Building a reflow oven. Have questions.

[MarkS]

Doing research on this topic, it seems fairly simple. However, several questions have arisen that I cannot find answers to, so I hope someone here can help.

1.) While I understand what "zero crossing" is, I'm not sure why switching on zero crossing is important. One project I've seen had its own zero crossing detector circuit, but I can buy an SSR with built in zero crossing detection. Why?

2.) Is there a recommended location within the oven to place the board? My naive inclination is the center of the oven.

3.) Same as #2, but for the thermocouple. Should it be installed from the top or back? Centered or to the side? Does it matter?

Thanks in advance,
Mark

[mindcrime]

Doing research on this topic, it seems fairly simple. However, several questions have arisen that I cannot find answers to, so I hope someone here can help.

1.) While I understand what "zero crossing" is, I'm not sure why switching on zero crossing is important. One project I've seen had its own zero crossing detector circuit, but I can buy an SSR with built in zero crossing detection. Why?

Maybe consult this? http://www.crydom.com/en/tech/newsletters/solid%20statements%20-%20ssrs%20switching%20types.pdf

In particular, note this bit:

Zero-crossing relays are ideally suited for most commercial and industrial loads, such as resistive heating elements, lamps and ballasts, and any other load with low initial impedance or capacitive
characteristics.

2.) Is there a recommended location within the oven to place the board? My naive inclination is the center of the oven.

To some extent the answer is "it depends". Ideally there would be no thermal gradients inside the cooking chamber, and the answer would be "it doesn't matter". In practice, there may be "hot" and "cool" spots in your oven. If so, you may need to tweak the board position to be in an ideal spot. If that comes up, once you identify the "sweet spot" you'd probably want to be consistent with your positioning. Thoroughly insulating the inside the oven with reflective insulation that reflects IR radiation may help smooth out any temperature gradients.

3.) Same as #2, but for the thermocouple. Should it be installed from the top or back? Centered or to the side? Does it matter?

One thing somebody pointed out to me, is that you really want to track "board temperature" not just "air temperature in the chamber". They will differ, although I'm not sure by how much. Anyway, to get a simulation of your "board temperature" without actually taping the thermocouple to your board, you can take a small piece of scrap PCB material, tape the thermocouple to it with kapton tape, and place it as near as possible to the board you're actually soldering. This is the approach I've been using and so far it seems to work reasonably well.

Also, FWIW, there's been a lot of recent discussion around building reflow ovens in this thread. If you haven't seen it yet, you might find some useful thoughts by browsing through there.

[MarkS]

You have been EXTREMELY helpful! Thank you!

[mindcrime]

You have been EXTREMELY helpful! Thank you!

No problem. To be fair though, to the extent that I know anything about this stuff, it's mostly due to the generous help I've received from other EEVBlog users, in that other thread. Folks like @fourfathom, @coppercone2, @beanflying, @wizard69 and many others have been critical in helping me figure out how to get something useful put together.

[jmelson]

2.) Is there a recommended location within the oven to place the board? My naive inclination is the center of the oven.

3.) Same as #2, but for the thermocouple. Should it be installed from the top or back? Centered or to the side? Does it matter?

Well, in general center the board between multiple heating elements.  Closer to the element gets hotter.

I have never gotten a thermocouple in air to work.  I've burned several boards.  What I did was get some micro-sized
thermocouple extension wire, stripped and twisted the ends and poked it into a plated through hole in one of the boards.  That way, the thermcouple reads the actual board temperature.

I also use a ramp and soak temperature controller I got on eBay.  These allow you to program linear temperature ramps and temperature holds.  I ramp to 180 C for a couple minutes, hold at 180 for one minute, then ramp to 230 C in a couple minutes, hold for one minute and cool to room temp.

I use this system to reflow from one big board up to 8 small ones at a time, and have done over 2000 boards with it.
I got a large GE toaster oven with 4 heating elements.  Two are above the rack, running left to right, toward the front and the back.  Two more, same arrangement, below the rack.

