I smoked 6 mc34063 today, not bad huh!

 

Poor things. They were so young too.

 

Jay Leno (a US comedian and talk show host of a popular night talk show) had a joke that pertains. He was doing Freto's commercials at the time, and had just totaled his motorcycle. The manufacturer of the bike told him to "Crunch all you want, we'll make more".

Smoking chips is a part of learning. If you have never destroyed a component accidentally, you are probably doing it wrong. We learn more from our mistakes than our successes.

 

MC34063 are like $0.60 apiece. It's not like blowing a more expensive component.
I have smoked at least three STM32F407 at $10 each so far. And it is a pain to unsolder a 100-pin LQFP chip every time.

 

I have smoked at least three STM32F407 at $10 each so far. And it is a pain to unsolder a 100-pin LQFP chip every time.

How did that happen?

 

I really don't know. In all cases the ground on the scope probe touched the 3V supply line. I know this should not destroy the chip. What I do know is that the chip got very hot after that. CMOS latch-up maybe?

 

Maybe, or could it be that the supply overshot after you removed the short, and this somehow broke the insulation in the chip? Do mosfets fail like that?

 

Smoking chips is a part of learning. If you have never destroyed a component accidentally, you are probably doing it wrong. We learn more from our mistakes than our successes.

Thanks for the encouragement Bill

 

MC34063 are like $0.60 apiece. It's not like blowing a more expensive component.
I have smoked at least three STM32F407 at $10 each so far. And it is a pain to unsolder a 100-pin LQFP chip every time.

Yes lucky they are cheap. and hope you don't smoke STM32F407 as many as I did.

 

The important thing is to learn from it.
The last thing I let the smoke out of was a surface mount LED. I was checking the soldering of that part of the board with 5 volts to see if the LEDs lit up, but I hadn't checked that there were no solder bridges under the resistors. Next time I'll measure the resistance of the resistors and check that it reads what it should.

 

The important thing is to learn from it.

Yes, totally agree

 

My first job I took out a $10K AC power synthesizer that could mimic any AC power source in the world for several hundred watts. My supply started to smoke and when I went to hit the OFF button I HIT the OFF button, and the ON button too.

Worst disaster was a scope probe slip. I was in on a Saturday pushing along some units when my probe shorted ground on a computer controlled test stand and thru the IEEE-488 bus took out the computer, a power supply, electronic load, and a spectrum analyzer. Better then $50K in equipment.

All of it got repaired, and my boss, while not pleased, just accepted what happened.

 

My first job I took out a $10K AC power synthesizer that could mimic any AC power source in the world for several hundred watts. ... ...

I feel better now, thanks for sharing.

 

I have been known to brand myself with part numbers, especially off transistors. Who knew they were hot? Must have enjoyed it though, I still do it occasionally.

 

Thanks for the encouragement Bill

i would agree to this also...


well, sometimes it is more of the failures than the successes..

more failures = more learning..

 

My well respected boss once told me after a $40K boo-boo on my part (I set a timing function backwards) that "the only people that don't make mistakes are those people that don't do anything! Did you learn anything from it?"

So, chalk it up to a learning experience and try not to make the same mistake again.

 

Did you learn anything from it?"

So, chalk it up to a learning experience and try not to make the same mistake again.

Yes I did learn from it: disconnect the power supply before replacing components on a breadboard.

I thought I can get away from it, but apparently I can't.

 

I learned that all timing relays don't use the same terminology. Some refer to delay time before action and others refer to active time.

 

Usually the rule is disconnect the power when you´re doing anything, unless you need the power on to make a measurement. Things like loose screws when working on a case can ruin your day.

 

I've used 34063 SMPS IC's for many differnt projects, and never "smoked" one. They are quite forgiving, especially if you use it's recommended current limiting resistor (which is in all the datasheet example circuits).

You definitely need to improve your test setup layout and test procedures.

One good improvement is to use a current-limited PSU to power the SMPS under test, and if you have mis-wired it or have mistakenly applied a short circuit etc then the Vin PSU will just go into current limiting and not fry your circuit under test.

A good bench supply with adjustable voltage and current limiting is a must, as is a reliable variable load dummy load with some type of load current adjustment.