Automated toaster reflow project
eagletalontim
Posts: 1,399
Hello all once again! This is a new project for me, but I need to get this accomplished one way or another. I have recently ordered some circuit boards and for some reason put 0603 parts on the board be accident. I could redesign the board, but if I am able to solder them with a toaster oven, why not give it a shot! I have already purchased a cheap 470W toaster from my local Family Dollar that has 2 heating elements, one on top and one on bottom. My goal is to use a Prop or an SX chip to control the toaster. If this toaster will not work, I can always put it back together and return it 
First things first. I need a way to monitor temperature inside the oven using some kind of temperature sensor (thermistor) that can handle the temps. If at all possible, I would prefer an already configured temperature reading device. Don't want to buy something I have to configure. For example, the temperature sensor connects to an ADC chip and I have to learn what the output is at X1 degree, then again at X2 degree and so on. I have no way to accurately read temperature inside the toaster now, so getting multiple readings for configuring is not an option. Is there ANYTHING available that is like this and not expensive? Not sure if anything is already made, but I was thinking a temperature probe that would go inside an oven and beep at a certain temperature. I could "hack" that device and use the signal from the beeper to trigger an event in the Prop.
First things first. I need a way to monitor temperature inside the oven using some kind of temperature sensor (thermistor) that can handle the temps. If at all possible, I would prefer an already configured temperature reading device. Don't want to buy something I have to configure. For example, the temperature sensor connects to an ADC chip and I have to learn what the output is at X1 degree, then again at X2 degree and so on. I have no way to accurately read temperature inside the toaster now, so getting multiple readings for configuring is not an option. Is there ANYTHING available that is like this and not expensive? Not sure if anything is already made, but I was thinking a temperature probe that would go inside an oven and beep at a certain temperature. I could "hack" that device and use the signal from the beeper to trigger an event in the Prop.
Comments
I'd just stick them in place with some water soluble flux and get one end tacked down. Then reposition if I had to and tack the other end. I have found that there are occasions where a surface mount resistor or a surface mount capacitor can be a quick modification and very handy. One can solder in extra by-pass caps, or rig a pull-up resistor without having to print a whole board.
Toaster ovens do work well. There are 3 key points.
A. Never to be used for food again.. as you might be working with toxic metals.
B. Timer control... usually via a relay or SSR on the AC mains.
C. Temperature control .. I suppose one could just use an oven thermometer, but you can get fancier.
Be aware that just because the PDFs give all sorts of fancy curves for proper soldering doesn't mean you have to engineer such tight controls. The main objectives are just to get items to solder and not to overheat the components when doing so. The homebrew solution is not ever going to be as tightly controlled as an automated production line for thousands of boards.
I actually fear that the fine pitched larger chips are going to demand a lot of care. I am getting older and I prefer DIPs. But there are some good reasons for reflow projects: such as R/C electric air projects that need small light boards.
Temperature control could just cut out the A/C mains via relay. But if you to provide a whole temperature cycle, it requires more sophistication.
Industrial oven temperature sensors are bi-metalic wire and various types are used for various temperature ranges. The wire itself is a long story in special handling techniques. I have doubts that a thermistor is going to do it all.
Solders have a range from 90 to 450 degrees centigrade. So pick a thermocouple that will preform well at that range. Get a premade on that can be installed or inserted in the toaster oven. And have the electronics interface with a microcontroller.
http://en.wikipedia.org/wiki/Thermocouple
http://www.google.com/search?num=50&newwindow=1&site=&source=hp&q=frying+pan+smd+reflow+soldering&oq=frying+pan+smd+soldering&gs_l=hp.1.1.0i22i30l2.3125.19679.0.21894.24.20.0.4.4.0.281.2856.7j9j4.20.0...0.0...1c.1.15.hp.ST7zocUgloI
Practice with some scrap boards to see what it takes to remove and re-solder some components without damage. Even an ordinary cast iron pan can be used for a quick experiment.
470 watts is really low for a toaster oven. Sounds like a warming oven instead.
