I bought a very small electric skillet at Goodwill; it's more or less square, about 5" on each side. The skillet is made of aluminum with a non-stick coating on the inside. The heating element is circular and about 4" in diameter. It's in a groove underneath the pan, and is held in place with screw clamps.

I plan on using it for SMD soldering, and have read that one of the major drawbacks of such devices is that they have hot spots over the element, and the PCB sometimes has to be moved around in the skillet. My question is, "Would a 4" square of 1/4" thick copper laid in the skillet be of any significant benefit in evening out the heat?"

Thanks.

 

Will you be using Pam cooking spray as a flux? I made a heater for press fitting ball bearings on to shafts, out of a electric sandwich maker. Ended up having to do the aluminum plate thing to get an even heat. Drilled and tapped blind holes in the bottom of the plate and drilled holes in the sandwich maker. Did that to get full heat transfer in the added plate.

 

You can put a chunk of metal in the bottom of the skillet, or a small ceramic tile, and I've even seen people using a layer of dimes (coins). Anything that provides a bit of distance and added thermal mass and some reduced conduction will do fine.

 

I didn't ask about a ceramic tile, but I should have. It's lots less expensive than the piece of copper. I'll try it first.

Thanks.

 

What about using a toaster over instead ????

 

What about using a toaster over instead ????

That may be the way to go, but my reading seems to indicate that the skillet method works better. Not to mention that I found my electric skillet at Goodwill for less than $4.

 

That may be the way to go, but my reading seems to indicate that the skillet method works better. Not to mention that I found my electric skillet at Goodwill for less than $4.

Yeah I did some reading and found that alot of people prefer the skillet method over the toaster oven.. .Sorry about that


Also look up hotplate reflow .

 

I am back to using the skillet to reflow SMT parts on PCBs. Although I have had some success, there are some problems.

Looking at a more or less "generic" heating profile, I see that reflow occurs at about 220 degrees C. So, I bought a dimmer, and experimented with it until I could more or less duplicate the "generic" heating profile (as measure on the surface of the skillet with an IR thermometer.) But when I tried it out, I quickly discovered that the top surface temperature of the PCB did not reflect the surface temperature of the skillet. I had to heat the PCB much longer than the heating profile indicated in order to get reflow. In addition, I found that at least part of the PCB was hot enough to become soft (FR4.)

Which leads me to the conclusion that (despite what I read on SparkFun, and some other places) heating the bottom of a PCB in order to reflow solder on the top is not really a good way to go. I don't really want to go the PID route, and wonder if anyone here has experience with manually controlling a toaster oven with a dimmer. If so, what have you found?

Thanks.

 

@tracecom

In the early days of "kinder, gentler, safer" chemical laboratory classroom, a micro scale organic chemistry set was available. No Bunsen burners, no expensive Fiberglas heating mantles with variacs, no, the kit came with heater wells and sand. The sand is the heat transfer media. Just push the test tube deeper for more heat in the electrically heated well. Once you got the feel of depth vs response, it was quite easy to control a distillation of a few milliliters. Time constant on heating and cooling was extremely fast. Boil/no boil in a second or two.

I am sure the same could work for your board. Some fine silica sand should not be wetted by standard solder. Slide the board into the hot sand, slide it out. Tap off like you know what. Done. Or, that's how it works in my imagination.

 

@tracecom
I've been thinking about a small project around the same lines ... but I'm going to try to use a 300W halogen worklight instead of a skillet. In fact, I'll most probably do that today.
I'll take some pictures and let you know how it went.

 

Alright @tracecom, today I did what I promised, and here are the results:

I designed and built this small PCB adapters for SMT devices. They're designed to fit in an 8 pin DIP socket and look like this:




What I did, is I cut an aluminum template (I own a CNC router ) to make it fit very snug:





So I inserted the adapter in the template, and found the soldering paste so hard to maneuver that what I did was tin the traces with it, using my ordinary soldering station set at the lowest temperature.



This is the 300W floodlight that I used for my little experiment



I placed the chip on the PCB, aligned with the tinned traces:




And placed the template with the PCB and the aligned chip on top of the floodlight, and then turned it on. Everything went really bright in there!:




In the end, everything worked ok, and the chip was soldered to the PCB. But the thing is that I couldn't see the moment in which the solder re-melted... the truth is, I think that (at least in this size of chip) it's better just to tin the traces with the solder paste, and then press the chip's pins one by one with the soldering tip set at the lowest temperature.
BTW, I haven't tested the chip yet, so I don't know if the lamp's heat damaged it or not. But I doubt it, I previously measured the lamp's temp for 5 minutes, and the maximum temperature reached was only 275°F




I plan to test the circuit tomorrow... I'll let you know how it went.

 

Alright @tracecom, today I did what I promised, and here are the results:

I designed and built this small PCB adapters for SMT devices. They're designed to fit in an 8 pin DIP socket and look like this:

View attachment 89411


What I did, is I cut an aluminum template (I own a CNC router ) to make it fit very snug:

View attachment 89412



So I inserted the adapter in the template, and found the soldering paste so hard to maneuver that what I did was tin the traces with it, using my ordinary soldering station set at the lowest temperature.

View attachment 89413

This is the 300W floodlight that I used for my little experiment

View attachment 89414

I placed the chip on the PCB, aligned with the tinned traces:

View attachment 89415


And placed the template with the PCB and the aligned chip on top of the floodlight, and then turned it on. Everything went really bright in there!:

View attachment 89416


In the end, everything worked ok, and the chip was soldered to the PCB. But the thing is that I couldn't see the moment in which the solder re-melted... the truth is, I think that (at least in this size of chip) it's better just to tin the traces with the solder paste, and then press the chip's pins one by one with the soldering tip set at the lowest temperature.
BTW, I haven't tested the chip yet, so I don't know if the lamp's heat damaged it or not. But I doubt it, I previously measured the lamp's temp for 5 minutes, and the maximum temperature reached was only 275°F

View attachment 89417


I plan to test the circuit tomorrow... I'll let you know how it went.

