4
\$\begingroup\$

I see lots and lots of builds of "temperature controlled iron" that simply either:

  • adds a timer to an uncontrolled iron
  • adds a dimmer to an uncontrolled iron

The first is the same as turning on the iron later. And the second is the same as paying too much for a 25W iron (because that's what your 75W iron will become).

Are there any projects out there really adding closed loop temperature control to an unregulated soldering iron?

I think this can be done way cheaper than buying an overpriced Weller, but before I start something I would like to see where others failed/succeed and I'm failing miserably in finding any preceding work on this.

\$\endgroup\$
  • 1
    \$\begingroup\$ Actually the second point is incorrect: The dimmer provides convenience of actually controlling heat. Are you going to unplug your 75 W once it gets too hot or just adjust the potentiometer on the dimmer? Still, good question. \$\endgroup\$ – AndrejaKo Jul 5 '12 at 18:08
  • \$\begingroup\$ can you expand on the dimmer? most builds, such as this one hackaday.com/2008/08/05/how-to-the-hackers-soldering-station does not sense temperature in any, so it has no 'control' over what it's outputing. it just make the 75W iron works as a 25W. if you leave the dimmer at minimum but leave the iron idle enough to fill it's heat reserves, it will probably be as hot as if you left the dimmer at max and used it right away. \$\endgroup\$ – gcb Jul 5 '12 at 18:25
  • 1
    \$\begingroup\$ @gcb the iron is in constant equilibrium with the heat it loses to the surrounding air, so it won't be anywhere near as hot at fully "dimmed" versus fully on. \$\endgroup\$ – Nick T Jul 5 '12 at 18:29
  • \$\begingroup\$ @gcb The dimmer method of operation relies on user to detect when the iron has enough heat. Here's an example: You're soldering a linear regulator and related components in SMD package using a 75 W iron. Since the iron would get too hot on 75 W to solder SMD capacitors, you set it to 25 W. Once you start soldering the regulators large tab, you notice that the iron is unable to produce sufficient amount of heat to increase the temperature of the tab and use the dimmer control to increase the power to 75 W. With more power, the iron solders the tab. \$\endgroup\$ – AndrejaKo Jul 5 '12 at 18:31
  • 1
    \$\begingroup\$ Yet another point which is problematic with irons which have no good heat limiting capabilities is that the temperature will rise with the increase of heat. So first joints will be at high temperature which may be too high for flux to operate or which may destroy components and as iron cools, you can reach the stage at which there's not enough heat to complete the current joint. \$\endgroup\$ – AndrejaKo Jul 5 '12 at 18:36
3
\$\begingroup\$

For an automatic temperature control, you would need three basic components:

  1. A way to sense the temperature of the iron at the tip
  2. A way to control the power supply to vary the temperature
  3. A microcontroller to read the temperature and determine how to adjust the power supply to achieve the desired temperature

Since you asked for preceding work here is a nice build log of a DIY mod to a soldering station including a lot of nice pictures, schematics, and firmware.

And Dangerous Prototypes looks to be putting together a kit of some sort. But some of the links are dead and information is missing. But they do have a nice custom PCB for the project that is currently out of stock as of this writing. I'm not sure if this is a work in progress or just an unfinished and forgotten about project. You could probably email them and ask for an update.

PS: I agree with AndrejaKo's assessment on the dimmer controlled iron. For some people, that's all they want'need. Here is a nice Instuctable on how to do just that.

\$\endgroup\$
  • \$\begingroup\$ That pcbheaven project is just what i was looking for! thanks \$\endgroup\$ – gcb Jul 5 '12 at 19:06
4
\$\begingroup\$

You would need some type of temperature sensor. Temperature controlled irons likely use either thermocouples or thermistors (probably the former as they are generally perfectly happy at high temperatures). If you were able to wire a thermocouple down near the tip of your iron, you could simply read the temperature and implement some controller (PID, fuzzy, whatever) to manage the power.

Alternatively, you may be able to read the temperature based on the change of resistance of the heating element. Monitor the voltage and current passing through the element to get the resistance (R = V/I), and use the temperature coefficient of resistance of the element (probably Nichrome) to calculate the temperature. It would take a great deal of care to get this to be decently accurate, as you'd have to subtract out the lead and possible connection resistance (both of which may also change with temperature).

In any event, your results are bound to be fairly poor. Beyond simple temperature control, there are often many other significant differences between a Radio Shack iron and a (e.g.) Hakko. Namely, the proximity of the heating element to the tip (good irons will have their heating element inside the tip), and the "density" of the heating element (the entire 65 W heating element in a Hakko is about 3/16" dia. x 1/2" long), both of which serve to couple the heat generated into the tip extremely quickly and efficiently. If you simply want a temperature-controlled iron, just buy one. The only reason to make one is for some design experience (but it's been done before, so really why bother? Spend your time on something totally new).

\$\endgroup\$
  • \$\begingroup\$ I dont like your wording in your first paragraph. IMO, PID is not directly related to fuzzy-logic. Sure, you can implement a fuzzy-logic based PID controller, but a 'standard' PID controller will probably do fine, too. Written as it is now, it reads like you have to use fuzzy-logic. \$\endgroup\$ – PetPaulsen Jul 5 '12 at 18:41
  • \$\begingroup\$ @PetPaulsen from what I've read, PID controllers are just a subset of fuzzy logic controllers. In any event, it doesn't really matter. Any sort of "controller" would work. \$\endgroup\$ – Nick T Jul 5 '12 at 18:44
  • \$\begingroup\$ Oh, that's new for me. Do you have any sources? Would be interesting to learn more about it. \$\endgroup\$ – PetPaulsen Jul 5 '12 at 18:47
  • \$\begingroup\$ interesting you mention that iron, i considered it a while ago and it was out of stock since forever. they were recommending a hakko one for $120. ended up buying this one this aoyue which i will be dissasembly soon \$\endgroup\$ – gcb Jul 5 '12 at 19:00
  • 1
    \$\begingroup\$ I think fuzzy logic is a subset of PID, not the other way around. (BTW, fuzzy logic was a big hype in the 90s, but I haven't heard the term in more than 10 years. AFAIK fuzzy is as dead as a dodo) \$\endgroup\$ – stevenvh Jul 6 '12 at 5:53

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.