Rick
Not sure if it gets hot enough just yet. That is what I am trying to figure out right now. I do have a cheap walmart meter that has a thermocouple, but I am not sure of the accuracy. I was also wondering if I could use the thermocouple to connect to the Prop and read the temperature. I have a 555 timer that I have used in the past to do an ADC conversion, but I am having problems getting the output signal to change based on temperature changes like I had last time. My battery on my o-scope just died so it has to charge before I can do anymore testing. One thing I do know is the thermocoupler has a resistance of 16 Ohm at 70 degrees F and the resistance increases as the temperature rises.
I'm not aware of a 555 thermocouple reader.
I had thought a thermocouple required a special amplifier and ADC chip or a specialized chip to read the thermocouple directly.
Do you have a thermocouple you plan to use? I've use this one from SparkFun for a drying oven and it seemed to work okay. I haven't used other thermocouples so I don't know how it compares to other thermocouples.
16 Ohms sure seems like a strange thermistor value.
There was some discussion about thermocouples recently. Make sure and read Tracy Allen's posts. I linked to some other thermocouple discussion in post #2 of the thread.
Since I have gotten this to work so far without having to purchase anything, I may stick to trying to configure this setup to work for my needs. The only way I can test the temps is by a cheap handheld IR temp sensor I have had for over 8 years. Not sure how accurate it is, nor the best way to hold it to get an accurate reading. If I move even 1/2" the temp difference can be +/- 20º F
Also, the thermocouple I have outputs a very small voltage which is what I am using though the LM324 to output higher voltages.
If you google "toaster oven reflow" you will find lots of solutions using various micro-controllers. I have always wanted to make a Propeller based controller. With it's multitasking ability, you could make a really nice one that would not be dependent on a particular oven model.
- A standard K Type thermocouple is what you want to use for temperature monitoring. More than one would help resolve cold spot issues in the oven.
- Solderpaste manufacturers specify the recommended reflow profile for processing an assembly with their paste. These profiles will lay out the temperature and timing for proper reflow and can be applied to a toaster oven in the same manner as an actual reflow oven.
- A standard 15 amp SSR can be used to properly turn the oven on and off to gain appropriate temps. The oven is usually left on a very high setting and then you control the on/off cycles to achieve specific temps.
- Most toaster ovens don't have fans, so you must be cautious when driving elements for long periods. The oven temperature in the center of the oven may be just reaching your target, but because of the element temp, you can easily overshoot your desired maximum temp because the element temperature will be much hotter then the center of the oven. Cycling the elements in a slow PWM fashion can reduce the overshoot risk.
- To eliminate the need for a specific model oven by others who want to use your controller, your controller will need the ability to characterize the oven and apply offsets to it's operating parameters. Here is my suggestion for characterizing a toaster oven.
- Attach two thermocouples to a set size of copper clad PCB (IE: 4"x4", 0.062" thickness). One on top, and one on bottom.
- Attach a thermocouple to the oven, a set distance from the elements (IE: 1.5" below the elements)
- Run a preset series of heating profiles to characterize the performance of the elements. This will gauge the element heat up time, heat conductivity of the oven, and heat transferability to the bottom side of the PCB (assuming there are no bottom elements)
- From the data obtained from this set of heating profiles, the oven's element performance has been characterized and consistent profiles can now be made by the controller because it will know the elements heat up time, overshoot time/temp, etc.
- A characterization process allows an end user to use almost any toaster oven with your controller. This will be a major win for a toaster oven controller.
- Controllers really should have a display to plot temp vs time for real-time monitoring of the reflow profile. This is one reason I think the propeller is ideal.
I wish I had more time to devote to a Propeller Oven controller. I think you could easily design and sell them if you can just nail down the software for the characterization.Some info on my process for reflowing PowerTwigs is detailed here on this excellent thread by TonyP12:
http://forums.parallax.com/showthread.php/130196-How-I-do-SMT-soldering-in-toaster-oven?p=983326#post983326
Regarding 3 minutes to melt solder....
