Weird Microwave problem - trips breaker when you shut the door!

A Canadian company made a programmable radiation treatment machine. If you entered the wrong setting, then backspaced and corrected it, the patient got the maximum the machine could deliver. Killed some people before they found the bug.

N

Yes, that was a classic. I always love how some people think software can take care of everything. There are some cases where a simple switch is far more reliable.

--- sam | Sci.Electronics.Repair FAQ Mirror:

Repair | Main Table of Contents:

+Lasers | Sam's Laser FAQ: | Mirror Sites:

Note: These links are hopefully temporary until we can sort out the excessive traffic on Repairfaq.org.

Important: Anything sent to the email address in the message header above is ignored unless my full name is included in the subject line. Or, you can contact me via the Feedback Form in the FAQs.

When electronic control boards were still on only half of the microwave ovens sold, it was a common-place to suggest to customers on the phone that if they were to unplug the oven from the wall outlet for a minute or two, the (software) fault they were describing would probably correct itself. Many times it did!

that

When ovens all had mechanical timers I got many a panic call from someone who couldn't figure out why their oven had stopped working but the hotplates were fine. I got quite skilled at leading them through the reset procedure over the phone - saved a drive.

N

Then you're relying on all the electronics to be functional as well as the software to be bug free, in real life there's just way too many chances for that to not work. They've been doing it the way they do for 30 years now and I've not heard of a single safety issue coming up in relation to that.

"James Sweet" bravely wrote to "All" (01 Jul 05 06:30:35) --- on the heady topic of "Re: Weird Microwave problem - trips breaker when you shut the door!"

JS> From: "James Sweet" JS> Xref: aeinews sci.electronics.repair:52068

JS> Then you're relying on all the electronics to be functional as well as JS> the software to be bug free, in real life there's just way too many JS> chances for that to not work. They've been doing it the way they do for JS> 30 years now and I've not heard of a single safety issue coming up in JS> relation to that.

Software isn't as unreliable as you infer since there are planes which operate "fly by wire" over our heads at this very instant. Hundreds of millions of passengers per year are trusting their lives to a line of code translated to 1's and 0's. These FBW planes even use joysticks! Like it or not, the world is now a giant video game... PONG rulz!

A*s*i*m*o*v

... Digital circuits are made from analog parts.

you

OK, next time I want to spend $20,000,000 for a microwave oven I'll keep that in mind. :)

--- sam | Sci.Electronics.Repair FAQ Mirror:

Repair | Main Table of Contents:

+Lasers | Sam's Laser FAQ: | Mirror Sites:

Note: These links are hopefully temporary until we can sort out the excessive traffic on Repairfaq.org.

Important: Anything sent to the email address in the message header above is ignored unless my full name is included in the subject line. Or, you can contact me via the Feedback Form in the FAQs.

Some ground worker duct taped over the instrument holes on a plane and killed everyone on board.

N

The process of designing, writing and most importantly, testing the software used in an aircraft are completely different from that for the firmware in a microwave oven. There's just no comparison.

"Sam Goldwasser" bravely wrote to "All" (01 Jul 05 14:35:24) --- on the heady topic of "Re: Weird Microwave problem - trips breaker when you = shut the door!"

SG> From: Sam Goldwasser SG> Xref: ae > "James Sweet" bravely wrote to "All" (01 Jul 05 06:30:35) > --- on the heady topic of "Re: Weird Microwave problem - trips breaker when you > shut the door!" >

SG> OK, next time I want to spend $20,000,000 for a microwave oven I'll SG> keep that in mind. :)

Oh, be serious! The door interlock mechanism is an inherited safety device from the era when microwave ovens were controlled by what was basically a washing machine timer switch. The door could be easily monitored by an optical sensor. If the beam doesn't line up the microcontroller doesn't power the magnetron. That's about 1 line of code and no way would it cost $20,000,000 to do that!

A*s*i*m*o*v

... Puddy-tat's not so bwave in Gwanny's microwave!

"James Sweet" bravely wrote to "All" (01 Jul 05 23:59:06) --- on the heady topic of "Re: Weird Microwave problem - trips breaker when you shut the door!"

