Better reflow hardware option

[kuon]

Hello,

I have an SMT AS-5060 reflow oven, but it's garbage. I mean, the frame/chassis is sturdy, but the temperature control is just useless. It heats the board super un-evenly, leaving burning marks at some places and not soldering others, it will do a different profile each time I fire it up...

For example, I did a board with those:

https://www.molex.com/molex/products/part-detail/pcb_headers/0908140204

and this paste:

https://www.digikey.ch/product-detail/fr/chip-quik-inc/SMD291SNL50T3/SMD291SNL50T3-ND/5130160

The UI of the oven went up to 230°, but all those molex connectors where burned/black, and half the paste wasn't soldered.

Only thing I was able to solder was low temp (138°) paste, because at this temperature I won't burn the PCB or components and I can keep the temp for a few minutes to ensure all the board is heated.

Anyway, I'm a bit disappointed because that thing was expensive and it's not doing great service.

My PCB are quite small, biggest I am working on is 15x10cm, so I'd say if it can accommodate 20x20cm it would be good enough.

I've seen many hack with toast over or hot plate, but I don't have the time to tinker with that.

I am wondering if you could recommend something that works, ideally I'd like to be able to reflow regular lead free without burning the board or SMT connectors.

[julian1]

I am trying to remember, if the problem with those units, is that radiant heat exposure from the exposed heating elements is higher than the programmed reflow temp?. eg. so that devices will get uneven heat depending on how isolated/proud of the board surface they are.

I wonder if a skillet/hot-plate, with a *good* tight thermocouple loop, using conduction heating from the base might be better - to avoid the radiant heat issue?

Put the thermocouple wire on the top of the skillet, or even attach it temporarily to the pcb board surface and then heat through from the base to ensure even heat. Has anyone tried something like that?     

Edit, hotplates for reflow are common. What I mean is a more strategic thermocouple placement, for control and to allow running proper profiles with mcu control.

[julian1]

Hmmm, this guy experimented with hotplate + thermocouple + mcu control.

But then discovered that the on/off power sequence to achieve a good heat profile and avoid overshoot only needs a couple of phases and the required timing could be determined experimentally. 

So the whole thing reduces down to, heating element + heating mass + stopwatch and manual switching. 

An IR gun thermometer pointed at the pcb surface temperature could help with consistency. Given the relatively stable temperature arc/profile and low thermal mass of the pcb, you could pre-determine the temperature turn-off point to achieve the desired profile.

 

[Peabody]

"This guy" is actually me.  I ended up donating the hotplate to my local  open source hardware group's makerspace, and reports are that it works well.  All my experimenting was done with leaded paste, so anyone wanting to use lead-free would need to work out their own power cycle.  And I've concluded that stencils are really essential for reflow work.  It's almost impossible NOT to put down too much paste if you do it manually.

[Doctorandus_P]

I do not have much confidence in the saw blade method mentioned above.

Without measuring temperature you way to dependent on process parameters. A bit of moving air of someone opening a door in the neighborhood or a new moon and it's way off.
You also won't be able to do 2 boards without letting it cool down completely in between (which may not be such an issue for hobby work) But keeping R12 in the reflow zone for 2+ minutes is not good.


You may be able to modify your existing oven into a much better (or at least much less sucky) one with a few simple modifications.

Put some mesh over the PCB, so heating is much less from direct IR radiation.
Put a small fan in it. with just enough uhmpf to move air around a bit. The idea is not to make it "blow", but just enough to mix the air of different temperature zones to an even temperature.

Fan motors do not like heat though. To get around that you can put a motor outside of the box and use a stainless (does not conduct heat well) steel axle through a small hole.

It's also possible that your oven is simply broken.
If this version has those quarts pipes with a thin wire coil inside, These coils sometimes move inside the tube which can lead to hot spots.

[kuon]

Well. I'm OK with spending a few hundred $ on a solution that work. I am just very disappointed by my current reflow oven (which did cost arount 1500$).

I don't really want to tinker with a hot plate, I'd like something that work out of the box. I was just wondering if hot plates with built in controller and cooling existed.

[coppercone2]

did you make sure its calibrated? it sounds like its grossly overheating the thing, or depending on how it heats, there might be a problem with the heater. I have seen hot plates that stop being uniform for whatever reason, but the heater is under ceramic so I can't tell you exactly why. When I looked in the manual, the hot plate specified a way of determining if its good using a thermal camera (uniformity spec).

Did you check for excessive oxidation on the heating element? is it rods like a oven, or a hidden heater (mesh or serpentine)? It could be that the impedance is too high and there is a hot spot. Heaters do not last forever and degrade.

Did you give it the 20cm spacing from other objects as dictated by the manual?

[Doctorandus_P]

I was just wondering if hot plates with built in controller and cooling existed.

Yes they do, key words like "pre-heater" " hot plate", "rework station" turn up many results in electronic shops, But I have to add, I never used them.

A few years ago I  saw such a device on one of the crowd funding sites. It had a flat top as often with a hot plate, but also an aluminum heatsink on the bottom which was horizontally mounted, so it won't work too well, until the built-in fan turns on. Firmware was explicitly designed for reflow soldering profiles.
 
A quick search:
https://html.duckduckgo.com/html?q=hot+plate+soldering+profile
Turns up a diy version with similar form as the crowdfunding version I saw back then
https://medium.com/home-wireless/how-to-build-a-hot-plate-reflow-oven-ecdc23ec9da7

I also did a few searches on a few of the crowdfunding sites, but was unable to find the thing itself.
It did turn up:
https://www.crowdsupply.com/pcb-arts/vapor-phase-one

which may be interesting to you.

[lukego]

Have you considered an Imdes vapor phase oven?

Just curious because vapor phase soldering has a vocal minority of enthusiastic proponents but doesn't seem to be widely used.

[kuon]

I did a very simple setup with an induction hotplate, a thick steel disc (1cm), and a thinner (0.5cm) aluminum disc on top.

I use induction because it has less inertia than regular hotplates and also it heats faster, and the temperature on the heated plate (my steel disc) is very well spread.

For reflow I start the plate, and I check the temperature with a small surface k-probe on the aluminum disc. When I hit profile top temperature on the alu disc, I move the alu disc to the side with clamps and I let it cool under my solder fume aspiration fan.

This worked perfectly well. I just did a few tries with few test boards to calibrate the profile, basically at what temperature I need to move the thing aside and what power to set on the hot plate. Now it's a manual process, but I don't think I need to automate it (it takes only a few minutes and I like being able to visually inspect that big pads like connectors are properly reflowed).

For some reason I though the profile would be very hard to get right, but the alu disc is very good at spreading the temperature evenly, and I was surprised how well the heat moves to the PCB. Also the induction plate can control the temperature with ease. For example, if I am like 20° too low, I step the power up and in about 5 sec I got my 20°C.