How do microcontrollers fail?

[alec_t]

I can't help feeling that with all that complexity the system is bound to fail . The fewer components/functions the better. Apply the KISS Principle. For starters, emergency lighting shouldn't, IMO, be dimmable.

 

[blueroomelectronics]

Who's paying for all this? A four layer board and SAM7 CPU IMHO is overkill on every level yet you seem to want to cheap out on stuff like voltage clamping and the like. I would also think NiCad is an odd choice as they have a very low power density compared to NiMH or Lithium, my personal NiCad inventory is zero, they've all died. Every emergency light I've ever seen uses bog standard SLA batteries.

I've looked at your past posts and have to agree with Mr Gibbs. If this is a hobby that's fine but try to learn from the members here and on the other forums you post at, as a commercial product stop now. If it is a commercial product what are you bringing to market that's unique or makes your product viable?

What we know.
It's a LED based emergency lighting system with a CPU that's more suited for a PDA than a UPS.
The LEDs are 100V chains (why not use a lower voltage closer to the battery voltage?)
Against the advice of the other members you're sticking with NiCad. If memory serves they had memory problems
You're building or have built a 4 layer PCB around the SAM7, how, why if you're still obviously not beyond the spec stage and are still designing it.

I've seen no power supply, charging circuit info. I'm really curious about what you've got in mind.

Oh please post a photo of the PCB, I'm guessing you're not willing to post the schematic.

 

[ericgibbs]

Totally agree with you....thing is, the expert design consultancies who claim to know what theyre doing, charge an absolute fortune. I don't.

Flyback,

Who represents your Company at these Consultative meetings.?

IMHO, its they who are fault.

 

[blueroomelectronics]

Totally agree with you....thing is, the expert design consultancies who claim to know what theyre doing, charge an absolute fortune. I don't.

Speaking of expert design consultancies.....its amazing how many of them take shortcuts without telling the client..........for example, one consultancy charged $40000 for a LED light which was in a tiny package and switching at 1MHz......the only reason they got it so small, was because they used 1206 sense resistors (in each led string) dissipating 225mW each...on pads that were abs min. The 225mW was with perfect equal sharing in the paralleled leds.........with unequal sharing, these 1206's would obviously dissipate more than 225mW.


The biggest cracker i saw was a company who designed a smps for a telecoms unit............were talking of a telco company youll know of..............i took one look at the feedback and coudl see something was amiss.....i told the gaffer, and he changed the opto to a same-part-number-different-manuf, and it oscillated so bad it put sidebands on the mob phone signal and was unusable......all 1000 proto units in the field had to be recalled.

Another cracker was a offline device with a reg in it consiting 2 off 5V zeners in parallel.....yes they didnt share.....yes they blew in the field....yes the custoemrs were angry.

The best had to be this software guy who programmed undertemperature shutdown into a device without telling anyone.......the temp sensor was very inaccurate and cheap......that nearly closed down the company concerned.

I went to one place and found 5 overvoltaged resistors in a small product which had just gone into production and was with customers (were talking alomost 240V on 0805 resistors. (duty one half).......i said i was leaving unless they ordered it be fixed...when i went into the car park after work.....my rear tyres had been let down, and the valves glued up.

I can list many more of the nice shortcuts that the expert consultancies take.

How did they get approved? You just can't (at least in most countries) just hook stuff up to the mains or telco without regulatory approval. On the other hand some countries it's a free for all as it's rarely enforced.

Most companies have the design looked over by someone before going into production. Yes there are many bad, under-qualified designers out there...

Please have someone check your full design before you go into production, even prototyping. Without seeing anything beyond a couple of basic schematics you've posted (did you actually built them to test them and I don't mean simulate?).

PS why would 240V be anywhere near a 0805 resistor?

edit: as noted dim-able is an odd thing for emergency lights. Perhaps they're emergency mood lights

 

[mvs sarma]

Hi Bill ,

i had to enjoy one 18F4550 failing at its RD0, RD1 and / or RD2 this week. It was a new piece made in Thailand. i had used the port pins for driving DDS chip AD9850.

I was checking other aspects like DDS module (e-bay item)) failure etc for three days. Software was tested on another pcb.

