SparkFun will be closing early at 3:30 Friday 5/27 and remain closed Monday for Memorial Day (5/30). Orders placed after 2pm MT on Friday (5/27) will process and ship out on Tuesday (5/31).
If you've ever played with SMD devices, you know what a pain it can be to solder tight pitch pins. Yes, we know others have used toaster ovens for reflow soldering purposes. Yes, we have our own reflow oven that cost 30 times as much. But where's the fun in that?
We have really built upon the shoulders of others here. There are many more knowledge individuals that pioneered the field. We are only here to re-hash and hopefully educate those in the art of toasting.
As of writing - we haven't reflowed anything. Well, at least not in our 110V fire starter. Basically, I just want to share the information I have obtained and the layouts I have created. And I'm too impatient to wait for the v2 PCB...
As usual, there are quite a few photos. I heard Hong Kong is going to Gigabit internet connections. Isn't it time for you to upgrade that 56k modem? My apologies to all users still on slow connections.
All the usual warnings apply: A toaster oven can kill you. It could certainly burn down your house. Lead is bad. Don't eat from a toaster oven that's cooked PCBs.
SMD Reflow Tutorials:
What can you do with a toaster oven?
These cheap-o units from Tar-chez can cost a meager $50. A toaster can hit 230C with relative ease. This is hot enough to melt solder paste allowing home hobbyist the ability to 'cook' PCBs. All that is needed is a controller to heat the toaster to a specific profile. This is all because of a posting on the SFE Support Forum...
Picture of the premo profile off our $2300 reflow oven. By the way, this reflow oven didn't work.
All we aim to create is an embedded system that takes in temperatures and actuates a relay that controls the heating elements - effectively controlling the temperature. That's where the Reflow Toaster Controller comes into play. But first, we must purchase and immediately destroy a perfectly good toaster oven!
I think I paid $60 at Target (local chain - basically WalMart) for this guy :
Oster Model 6293
Inside tray. A dry-run test proved this baby can heat up like crazy. The lights dimming across my house gave me a fuzzy feeling in my stomach as well.
The front control panel - soon to be circumvented.
Inside. Notice the two heating elements on top and bottom. It's got to be better than just two, right?
Don't immerse in water?!? Damn. 1500W is important to know though.
Getting there. It was actually pretty hard. I'm not sure if I can get it back together...
Guts. Back-side of the control panel knobs.
The annoying bell.
The entire oven seems very simple, yet well designed to withstand the extreme use of 10-12 years of high-temperature stressing.
An overhead shot of the AC transformer?
I tore into the timer dial hoping to find some sort of a relay that could be digitally controlled.
The timer dial exposed: In retrospect, this was a foolish thing to think. All low-cost ovens such as this Oster operate with the bare minimum of parts. A power relay just wouldn't be included.
Controller Time -
Ok - so that's the oven we would like to use. But how to control it close enough to reflow expensive PCBs?
First we need a thermocouple. Thermo-what? A thermocouple is a type of temperature probe that can withstand the very high temperatures (>100C) of the toaster oven. There are a lot of different thermo-whats out there. The type-K thermocouple is the most common (at least in my opinion). Type-K refers to the two types of metals that are combined to create the thermocouple sensor. I don't dare bore you with the specifics. The Analog Devices datasheet for the AD595AQ IC explains the entire principle very nicely.
Not surprisingly, you can pick these up from quite a few vendors on the internet. The tricky part is taking the voltage change distributed by the physical nature of the thermocouple and turning that into a temperature. Segway back to the AD595AQ! This very expensive DIP package was built for such a thing. Attach a K-type thermocouple, power the thing at 5V, and you've got a 10mV/C output! There are many other ways to skin this cat, many are cheaper, but all are far more complex then this tight package. And it WORKS! The readings agree nicely with our infrared temperature gun.
We are only looking for +/-1 or 2 degree C accuracy over the oven. So we can take the 10mV/C voltage levels directly into the 10-bit ADC (4.88mV per bit) found on most 16F PIC microcontrollers. If you've read anything on the Spark Fun Support Forum, you'll know I love the 16F88 PIC. Simply because it is a work horse, it's got an ADC, it's cheap, and it can be boot loaded!
So we've got the microcontroller to read temp data, now we need to turn on and off the oven. With a 1500W oven on 120VAC, that's 12.5A that needs to be turned on and off. Any normal relay would probably melt or fail after a few hundred cycles. While my co-worker (Pete) really wants to see the first option occur, I opted for a relay that wouldn't tack weld itself together. A short search on Digikey produced many options. My only recommendation is not to confuse the coil type with the paddle rating. I thought I needed a 120VAC relay. I accidentally ordered a relay that required 120VAC to actuate. My poor PIC couldn't actuate a relay like this. Instead, find a relay with a coil rating of 5V or 12VDC. Most relays will be able to control AC voltages, it is the coil voltage that needs to match your DC circuit. We chose a 12VDC coil the second time around with a 30A rating. Testing in time will tell - we haven't got them yet...
So the PIC can take temp readings, and based on those readings, turn on/off the relay. Here is the schematic for version 1 of the ReflowToaster. The PCB came out very well!
Simple through hole design
Just a few small problems with the layout. I messed up the silkscreen/mechnical diameter of the two bulk capacitors. I also forgot the current limiting resistor into the base of the BJT. I thought it would work... It doesn't.
Here it is fully populated.
In action on my desk...
At this point, I surprisingly have the board fully functional. The limiter was the relay. When comparing the size of my dinky relay (PCB Mount - 12VDC / 10A) with the massive heat shielded connectors of the 120VAC main wires, I decided to pause and contemplate my life. I'd rather not screw this one up.
So I decided to order a bigger, heftier relay with Quick Connect lugs. I'll still have to cut into the main wires and splice with heat shield, but I'll feel a lot safer in the end.
Instead, I decided to re-design the toaster board. A fellow user on the SFE Support Forum actually recommended this project. I initially wanted to make a board that relied entirely on a Windows VB program to control the profile. After an hour or so of programming, it was obvious it would be much more functional some day if it was stand alone. But I'll keep the RS232 interface for boot loading and data logging.
So I added a 2x16 LCD, three control buttons, and a hefty on-board relay with solder lugs. Here is the schematic. The PCB tripled in size, but I think it will be very nice in the end.
Yea, it doesn't look like much - but it's going to be very cool.
So now we wait for the new PCBs to come in. The code was really simple and is not yet complete. The basics are there though : ToasterControl-v01.c
We hope to put a kit together for under $100 for those that would like to replicate our fire hazard. We'd love to get your feedback on the forum! Let us know about your reflow experiences as well.