JS> From: "James Sweet" JS> Xref: aeinews sci.electronics.repair:52141

JS> The process of designing, writing and most importantly, testing the JS> software used in an aircraft are completely different from that for the JS> firmware in a microwave oven. There's just no comparison.

I feel you are exagerating somewhat. True there is no comparison between the two but the principles are the same. Both use positive goal oriented programming where error checking and input validation is often an after thought. The truth is that "shit" happens and both have no way to handle "it" when "it" does. Hell, even with human pilots at the controls they still managed to crash them without help of software.

A*s*i*m*o*v

... 'Keep the smoke inside.' -- 1st Rule of Electronics.

you shut the door!"

And what happens if the software has a bug is gets kicked into a funny state by a power glitch? Hypothetical product recall notice for the HiTeck Internet-Ready MicroNuke 1: "We have determined that a particular key combination will allow the oven to run with the door open. If you've already experienced this, We sincerely hope you haven't put your head inside to try to figure out what was going on. Owners are requested to return the ovens for a software upgrade or to cut off the power cords and throw them away. The software upgrade may also be downloaded from our Web site. We believe this will solve the problem but of course, bugs are to software like flies are to honey."

--- sam | Sci.Electronics.Repair FAQ Mirror:

Repair | Main Table of Contents:

+Lasers | Sam's Laser FAQ: | Mirror Sites:

Note: These links are hopefully temporary until we can sort out the excessive traffic on Repairfaq.org.

Important: Anything sent to the email address in the message header above is ignored unless my full name is included in the subject line. Or, you can contact me via the Feedback Form in the FAQs.

device from the era when microwave ovens were controlled by what was basically a washing machine timer switch. The door could be easily monitored by an optical sensor. If the beam doesn't line up the microcontroller doesn't power the magnetron. That's about 1 line of code and no way would it cost $20,000,000 to do that!

Would you like to rely on this line of code when your 4 year old gets up the heat her breakfast? Buggered if I would. Even elevators with 100+ years of debugging kill people every year. I don't want my kid's eyesight ruined because some slave in a Chinese factory coughed at the wrong moment.

N

The power glitch issue is the most realistic problem, as a user of microcontrollers in my own projects, I've had to track down some really obscure bugs where a program would get in a state that it should never be able to get into, a random bit gets flipped somehow in a register and suddenly the program is doing very strange things and it's difficult to pinpoint why it got there.

In the end, nothing beats the dependability of a time proven mechanical interlock. You need the microswitches anyway, may as well have them work directly with the power rather than interface to it through other components. As someone else mentioned, the triacs that control power to the magnetron can and do fail, usually shorted which would disable even the most stable software control. The present design is cheap, effective and reliable, there's just no good reason to change it.

You all have some good points. I do agree that mechanical safety devices should be used as well. For instance, microwaves should (if they don't already) physically disconnect the magnetron from the circuit if the door is open so that even if the microprocessor tries to turn it on, it can't.

The reason I brought that up was that it seemed strange to protect against microwave exposure by using something that seemed equally dangerous: a short circuit. Also, I remembered reading somewhere (probably either here or on the alt.home.repair group) that aging circuit breakers might fail to trip when they should. That may not be true, but I couldn't help wondering what would happen if for some reason the breaker didn't trip. So I assume that all microwaves also have internal fuses which should at least blow in that situation.

--
Travis Evans
[The email address on this post is valid, but may change from time to
time.  Make sure you use the latest email address; if you use an old
one, I will not receive your message.]

Yes.

--- sam | Sci.Electronics.Repair FAQ Mirror:

Repair | Main Table of Contents:

+Lasers | Sam's Laser FAQ: | Mirror Sites:

Note: These links are hopefully temporary until we can sort out the excessive traffic on Repairfaq.org.

Important: Anything sent to the email address in the message header above is ignored unless my full name is included in the subject line. Or, you can contact me via the Feedback Form in the FAQs.

I think that was covered on the very first post. Perhaps there is nothing wrong with the microwave!

By the way, anybody else wire up a Heathkit microwave oven way back then?? The year was 1971 I think. I was saving big bucks for buying that $400 kit!! I still think a simple rotating dial is the ultimate control for simplicity and speed.

greg

This was quoting a bit about, trying the microwave on a different breaker.

greg