I'm looking into modifying a toaster oven into a DIY reflow oven, roughly using this guide:

http://www.whizoo.com/reflowoven

One necessary part is a thermocouple, but I'm unsure how to choose one. The kits of theirs that you can buy (I don't plan to) don't mention a brand of thermocouple. What I see on Mouser and Digikey are very expensive. I see some on Amazon that are cheap and look like the one that came with my multimeter. For example I looked at this one:

https://www.amazon.com/Perfect-Prim...ie=UTF8&qid=1519617001&sr=1-1&keywords=TL0400

It's rated to 752 degrees F, but for now I'm only assuming that's the rating of the tip, not the cord. And unfortunately I can't find any datasheet for it. Since the thermocouple tip will ideally be in the middle of the oven (I assume), part of the cord will be at about the same temperature as the tip. Should I assume the cord is rated to the same temperature as the tip? It says the wire has glass braid insulation, which kind of implies high temperature to me.

 

You can use a K type probe with an Ad595 chip, these are designed to give out 10mV per deg C, ideal for oven controls upto 500C// 930F, with an accuracy of +/- 1%.

http://www.analog.com/en/products/a...erface-amplifiers/ad595.html#product-overview


It all depends on which chip you want to control the heaters.. .

 

I'm looking into modifying a toaster oven into a DIY reflow oven, roughly using this guide:

http://www.whizoo.com/reflowoven

One necessary part is a thermocouple, but I'm unsure how to choose one. The kits of theirs that you can buy (I don't plan to) don't mention a brand of thermocouple. What I see on Mouser and Digikey are very expensive. I see some on Amazon that are cheap and look like the one that came with my multimeter. For example I looked at this one:

https://www.amazon.com/Perfect-Prim...ie=UTF8&qid=1519617001&sr=1-1&keywords=TL0400

It's rated to 752 degrees F, but for now I'm only assuming that's the rating of the tip, not the cord. And unfortunately I can't find any datasheet for it. Since the thermocouple tip will ideally be in the middle of the oven (I assume), part of the cord will be at about the same temperature as the tip. Should I assume the cord is rated to the same temperature as the tip? It says the wire has glass braid insulation, which kind of implies high temperature to me.

You have an interesting challenge here because there are quite a few different kinds of thermocouples, each one produces a different voltage to temperature output curve. This means that you must have the thermocouple match the calibration of the device that you use to display the temperature. The alternative is to have an accurate high resolution millivolt meter and a set of look-up tables that cover that type of thermocouple. The most common types are "J"' "K", and "T", each will cover the range that you need for a soldering oven. For your application just a simple welded thermocouple will work, so you can get by quite cheaply.

 

Long time ago I used to use a toaster oven (TO) to cure epoxy mounts. The problem I experienced with TO's was "Thermal Inertia". I know, no such thing. But wait a minute, there IS such a thing. The heating elements can heat up to 5000 degrees F. The TO runs the elements until it reaches a set temperature then switches off the current. Still, the element is quite hot and the temperature ramps way above the set temperature. Eventually it cools enough to where the temperature starts to fall. Here again you get thermal inertia. The temperature drops below a set point and the current comes on. But the element doesn't immediately heat up. So the temperature continues to fall until the element heats up enough to overcome the downward fall of the temperature. It's quite a thermal roller coaster inside a toaster oven.

They make and sell controllers that as you approach the set temperature the controller begins to modulate the power to the heating element, thus lowering the output before it gets to set temperature. Then it maintains a small current high enough to maintain the temperature. I looked into buying something like that but the cost back in the 90's was prohibitive. Besides, I found that I could use an aluminum box with a 60 watt light bulb inside it to sufficiently warm and accelerate the cure of my epoxy mounts. Of course, a 60 watt light bulb is not going to reflow your solder, so something like an oven would likely be the way to go. Just be sure you know how high the temperature will soar above your set point. Maybe, for instance, if it exceeds your set temperature by 100 degrees F you could set your temperature 75 degrees F below your target. The oven will achieve reflow temps and go above by only 25 degrees. Certainly long enough to reflow your board.

[edit] excess heat and excess dwell time can harm more sensitive components.

 

Long time ago I used to use a toaster oven (TO) to cure epoxy mounts. The problem I experienced with TO's was "Thermal Inertia". I know, no such thing. But wait a minute, there IS such a thing. The heating elements can heat up to 5000 degrees F. The TO runs the elements until it reaches a set temperature then switches off the current. Still, the element is quite hot and the temperature ramps way above the set temperature. Eventually it cools enough to where the temperature starts to fall. Here again you get thermal inertia. The temperature drops below a set point and the current comes on. But the element doesn't immediately heat up. So the temperature continues to fall until the element heats up enough to overcome the downward fall of the temperature. It's quite a thermal roller coaster inside a toaster oven.

