There are almost as many ways to reflow SMD boards as there are life hacks. Today we are adding another method to the list. [Dasaki] Turned a halogen spotlight into a surface mount oven, and that’s all the bells and whistles. Watch the video under the break.
In fact, we’ve already seen low-tech versions of this hack, but we don’t want to use it every day. [Dasaki] It takes 100 planks to complete, so it’s worth putting in the effort to do it well.
The spotlight was a natural choice – it had its own built-in camera, and when [Dasaki] tested it, it warmed up to 300°C at 5°C/sec and showed no signs of stopping. Contrast that with toasters that start complaining at around 250C if they can. Aside from the lamp itself, everything else is a standard temperature controller, a solid state relay, and a thermocouple to interact with the microcontroller. The code is on his GitHub.
We’re partial to frying pans ourselves, but Bill Hurd prefers professional halogen/infrared appliances. In fact, there are many ways to get the job done, from the hacky to the almost professional, and you’re bound to find a reflow oven design that suits your budget and tolerance for bullshit. The [Dasaki] halogen light oven looks like a worthy addition to our list.
Well, actually it may not require any additional electronics at all, just a timer. Of course, not as beautiful as temperature control, but it will be a very simple, cheap and compact reflow oven.
Unless your profile needs to go from low to 300C in 5 seconds, I doubt it will work!
Without the feedback loop, the lamp will most likely go into melting mode… it’s not designed to reflect that much heat into it…
Really? What can be damaged? There are metal reflectors, quartz light bulbs and tungsten elements, and it’s white hot. I can’t imagine where it will hurt.
Despite all the losses from radiation and convection, tungsten is still white hot… guess what happens when you reflect heat back into it… it gets even hotter.
When tungsten is heated, its resistance increases, which reduces energy consumption. I’m not sure how close the lamp is usually to the edge, but incandescent lamps have a natural passive negative feedback that partially corrects for voltage and ambient temperature variations.
hanelyp: Only to a certain extent. According to https://www.tungsten.com/materials/tungsten/, at 2727°C (3000 K), which is close to the operating temperature of a halogen lamp, the resistivity of tungsten is 90.4 µOhm-cm, and if you increase it more than at 1000 °C, it only rises to 108.5 µΩ-cm. This change is only 20%.
5 C/sec faster than recommended by almost all profiles. You can use a slower-than-usual ramp to cut the reflow cycle down to half a decent amount, but it’s not really as efficient as a proper soak period. However, if your heating element is ramping faster than usual, you must have some way to slow it down. Fixed dimmers and timers might work, but electronics aren’t expensive at the moment.
Great idea. Can it be turned upside down so the heat comes from below? I like hot plate reflow because the heat comes from below, so large components don’t absorb as much heat. When the components are only on the top side, everything warms up evenly, which is usually the case. The disadvantage is that the entire PCB must be heated. You need to pay attention to the maximum temperature so as not to damage the PCB, and the profile needs more time to overclock. My problem is that even the IR burner has a slow response time, making it difficult to control. I would like to use a halogen electric stove for better response, but they are expensive and hard to find. Using a halogen work light seems to solve my problem if the halogen work light can be set to heat from below. Maybe some PCB support wire or IR clear window can do it. Any ideas how to do it?
The FR4 PCB is a good insulator. Too much energy is wasted heating the board from below. Because of its inefficiency, hot plate welding is rarely used in industry.
I bought one of these cheap when they were on sale at Walmart, presumably because that particular model was discontinued, but I could use something like this:
The glass cover is based on a non-contact IR thermometer and acts as an excellent IR reflector. Because the emissivity of the dark Teflon surface is quite high, I used the same infrared thermometer to plot the temperature profile. A series of 30 second power adjustments at full on followed by 30 second power adjustments at full off proved to be very effective for approximately the appropriate ramp length and firing time, with one minute for the final fast ramp of solder flow. Cooling is achieved by carefully removing the board shortly after the solder flows. Not an ideal cooling period, but I’ve had no issues with components being exposed to more heat than necessary.