I fully expect the IC will be okay. What was the purpose of the hole in the aluminum plate?

 

To solder SMDs like this, tin ONE trace, solder the one pin. Press down lightly on the chip with a screw driver or tweezers (or finger) then touch the trace one more time to allow the smd to fall firmly on the pcb. Then tack all the rest of the pins.

It is hard (impossible) to get all pins soldered firmly to the board when the chip cannot sit squarely on the board - which happens when all the traces are tinned and all scan not be feel owed at the same time.

That size chip is easy with a soldering iron.

I understand it was a feasibility test but the comments of tinning all the traces is kind of in the wrong direction.

 

To solder SMDs like this, tin ONE trace, solder the one pin. Press down lightly on the chip with a screw driver or tweezers (or finger) then touch the trace one more time to allow the smd to fall firmly on the pcb. Then tack all the rest of the pins.

It is hard (impossible) to get all pins soldered firmly to the board when the chip cannot sit squarely on the board - which happens when all the traces are tinned and all scan not be feel owed at the same time.

That size chip is easy with a soldering iron.

I understand it was a feasibility test but the comments of tinning all the traces is kind of in the wrong direction.

Thanks for the advice. It is thoroughly appreciated.

I've been soldering smts of that size for a few years now, but now I know that my mistake was using 60/40 solder for that purpose... I had to develop that technique because I couldn't control the soldering process very well at first. What I was forced to do was tin all the traces, place the chip, and then solder the first pin by pressing very firmly against the PCB. What happened next is that the pin would give way and bend, and the solder would do its job. Then I'd do the same thing with the pin diagonally opposite to the first one, and after I made sure that the chip was properly aligned, I'd work on the rest of the pins.

This was better than just tinning one trace and then work on the rest, because for a good solder joint it's necessary for it to get under the pin. Otherwise the solder will not join the trace and the pin very well, and will tend to break at the only junction they actually have, which is the tip of the pin... 60/40 solder is too viscous and has a surface tension so strong that it becomes hard to handle and tends to form lumps and blobs at the ironing tip. Also, some circuits that I make are exposed to strong vibration... that's how I know about the break at the tip problem.

Anyway, things became much harder when I discovered that some chips (especially analog and ADCs) are very sensitive to temperature. And 60/40 solder does not flow very well below 400°F... it takes longer for the tip to melt it and perform the weld. That extra time that the chip is suffering at high temperatures also affects it. I was eventually able to control the pressure being applied, the temperature, and the amount of time to a point in which I'd get an 80% success rate (that is, not seriously damaging the chips) on chips with a 0.025" pitch.

But then I discovered that you could actually buy low-temp soldering paste in syringes instead of jars! And this is my first experiment with the thing...
So on this first try I found out that soldering paste is not very cooperative when you try to "paint" traces with it... it's lumpy and too "dry" and "sandy" for it to work like an ink of sorts... as I had expected it would. The thing is designed to be used with stainless steel stencils, which I don't have. That's why I decided to tin the traces first, as I had done in the past....
But lo and behold! On this first try I also discovered that soldering paste is extremely fluid! and what you have just recommended makes perfect sense now... I'm going to form a small solder blob at the ironing tip, and with that high fluidity I think that the solder is perfectly capable of getting below and behind the pins even when they're fully resting on the traces.

I am sure going to try what you've just recommended next time I have to solder SMTs. Which will most probably be this friday. I'll let you know how it went.

 

I fully expect the IC will be okay. What was the purpose of the hole in the aluminum plate?

The hole in the aluminum plate holds the PCB in place, and also allows it to make direct contact with the floodlight's glass below it. Of course that last small advantage is practically overkill, since heat will easily travel through the aluminum anyway.
Also, it's much easier to handle the aluminum plate with the inserted PCB than just that tiny PCB by itself.

 

The hole in the aluminum plate holds the PCB in place, and also allows it to make direct contact with the floodlight's glass below it. Of course that last small advantage is practically overkill, since heat will easily travel through the aluminum anyway.
Also, it's much easier to handle the aluminum plate with the inserted PCB than just that tiny PCB by itself.

Okay. I had thought that you were going to apply the heat to the component side of the PCB. That would be more like a reflow oven and might work better. Most of what I reflow are small PCBs (10cm x 10cm) with a mix of SMT and through hole parts. I reflow the SMTs and then hand solder the through hole parts. Thanks for your posts.

 

@cmartinez

This might be what you are having troubles with 60/40 solder...

...using my ordinary soldering station set at the lowest temperature.

The goal is to melt the outside without heating the inside.

You are the grill master, so think about it like trying to make a steak that is charred on the outside and rare on the inside. That cannot be done if the coals are not super hot.

It is best to use a 700-degree F fine-point solder tip, touch it for a 1/4 to 1/2 second and the solder flows oil thin instantly and wets out on the parts well. Using thin solder helps too, so you reduce the thermal mass and don't have to wait for the solder to melt.

 

I've soldered tiny parts like this using only hand tools, either a small tip iron or a hot air pencil. Of the two I perfer the iron as the pencil heats the part too much IMHO, where the iron concentrates heat just where it is needed. Hot air is only needed when doing small BGA parts, and I have no method for large BGA parts.

If your board needs to be tinned (the really cheap adaptor boards I get from China don't need such) I would recommend tinning first, then use solder wick to remove all but a skim coat of solder. Then re-apply on each lead when heating it.