Toaster Oven reflow

[Zero999]

I am concerned about the fire hazard, I wouldn't leave it unattended because there is no back-up if the TRIAC fails short. I thought about using a relay.

Good point. You could use a thermal fuse or a bimetal switch which is resettable or better still use both for piece of mind.

[Chasm]

Good point. You could use a thermal fuse or a bimetal switch which is resettable or better still use both for piece of mind.

Since this is a failsafe I'd opt for a thermal fuse.
It should never trigger in normal operation, if it does trigger I really don't want the device to restart after a cool down period.

[Zero999]

Come to think of it, I agree, although having a bimetal trip at a lower temperature as well as the fuse, would save money when you're experimenting.

It's also good to have two layers of protection, the bimetal strip, just in case there's a mistake in the code and a fuse in case the switch contacts weld closed - such a system should easily pass both industrial and domestic safety standards.

[CafeLogic]

I thought about a thermal cutoff and even noted that in the source code. The problem is that it requires all sorts of experimentation. You have to find the temperature of the adjacent cavity when the toaster reaches a dangerous point and then get the thermal cutoff for that temp. Since I have zero intention of making a commercial product, I'm not going to bother. I overspeced the TRIAC but really the solution for my use is common sense. Make sure the area is clear of flammable objects, and don't go to lunch while it is reflowing.

I was actually surprised at the lack of protection in the toaster. Microwaves always have 1 or 2 thermal cutoffs plus current protection but this had nothing. Just the mains running straight to the element through a bi-metal switch. So anything I do is an improvement over the safety I experience while making toast.

[CafeLogic]

just in case there's a mistake in the code

On one occasion while I was doing the software it locked up while the element was on (stack overflow). Of course I have since enabled the watchdog but that is a very legitimate concern, especially when it is complicated. My opinion is that a single engineer simply does not have the resources to make a commercial product like this, you need multiple people to review it and you need adequate resources for product testing and refinement.

[CafeLogic]

Full details have been posted

http://www.cafelogic.com/articles-2/a-reflow-controller-for-soldering-with-a-griddle-or-toaster/

[Zero999]

I was actually surprised at the lack of protection in the toaster. Microwaves always have 1 or 2 thermal cutoffs plus current protection but this had nothing. Just the mains running straight to the element through a bi-metal switch. So anything I do is an improvement over the safety I experience while making toast.

There are different categories of safety protection according to the risk and consequences of a failure.

I think here it comes down to how long the appliance is likely to be left on unattended. It's unlikely you'll leave the toaster switched on while you're out of the room for a long length of time and there's normally a timer with a maximum delay of 10 minutes or so. In other words you'll probably smell burning plastic before the toaster catches fire and if it does you'll be there to turn off the power and throw a fire blanket over it. On the other hand a microwave could be left on for more than an hour unattended, which is long enough to burn the house down.

On one occasion while I was doing the software it locked up while the element was on (stack overflow). Of course I have since enabled the watchdog but that is a very legitimate concern, especially when it is complicated. My opinion is that a single engineer simply does not have the resources to make a commercial product like this, you need multiple people to review it and you need adequate resources for product testing and refinement.

Yes software can never be relied on its own for safety critical applications. There need to be other layers of protection, unless the design consists of three MCUs running different firmware, monitoring each other to ensure it fails safe. The whole design needs to be peer reviewed by several safety engineers.

You also need to be careful with devices such as switches, relays and semiconductors which often fail dangerously i.e. closed circuit. Often the best option is to buy a specially approved safety relay.

I like your write-up by the way and even though it's obviously potentially dangerous the disclaimer should stop you being sued lol.

[Mechatrommer]

...unless the design consists of three MCUs running different firmware, monitoring each other to ensure it fails safe.

sorry i'm not familiar with the design, but the way i see it, they need to be separated power supply circuit too. even if many mcus, but if one supply is shared, if that supply fail, all the mcu will fail too.

[Zero999]

sorry i'm not familiar with the design, but the way i see it, they need to be separated power supply circuit too. even if many mcus, but if one supply is shared, if that supply fail, all the mcu will fail too.

I don't see how that's practical, normally there's only one power source available and the device should be designed to fail safe even if the PSU fails.