They make and sell controllers that as you approach the set temperature the controller begins to modulate the power to the heating element, thus lowering the output before it gets to set temperature. Then it maintains a small current high enough to maintain the temperature. I looked into buying something like that but the cost back in the 90's was prohibitive. Besides, I found that I could use an aluminum box with a 60 watt light bulb inside it to sufficiently warm and accelerate the cure of my epoxy mounts. Of course, a 60 watt light bulb is not going to reflow your solder, so something like an oven would likely be the way to go. Just be sure you know how high the temperature will soar above your set point. Maybe, for instance, if it exceeds your set temperature by 100 degrees F you could set your temperature 75 degrees F below your target. The oven will achieve reflow temps and go above by only 25 degrees. Certainly long enough to reflow your board.

[edit] excess heat and excess dwell time can harm more sensitive components.

CAUTION: excess heat is likely to destroy sensitive components. AND, if you use lead free solder, the required temperature is hotter than if you use regular solder. My experience with using an actual small reflow soldering oven was that I can do a better job soldering by hand. Of course that was using a good magnifier and a needle point soldering iron, and having many years of practice soldering.. But with patience and a steady hand it is also possible to do surface mount by hand, and that includes those nasty quad flat-pack devices.

 

Generally there are oly one choice - to buy an expensive what warranties a metrological accuracy or buy a cheap which needs YOU to graduate the passing T to Voltage units.
For example - nonlinearity.
For example - damn contact materials for plug, thus changing the measure by ambient Temp. etc etc.
In soviet epoch we was welding the ball on two different wires and it worked more than well. But nowadays is cheaper to buy and forget about inaccuracies or need for individual calibration.

 

All temperature ovens have inertia. You need a PID controller to prevent oscillations.

 

All temperature ovens have inertia.

True. But gas ovens don't have nearly the thermal inertia electric ovens have. Infrared ovens are much more precisely controllable. Hot air ovens are also better suited for solder reflow machines than radiant electrical ovens.

 

I believe what you want to do is get out a clean white sheet of paper and a pencil and start by listing your temperature parameters and the type of temperature control you need or want. As to using a type K thermocouple they come in a wide range of flavors we might say. Something mentioned was " Low-mass K-type thermocouple" so that needs considered. Low Mass gets you a type K thermocouple which will respond quickly to changes in temperature. Thermocouples also come with or without a sheath and with or without grounded tips in the case of a sheathed design. Thermocouple placement in the oven is also critical.

As to uncertainty? A standard type K thermocouple will give the following uncertainty:
Type K:
MAXIMUM TEMPERATURE RANGE
Thermocouple Grade
– 328 to 2282°F
– 200 to 1250°C
Extension Grade
32 to 392°F
0 to 200°C
LIMITS OF ERROR
(Whichever is greater)
Standard: 2.2°C or 0.75% Above 0°C
2.2°C or 2.0% Below 0°C
Special: 1.1°C or 0.4%

Where are you located?

Also as a side note most of these conversions use an SST (solid state relay) to control the actual heater elements. Make sure any selected SSR is capable of the current load. Additionally if you plan to babysit the process no big deal but if you plan to leave the oven unattended you may want to consider an overtemp control.

As to your linked thermocouple.
Wire Insulation Identification and Application Guide is a good overview of thermocouple lead insulation and temperatures. My guess is your link includes the wire insulation limits.

Ron

 

RE:""You need a PID controller to prevent oscillations.""
Choices - PID regulator plus K-type TC measurer for 15-20 Eur SESTOS (+/- 0,5 C, ebay.com) or 100 Eur +/- 1 C for Omicron, or 25-30 Eur for (UK) Omega, or now for some 13-15 USD are REX [made on planet Earth], but I havent checket them yet.

 

As to your linked thermocouple.
Wire Insulation Identification and Application Guide is a good overview of thermocouple lead insulation and temperatures. My guess is your link includes the wire insulation limits.

Ron

Thank you, that gave me what I was looking for.

 

Any other questions or problems just ask.

Ron

 

As a side note years ago I came upon a PID loop example written in VB 6.0 (Visual Basic 6.0). The program was originally written by Max Seim of the 3M Corporation. I compiled the code which is based on filling and maintaining a water tank. The program can be run using automatic PID or you can go to manual control and try to maintain the tank level which is a set point. Download the compiled program and extract the files from the downloaded folder and install it on any Windows OS. I have run it on Windows 98 SE through Windows 10.If anyone has an interest in the old VB source code it can be found here. You will need a version of Visual Studio to view the .vbj files source code. As old as the program is you get a good idea of PID Loop control and can view things happening on the charts embedded in the software.

Ron

 

Never believe Anything on Amazon. Many people are just "sellers" and haven't a clue what they are selling.

"K" is an extremely common thermocouple. If you can't use K, you try something else. Atmosphere and temperature range are the first parameters to look at.