Note that I only have to make one plate at a time and it is in the center of the pan, so that’s all I do. Temperature fluctuations at various points in the pan near the heating element can result in wildly different temperature profiles for each plate position, such that hobbyists cook one plate at a time.
The one in the kitchen is hopeless to cook, because the delay of the mechanical thermostat is too large, and there is a large thermal delay between the heating element and the thermostat sensor. Thus, you must constantly correct the temperature manually.
I also use one for pickling, since I pickle at 70 degrees C (below boiling), I solved the delay problem, but hung the pickle pan over the bottom of the pot with 2 inches of water in it to allow the thermal mass to pass through. However, I let the temperature stabilize for about 20 minutes before pickling. I also have a temperature sensor in the water and I can adjust the temperature manually.
So I guess what I’m trying to say is that these pans are probably the worst when it comes to temperature control.
LWIR thermopiles (probably 5.5 to 12 µm) can’t read it unless it’s really special glass. You measure the temperature of the glass, not the temperature under the glass. https://www.newport.com/medias/sys_master/images/images/h07/h37/8798003331102/EG-pg15-8b-600w.gif
I hope this article is a joke 
This light emits mainly IR waves, which means black components heat up more than white ones, which will cause uneven temperature distribution on the PCB, heating the board to 300 C in 5 seconds is a Bad thing Thought that it does not fit any known solder paste profile
“Like all incandescent lamps, halogen lamps emit a continuous spectrum from the near ultraviolet to deep infrared. Since the filament can operate at higher temperatures than non-halogen lamps, the spectrum is blue-shifted, resulting in a higher efficiency. Color Temperature Light with higher energy efficiency.
I’ve also read comments about raising the temperature to 300°C, which indicates how much heat the lamp can put out in the initial pot life test, not the profile it intends to use for the actual reflow. Obviously SSR will be used to reduce heat dissipation.
Lamps produce all spectrums of light, but they heat infrared waves, and they can’t even heat a PCB, which means you can’t solder all the components under the same conditions. Some components (large mass, white) can be heated up to 230C, and others (small black components) can exceed 250C, the difference between heating parts with IR oven can be 5-10-20C (depending on the degree to a different degree), namely so they are not used in manufacturing, hot air convection ovens are the tools people use to ensure +-1C heat distribution between PCBs for all parts in reflow.
Objects that appear black or white to humans may look completely different at other wavelengths. Since humans can’t really “see” if an object is black in infrared light, it’s useless to say that black objects heat up faster. Incidentally, almost everything is actually black at 10 µm (thermal radiation), except for metal surfaces and certain types of crystals such as ZnSe, ZnS, Ge, GaAs… such as Flir or those used by laser cutters . Of course, this does not mean that infrared ovens are “not used in production” for any other reason.
Well, I’m wondering how heat can be generated without infrared waves. The black problem is always there unless you are heating from below or THT and only using a wave of solder.
Regarding the temperature profile: you are right, but I read it like it was just a speed test. A connected controller can easily slow it down.
However, infrared reflow ovens are one thing. So I don’t think the “black component overheating” issue is actually a problem – I’m not sure how black the black plastic is in IR anyway.
Just because an object is black in the visible spectrum does not mean that it is also black in other spectra (eg IR).
IR is not a component issue. There are complete IR reflow ovens in the industry, and I use a vapor phase reflow oven at my company with a built-in IR preheat (and glue dryer) oven. Heat is always distributed unevenly due to the thermal conductivity of the various components/materials. You can solve this problem by simply waiting for a certain temperature.
IR heaters and IR reflow ovens are different types of beer. The temperature difference is 30 degrees, 20 years ago IR reflow ovens were very common in the industry, nobody produced them for the last 10 years because there was no demand for them in the industry.
Personally, I would love to have the latest boardroom equipment, the most advanced commercial 3D printers, and everything imaginable in my garage. Although reality does not allow this.
Infrared reflow ovens have been used in industry for decades. If anything, a halogen power supply will be less efficient because it has less IR.