[CafeLogic]

I think here it comes down to how long the appliance is likely to be left on unattended. It's unlikely you'll leave the toaster switched on while you're out of the room for a long length of time and there's normally a timer with a maximum delay of 10 minutes or so. In other words you'll probably smell burning plastic before the toaster catches fire and if it does you'll be there to turn off the power and throw a fire blanket over it. On the other hand a microwave could be left on for more than an hour unattended, which is long enough to burn the house down.

Unfortunately, I know exactly what the reason is. Some time ago, my then girlfriend decided to microwave her hot-pocket on top of a plate covered in strips of conductive foil (the bottom was labeled "do not place in microwave"). I come downstairs after she is yelling for me and there is pile of flaming smoldering ash inside the microwave. I could just imagine the massive arcs that were flowing in between the strips of foil. The microwave was dead, I later found the thermal cutoff was blown which probably saved my ass. So I am pretty sure all the circuit protection in a microwave is because of the potential for the high power electromagnetic waves to create electrical arcs and thus cause fires.

Add to that the electrocution risk. Here in the US, we have 120V for our normal appliances and 240V for big appliances. I am not going to say that 120V is not dangerous, because it is, but it takes an unusual set of circumstances to seriously hurt yourself with it, mostly involving water. The microwave on the other hand, creates many thousands of volts, enough to seriously injure someone in an instant on a dry day. The microwave is one of the most dangerous appliances in the home.

[Zero999]

You're right the microwave is the most dangerous appliance in the house, what's more using an isolation transformer to power it doesn't help because the high voltage neutral is bonded to the metal case.

Did you try replacing the thermal fuse? It may have been all right.

[CafeLogic]

Some stats to back me up. I found these a while back and thought they were interesting. This is from the US Consumer Product Safety Commission, covering electrocutions in the US in the year 2000. The most common is work place stuff involving arc flashes and construction workers from HV power lines, but this only covers the smaller percentage of consumer-product related deaths. Here is a partial breakdown.

Total deaths 150:

Cause	Number	My thought
Metal ladders hitting overhead powerlines	22	Ouch.
Wiring	25	Willing to bet this mostly involves missing or defective GFCI outlets in the bathroom
Air conditioners	8	This is 240V here (high humidity/rain could make trouble)
Pumps	7	Pool pumps, sump pumps and the like
Water heaters	3	240V
Refrigerators	2	I have no idea how this happens
Power tools	15	These are often 240V
Lighting	12	They specifically mention underwater lighting but could also be HV fluorescent ballast
Microwaves	7	These make up 50% of small appliance incidents, there are no toaster incedents listed

I attached the report in case you want to see the whole thing.

[CafeLogic]

Did you try replacing the thermal fuse? It may have been all right.

Yes, I found the fuse on eBay, worked like new.

[CafeLogic]

I just checked and the CPSC came out with a new report in June. Dramatic reduction, only 46 deaths in 2006. Prob more widespread use of GFCI in homes and products.

http://www.cpsc.gov/library/foia/foia10/os/2006Electrocutions.pdf

This caught my attention:

However, a
partial explanation for the increase of recreational equipment-related electrocutions in 2005 is a
single incident where four individuals were electrocuted while moving a tent, which happened to
come in contact with a power line. This one incident accounted for more than half (4 of 7) of the
recreational equipment-related electrocutions for the year.

That sucks

[Zero999]

240V in the US isn't really much more dangerous than 120V as it consists of two 120V live wires, 180^o out of phase with each other so the voltage relative to earth is still 120V. The person would have to be pretty stupid to make contact with both live wires simultaneously.

We also have 110V appliances used on construction sites in the UK but it's safer than the US because it's two 55V phases 180^o apart (single/split phase) or two 63.5V phases 120^o apart (three phase).

In the UK most underwater lighting has to be 12V to prevent electrocution.

[CafeLogic]

240V in the US isn't really much more dangerous than 120V as it consists of two 120V live wires, 180^o out of phase with each other so the voltage relative to earth is still 120V. The person would have to be pretty stupid to make contact with both live wires simultaneously.

lol, I never even considered that but you are right.

[RayJones]

But your eyes roll an extra 10 times per second despite our 240V over ground in Australia  

[Chasm]

On a second and closer look at the pictures the ARM board seems to be awfully close to the HV parts.