Homebrew oven feedback

[Jason]

Just like to report that after seeing Mike's vid with the Sanyo oven
I tried something similar. I bought a VERY small oven with a top and
bottom element for £19 and controlled it by hand whilst monitoring
with a K-type probe. This oven has metal baffles around the elements
to, I guess, shield the IR radiation and encourage heating through
convection instead. Not sure if this is a good thing or not.
I'm not convinced my K-type probe is registering the same temerature
as the PCB and solder paste so will investigate further.

[Gopher]

Sounds like fun, in a forced air convection oven they will brag about the efforts they go to to reduce or eliminate IR. That said IR is still a valid reflow method, albeit with some limitations, and is still used in some modern ovens for the preheat stage. That said heater elements in an oven will be kicking out all sorts of frequencies which would no doubt exacerbate the problems IR ovens can demonstrate, baffles are probably helping.
Some people also seem to be tinkering with DIY vapour phase which looks even more fun

[Mike]

I've never bothered with temp measurement - just set power to to 'only slightly glowing', wait til it reflows then turn off an open the door. I've done hundreds of boards with no issues - even when I've accidentally left it in for 6 mins instead of the usual 3-4, the only issue has been a discoloured PCB surface.
I don't use leadfree though, so I have a decent margin between "reflow" and "incinerate". I can't see why you couldn't pretty easily use Nitrogen or Argon with a toaster oven though to make leadfree more practical.
I've heard good reports from people using fan convection ovens, just as long as the fan isn't powerful enough to blow parts off the board.
I've never actually done tests but my gut feeling is that too much IR could cause scorching on parts with low thermal conductivity (e.g. FFC connectors) due to different absorbtion rates of different surface colours, however I've also seen people reporting success with ovens with elements that glow bright orange-yellow, so maybe not as big an issue as I think. Wouldn't be hard to do some tests of reflow time vs. element power.

Vapour phase looks an ideal solution, maybe using a modified deep-fat fryer with added cooling coils, but until someone comes up with a medium that is cheaper and more readily available in small quantities it's not going to be a sensible alternative to a cheap toaster oven  

[Gopher]

You can get low temp lead free paste. I'm sure it has its good and bad points and applications where it cannot be used but I have never used it. Most data sheets seem to point at going past 260C being pretty bad newts, as Mike says with SAC 305 melting at 217C, an oven without any control belting out lord knows what leaves you in a much riskier position than with leaded paste. A lot will depend on what you are happy with, what you are trying to reflow, how long you want it to last, and where the finished product is going. If its a prototype going to your test bench who cares is some things have gone a little brown, as long as ramping up and down isn't so quick you get thermal shock most techniques are going to work, if not to the degree one might expect in full production.

[Jason]

I'm using a 217deg Multicore paste. If I follow the recommened
profile for the components then I get a really faint browning of
the board but I'm happy with that.

Using the same paste with the hot air rig (before I bought the
oven) I was having problems with the flux bubbling and popping
0402s off before the metal liquified and the surface tension
kicked in to hold it in place. I don't know why that was (and, no,
they weren't moving due to too much airflow! ;-) )

I was looking at cheap import PID controller but a friend of mine
pointed me to a UK supplier that are doing a 8 segment / 8 prog
unit for 60 odd quid.

At some point I'm going to have to upgrade the oven to a
much bigger unit if I want to get reasonable size panels from
the RV1S cooked evenly.

[SteveW]

Vapour phase looks an ideal solution, maybe using a modified deep-fat fryer with added cooling coils, but until someone comes up with a medium that is cheaper and more readily available in small quantities it's not going to be a sensible alternative to a cheap toaster oven  

I ran a few boards with my fryer & Galden setup. All was fine, but vapour definitely escaped. Time for a cooling hat - I'm thinking of an aluminium pyramid (made from bent sheet, riveted shut) with a vigorous PC fan on top, blowing down the sides. Might need some extra heatsinky features inside or out, we'll see how it goes. Tuck a PIR matrix device in the top, like a MLX90620 (16x4 array, good for 300oC) to watch board temps, and a couple of themocouples with a heat path to the hat, to monitor vapour level, and I think I'm set for a super-cheap VP. Might stick a cooling loop in the fluid at the bottom, too, to cool things off at the end of a run.

For a small board, and no urgency on cycle time, I'm not planning on lowering the board into a vapour cloud, more creating a vapour cloud on demand, then letting it condense back out at the end. The lack of a fast chilldown may concern me later, but I don't really want to be stirring the air inside the chamber (do I?)