So I'm probably retarded, but I cannot seem to wrap my head

>>1007656
Get a job and move out of mummy's basement

>>1007656
have you actually seen this done before? Reflow oven temperature zones are static, so ramping a toaster oven doesn't really compare. You would need to invent your own oven curves since the ramp times wouldn't be similar at all.

If you are serious about doing it then thermocouple each board and use that for the feedback loop directly. You'll be able to tell when you hit thermal equilibrium to end your preheating zone. I don't know how critical the TAL is, since your oven probably wont be able to match how fast a reflow oven gets to it, but the cooling zone is very important so that your solder crystalizes correct.

You'll need something with SCR heating elements to get the control you want. I'd just pay an SMT shop to make whatever it is you are trying to make though.

File: kester[1].jpg (94KB, 594x409px) Image search: [Google]

94KB, 594x409px

>>1007729
shits all over the internet for hobby grade shit. And this is what I was talking about for reflow curve, although I'm sure temperature profile would be a more appropriate name. I was under the impression that there's one provided in the solder paste datasheet to inform what temperatures and times the paste needs.

And of course I have a thermocouple for feedback. How else would I know anything about whats going on inside?

The elements are basically hooked up to mains through a solid state relay which is pulsed by the arduino. What duty cycle for what stage is hard to say without a lot of other data, so that's why I wanted to use PID to adjust it on the fly to approximate the temperature profile.

I'll check out the thrift shops around town for a quartz halogen oven to gut and maybe throw some fiberglass insulation around the inside walls to improve performance.

I've done this with a SSR and a thermocouple, just using the built-in heating elements. Purely open loop though (well I closed the loop manually). I don't think a longer ramp hurts anything. Do some tests with solderpaste on scrap PCB, my setup tended to run hot for some reason.

>>1007749
Your components will also have there own curves. And none of them will agree. It's literally impossible to get it all to speck on consumer grade crap. Just get close enough and be happy. It will all work just fine anyway.

>and maybe throw some fiberglass insulation around the inside walls to improve performance.

idk if you want to do that, some of the drop offs on the end are hard enough to match w/o extra insulation.

I always liked Ben Hecks version https://www.youtube.com/watch?v=NCGzKDTFBSQ

I just wish he'd shared his code for it.