chip itself.

microcontroller requirements:
< must be highly reliable (not a hobby controller like pics and avr. this is a product not a college project.)

As others have already indicated, there is no such thing as a "hobby controller." PICs and AVRs are high-quality, capable, and reliable microcontrollers that are widely used in industry, and the fact that they are popular with hobbyists has no bearing on that.

If you are making it for NASA...

The fact that the target application is a water level controller, and that it is to be manufactured in high volume, are subtle hints that it probably is NOT being made for NASA.

...then relays are more reliable than PIC.

That's a rather remarkable assertion, to put it mildly. Do you have any published reliability data to back it up?

And finally, how do you propose to use relays to replace the functionality of a microcontroller chip? I didn't know relays were that versatile.

 

[DerStrom8]

And finally, how do you propose to use relays to replace the functionality of a microcontroller chip? I didn't know relays were that versatile.

Considering micros are basically just a bunch of switches, and relays are switches, I'm sure it's possible. However it would take millions of relays to take the place of a micro, and that would cost millions of rimes more. Not to mention troubleshooting if something goes wrong. Better to replace one micro than millions of relays.

IMO zahwi's statement about relays being more reliable than PICs is unfounded and downright ridiculous.

 

[zahwi]

Your tone is coming across as a bit disrespectful, I suggest you be a little more polite.

Matt

"Sorry, utter rubbish". Is this polite enough?!
You the Mods have plenty of respect for your own selves, you don't need my respect and I don't respect you anyway.
If you can't take or understand a simple joke like the one about NASA and the relays then I'm with the wrong company.
Goodbye ETO.

 

[DerStrom8]

"Sorry, utter rubbish". Is this polite enough?!
You the Mods have plenty of respect for your own selves, you don't need my respect and I don't respect you anyway.
If you can't take or understand a simple joke like the one about NASA and the relays then I'm with the wrong company.
Goodbye ETO.

Suit yourself.

For the record, I would have asked you to be more respectful if you spoke that way to any member. It has nothing to do with Nigel being a moderator.

 

[large_ghostman]

"Sorry, utter rubbish". Is this polite enough?!
You the Mods have plenty of respect for your own selves, you don't need my respect and I don't respect you anyway.
If you can't take or understand a simple joke like the one about NASA and the relays then I'm with the wrong company.
Goodbye ETO.

Errmmm pssstt you forgot your ball!