"C", a rare one is used for very high temperatures.

"R and S" have platinum in them so they are more expensive.

"T" is great for room temperature and cyogenics (very low temperature). A diode is even better for low temps.

RTD's are great up to about 200 C. They are a different breed altogether.

Thermisters are just plain wierd.

You have to watch thermocouple measuring systems because some don;t cover the full range of the TC. We had some that started measuring at 200 C.

Extension wire is just that, used to extend a thermocouple with junctions with the same electrical only properties.

Connectors are in a miniature, sub miniature and standard variety. They come in high temp versions and even chassis mount.

Grounded, ungrounded and exposed junction probes are also choices.

A Thermocouple relies on "cold junction compensation". The temperature of the connection to the electronics must be known exactly and that connection must be isothermal. Both at the same temperature. Thus these connections sometimes have a high mass or covered with insulation with the sensor. I used a measurement computing sensor interface where a fan in the vicinity would really mess up the readings.

www.omega.com is a very good source of information.

The amazon part referenced is likely the sheath temperature. If you were measuring the inside temperature of an oven, the insulation would have to go in there too.

I've used sensors in crygenics, melting of copper, RTD's for better accuracy and "T" type for room temperature measurements and crygenics.

This http://www.zallus.com/msp430-reflow-oven-kit/ so far has been the nicest re-flow oven controller that Ive seen. No experience. I just liked it.

PID is easy or it was easy for me to do from scratch.

 

At my previous employer I used an industrial reflow oven Atco PRO 1600 https://www.atco-us.com/products/item/5-pro-1600-smt-reflow-ovens . The thermocouples on the unit were all type K and had a fiber insulation. We ran some product to fairly high temperatures - around 300 Celsius and thermocouples worked well. They were made by Omega Engineering (www.omega.com). One other point to consider is the gauge of the thermocouple wire. Go with a 20 - 30 AWG wire. Thicker wire will have a delayed response.

 

At my previous employer I used an industrial reflow oven Atco PRO 1600 https://www.atco-us.com/products/item/5-pro-1600-smt-reflow-ovens . The thermocouples on the unit were all type K and had a fiber insulation. We ran some product to fairly high temperatures - around 300 Celsius and thermocouples worked well. They were made by Omega Engineering (www.omega.com). One other point to consider is the gauge of the thermocouple wire. Go with a 20 - 30 AWG wire. Thicker wire will have a delayed response.

Omega thermocouples have always worked well, for a whole lot of years, and so they are definitely a good choice. A local company, George Instruments, also produced very similar products that also worked well. But it has been 15 years since I dealt with them so I don't know the current status. Thinner unsheathed thermocouples do provide much faster response but they may not last as long i some applications, such as in flames. An oven is a gentler environment so their lifetime should be adequate.

 

THERMAX type K thermocouple wire.
Special Limits of Error grade wire,
24 gauge,
PVC ripcord.
# PR-KK-24.
Length of 50'.
$25.00
Free Shipping
http://www.ebay.com/bhp/type-k-thermocouple-wire

 

Thermocouple wire is another option that I have seen used. In fact I have also used it to make thermocouples. The limitation is in the welding process, because it is possible to contaminate the junction while welding the wires to whatever is being measured. But it is also fairly simple to produce an excellent thermocouple that delivers good results.

 

...The limitation is in the welding process, because it is possible to contaminate the junction while welding the wires to whatever is being measured. But it is also fairly simple to produce an excellent thermocouple that delivers good results.

Ball of any thermocouple is contaminated by definition.
When two alloys Ni+Cr and Ni+Al (Type K) are welded, in ball we have third alloy Ni+Cr+Al (contamination), which have thermoelectric power completely different from thermoelectric powers of original alloys.
So, in thermocouple we have 3 different thermoelectric EMF from couples: EMF1 (Ni+Cr and Ni+AL), EMF2 (Ni+Cr and Ni+Cr+Al) and EMF3 (Ni+Cr+Al and Ni+Al).
EMF2 and EMF3 are connected in opposite polarities, therefore they get canceled.
EMF1 remains.
Any contamination of welding ball does not affect on thermocouple accuracy.

 

Glass braid insulation is probably the best choice and generally reasonably inexpensive. Ceramic fibre is also suitable, but usually is more expensive. Teflon (PTFE) is the only polymer that might survive, but generally isn't suitable for operation over about 200°C so it is marginal for reflow work.

Do not be concerned with a temperature error of a few degrees due to the measurement circuit. The temperature measured by the thermocouple will, at the very best, be a reasonable approximation of the temperature of the components on the circuit board. Linearity error in any ordinary thermocouple will be far below the uncertainty that results from measuring the temperature of the air rather than the actual temperature of components. A part in 0402 is going to reach reflow temperature much faster than a TO-263. Acceptable results can be had, but there will always be a need for some experimentation and tweaking.