Time in the oven and exposed to high heat is as important as the temperature. The longer time in the oven, the higher risk of damage.
http://www.reprap.org/wiki/Thermocouple_Sensor_1_0
You can buy the above kit, or just the chip. But the wiring has to be absolutely right. With thermocouples, the wires are the sensor, have to be an exact length, and have good junctions. If not, nothing will work right.
Does this sound like a feasible option? I may need a bigger toaster if the heat does not rise quick enough. Got to do some tests on that. I guess I should draw a graph up to see how this toaster does.
So where to start? I know I will need a thermocouple and I like the features of the AD595. Question is, will any Type K thermocouple work? If so, how can I see if I already have a type K thermocouple? Is there a full schematic of an AD595 connected to a Prop? I am not sure what ADC chip will be needed to go with this nor how to wire it to be as accurate as I need. The kit in the above post is nice, but it will not connect directly to the prop without an ADC. If I need good accuracy, I don't want to keep buying parts just shooting in the dark hoping I can get something to work.
I have a MAX6675 chip but I have used it yet. It seems to me the MAX6675 would be less expensive than getting the AD595 and ADC chip.
I used the schematic to SparkFun's kit to see how to wire up the AD595. The MCP3002 was one of the first chips I learned to use and I didn't want to try to figure out how to interface with the MAX6675 chip so I figured I'd more likely have success with the AD595 chip.
One thing I've learned is that normal mechanical relays are not very durable when used to PWM an oven. Even with 0.1Hz frequency (oven turned on once every ten seconds) I went through two mechanical relays in as many years. Both relays failed in a closed state which ruined the drying test in progress.
MPJA has some relatively inexpensive SSRs that work well. I haven't had any of the SSR I use fail.
Edit (I hadn't seen your last post): The main advantage of the MAX6675 I see is the price. If you look at the some of the other thermocouple threads, you'll see Tracy Allen has recommended some other chips. The only way I know of telling what kind of thermocouple you have is to measure it's voltage at a known temperature (easier said than done). I haven't tried this myself.
If you go with the MAX6675, you can either buy a small PCB from SparkFun to break it out to a DIP footprint, or use the smt area on a Propeller Project board. I initially didn't like the smt area on the Project board but I've found it very useful since most of my projects end up have at least one smt part.
I have been browsing on Digikey for an SSR, but you have linked a 40A 480V relay. Is that much amperage really needed? The toasters I have looked at range from 1200 to 1500 watts, and the one I have now is 470W. This would mean at the most (1500W), I would need to handle 12.5A. Here is one rated at 16A @ 240V. Would something like this work : http://www.digikey.com/product-detail/en/S216S02F/425-2414-ND/720364
EDIT : If I am going to control each element individually, the amperage requirement will be much lower, correct?
Unfortunately the company seems to have gone out of business. They were located in Florida. They were called Silicon Horizons. If you do a Google search for it or even on YouTube you can see it work.
No, it is not.
You need one rated for the amps you'll use. 12.5A seems about right. Most of our circuit breakers are rated at 15A so I'm pretty sure you don't need one over 15A. I don't think the 8A SSR from Parallax would be enough.
I haven't purchased a SSR from Digi-Key before. You do need to make sure and get one for AC. You can't use a DC SSR on an AC line (I think most high current SSR are AC).
The relay you linked to at Digi-Key looks like it should work (I'll probably get a couple with my next Digi-Key order). You many need to use a heatsink with it.
I'm pretty sure you can drive a SSR directly with the Prop. No transistors are diodes needed like normal relays.
While SSR would be ideal, I think a mechanical relay would work fine. They're generally cheaper.
So with the Thermistor and the SSR taken care of, I need a way to tune the timing and heating. My thoughts on this are through a Visual Basic program on my computer that can track the temp and time and allow me to adjust the PWM through the program for different setups. I already have another project communicating with my computer through a serial connection, but I would like to try to use USB since my laptop does not have a serial connection and my toaster does not sit next to my computer. Is there any documentation on how to connect to the Prop through USB? I don't want to use a USB to Serial cable. I am thinking a MAX232 is what is used, but I can't remember where I saw this before.