Of course back in the days of SnPb as there were no moving parts/fans to dissipate heat and it was cheaper to manufacture, but I haven’t seen any manufacturer of industrial reflow ovens offer IR ovens since ROHS came out with lead free ovens in 2006, why??? Because the window for lead-free processes is narrow, and setting up profiles in an IR oven is time-consuming and always compromises for each board shell. I just googled and couldn’t find anything other than a Chinese desktop reflow oven. This forum post says that’s it: “IR reflow is yesterday’s technology. There are many problems associated with the principle of IR heating, namely the ability to transfer heat (shadow effect) and uneven heating of the printed circuit board. To avoid this, forced convection ovens have been developed. Reflow oven for SMD soldering must have forced convection principle.”: http://www.smtnet.com/Forums/index.cfm?fuseaction=view_thread&CFApp=1&Thread_ID=9783&#Message39257 Reflow oven for SMD soldering must have forced convection principle.”: http://www.smtnet.com/Forums/index.cfm?fuseaction=view_thread&CFApp=1&Thread_ID=9783&#Message39257 The reflow oven for SMD soldering must have the principle of forced convection. The reflow oven for SMD soldering must be forced convection. ”：http://www.smtnet.com/Forums/index.cfm?fuseaction=view_thread&CFApp=1&Thread_ID=9783&#Message39257 ”：http://www.smtnet.com/Forums/index.cfm?fuseaction=view_thread&CFApp=1&Thread_ID=9783&#Message39257 “: http://www.smtnet.com/Forums/index.cfm?fuseaction=view_thread&CFApp=1&Thread_ID=9783&#Message39257 “: http://www.smtnet.com/Forums/index.cfm?fuseaction=view_thread&CFApp=1&Thread_ID=9783&#Message39257
The visible part of the spectrum is converted into heat when absorbed by black. How big is the gap? Who knows, the OP seems to be happy with his reflow lamp.
Perhaps there is a way to solve the infrared problem by using small aluminum panels to distribute heat through natural convection.
I always thought of them as portable electric fireplaces. Ideal for heating canteens, light is just a welcome by-product. Using them as light sources is a life hack for me!
Yes, I mainly use 500W halogen bulbs as infrared heaters when working outside in the winter. Just shining on you, they help them so much.
In a way, very inefficient. This is great if you are installing drywall in an unheated building during the winter. It’s also great to drive away the winter blues by lying a little under the “sun”.
When I worked at a construction site, Mexican drywall workers would set up lanterns on the ground and use them to cook delicious meals for lunch. They’ll even put a tortilla on top of one and add beans, meat, etc. Almost sounds like a racist stereotype as I write this lol but I’m not kidding
The next Hack on Topic will definitely feature IR LEDs instead of halogen bulbs. To add to (1/cool) I say B-)
You don’t need reflective foil to do this. All closed cavities behave as ideal blackbody emitters, regardless of the material’s natural emissivity. the science! Also, instead of using an SSR, you can use a dimmer switch to turn the light on and off, shortening its lifespan. If you want to get creative, use a smart dimmer with a built-in serial interface and you can still automatically control the temperature.
You can purchase optocoupler/triac drivers with built-in zero-crossing switches. Using one of these and angle modulation (MSB first) is probably best for halogens.
I seriously doubt it. Of course, halogen lamps lose the benefits of the tungsten-halogen cycle when operated at low temperatures, but the rate of evaporation is greatly reduced. Of course, halogen lamps last twice as long as standard tungsten lamps, but tungsten lamps operated at 75% of rated voltage will last 10 times longer than full voltage operating lamps.
I have seen dozens of tungsten coated halogen bulbs because they have been dimmed throughout their lifetime.
It is the temperature of the quartz body that determines whether the halogen cycle will work, and the temperature of the filament determines the rate of evaporation.
It is a myth. We have been using dimmer halogen lamps in the TV industry for many years with no apparent side effects. These are not “special” bulbs or dimmers.
I think it’s my confirmation bias. Used a dimmer on a cheap bulb that came with the kit and the bulb burned out within a week. Then I bought heavy duty service bulbs, did not use a dimmer, they lasted much longer. 
Foaming occurs due to temperature. Everything inside the cavity has more or less the same temperature. You must insulate the work surface, but it does not matter if the inner surface is foil or black.
Post time: Nov-15-2022