Finally just replaced by a new device and the whole design started working. I am yet to analyze which of these pins failed.

 

[Flyback]

PS why would 240V be anywhere near a 0805 resistor?

....when i moaned about that my tyres got let down.

dim-able is an odd thing for emergency lights

....the ELU must be universal and work with 3V to 8V batteries, depending what the customer wants...some cust want 5W for 3 hrs...some want 1W for 1 hr, etc etc...so we dim the leds as and when etc etc.

Who represents your Company at these Consultative meetings.?

..i dont know, the big boss's i spoze....big boss's are never engineers....as Winston Churchill said "Engineers should be kept on TAP , not on TOP."


Who is an expert....who knows....i once worked at a big SMPS company.... it wasnt a pure smps company, but had smps in all its products...you will have heard of the company...it is one of the big names that you will be familiar with.
One day, the Chief Engineer gave me two PWM IC's from the same family (30-40W offline flyback). He asked me to make prototypes with them and compare them....one had a higher rds(ON) than the other one.....The Vin was 400V because they were to feed off a post PFC rail.

I did the job, and the one with higher rds(on) had better efficiency than the one with lower rds(on)....also, (they had internal fets), you could clearly see with the "heat scope" thingy that the higher rds(on) one was running cooler than the lower rds(on) one.
I presented these results at a meeting to all the engineers, and the chief engineer started grumbling at me in front of everyone that i must have done all the tests wrong because he said "how could the one with higher rds(on) have been more efficient than the one with lower rds(on)."

Readers here will know that when switching off a 400V bus the switching losses are dominant..and fets with higher rds(on) tend to have lower Qg......but this Chief Engineer, who designed most of the companies products, was there telling me nonsense.
It just goes to show....who really does know what they are talking about...and who says so?
.................................................................
returning to the discussion..its a sam3 not sam7

 

[misterT]

If you don't mind. I am very curious to know who are you working for now? Do you have a business of your own? What kind of people do you have in you team (working with this product)? Is the emergency light something that was ordered from you, or something that you are planning to market once it is done?

 

[Flyback]

I work in a company with other senior engineers.........the chief engineer here is one of the best smps engineers in the world....he checks over my work.

 

[misterT]

I work in a company with other senior engineers.........the chief engineer here is one of the best smps engineers in the world....he checks over my work.

I feel like you are in a tough place there now. I don't know how far in you are with the design, but if you feel like you have to take few steps back to make a better design, you should be encourage to do so. I've seen couple of software projects where months of work have been scrapped because of much better solution was found.. thanks to the previous work. And the end result was elegant and easy to maintain and easy to develop further. If they did not have the balls to go back, they would have had a complete mess as a final product. And that is a pain in the ass to support and continue.

Edit off topic: "they would have had a complete mess as a final product".. is that proper english.. I know I make lots of mistakes, but that one puzzles me.

 

[Flyback]

yes thats perfect english....................
the place i am at now isnt like the others where i had tyres let down

 

[misterT]

yes thats perfect english....................
the place i am at now isnt like the others where i had tyres let down

How far is the design? Software/hardware? The original topic was about reliability of microcontrollers, but I think that is the last thing to fail in the system.. You really need to re-evaluate the design before it goes into production.

 

[blueroomelectronics]

Hi Bill ,

i had to enjoy one 18F4550 failing at its RD0, RD1 and / or RD2 this week. It was a new piece made in Thailand. i had used the port pins for driving DDS chip AD9850.

I was checking other aspects like DDS module (e-bay item)) failure etc for three days. Software was tested on another pcb.

Finally just replaced by a new device and the whole design started working. I am yet to analyze which of these pins failed.

Wow, I've yet to have a PIC just fail on me. That said I'm also surprised at the Errata sheets that Microchip can still bugger some part of the IC that they've been doing for decades.

 

[nsaspook]

Wow, I've yet to have a PIC just fail on me. That said I'm also surprised at the Errata sheets that Microchip can still bugger some part of the IC that they've been doing for decades.

Unless there is process change that mandates a new litho mask and layout (stepping), fixing non-critical Errata almost never happens anywhere unless a really big buyer demands it.