Jon

[mindcrime]

I got a large GE toaster oven with 4 heating elements.  Two are above the rack, running left to right, toward the front and the back.  Two more, same arrangement, below the rack.

Sounds very similar to what I have, other than being GE, vs Black and Decker. Mine likewise has 4 total heating elements, two above the rack, and two below, running left to right. It's a 1500W unit and has no problem getting up to the desired temperature in time, especially after adding some insulation to the inside of the oven.

[SuzyC]

I am building my own reflow oven as well, but trying to minimalize what it takes to get consistent results. I have a 1200W two-element oven with a 11-Liter capacity. I ebayed a 7-cm 3/16 round probe type K thermocouple and a thermocouple amp with digital output to interface with my PIC mcu.

I drilled a suitable hole and pushed the thermocouple probe into the center/middle of the control area metal side to measure temperature.

I am able to reliably read the thermouple output with my MCU programmed to send temperature readings in ASCII to my PC via a serial port.

The 7-in long thermocouple probe was positioned a few mm above the grill. The 4x4cm perfboard was centered in the oven in every test I made.

I made several experiments with the same piece of single-sided FR4 perfboard and used 60/43 paste applied with a toothpick.

I made each experiment with placing several SMD 0805 resistors at each corner and near the center of the perfboard.

I used the most primitive method to manually control the time/temperature just using an on/off switch and a timer.

Everything seemed to be working fine and I got good results several times, especially after tuning my reflow profile each run.

The result was that reflowed connections looked reliable,  even if a few pads were not looking exactly perfect.

There were some balls forming and at first, some tombstoning, but the finished board flaws could be quickly touched-up with a soldering iron.

The problem is, I reused the same perfboard four times, after desoldering the parts, I would add paste, SMD parts and repeat.

It was on the fifth experiment, around 20 seconds after the final ascent to the reflow melting point that the perfboard suddenly caught fire and flooded the room with noxious fumes.

It took 15 minutes of windows open cross-circulation breezes to get rid of the smell.

A wise man once said, "Never drink too often from the same well!".

Is it at all possible to use a thermocouple probe?

Was the fire caused by drying out the board or possibly because the thermocouple probe failed to report the hotter real temperature of the perfboard?

[mindcrime]

A wise man once said, "Never drink too often from the same well!".

Was it possibly because the thermocouple probe failed to report the hotter real temperature of the perfboard?

My guess is that the problem was simply re-using the same piece of material over and over again.
I'm not a chemist or anything, so I don't know all the details, but generally speaking heat affects materials in a chemical sense. And repeated exposure to high levels of heat can actually make some materials more susceptible to combustion. 

If you put a piece of material in, and afterwards notice it starting to turn brown or char, or starting to curl up around the edges, I'd throw it away and not use it again.

[SuzyC]

Now you tell me!

[mjkuwp]

I made my own many years ago and have used it dozens or maybe even hundreds of times with success.

I put a copper sheet on the wire grill and taped my thermocouple to that.  This works for me but I stick to surface mount parts only.

I also blackened/tarnished the copper to try helping it to absorb more of the energy.   my donor oven was a cheap one but it does have a convection fan which I run.

it is a pretty easy project... or it can be a really difficult project depending on the scope you set for yourself.

one more thing, my oven is nominal 1500 Watts but I took it apart and shortened the heater wires to reduce resistance and drive the current up.  I ended up with around 1800Watts.  I was aiming to reach near 20 Amps which is the maximum my circuit can draw.

[MarkS]

Again, I appreciate the help. Thank you, everyone!

[rstofer]

I bought a Black & Decker Infra-Wave toaster oven.  This is no longer in production but others will work.

I used an earlier version of this controller.  I mounted the display and start button to the front of a metal enclosure.  I used a wall wart for controller power.

https://www.rocketscream.com/blog/product/tiny-reflow-controller-v2/

I mounted a zero-crossing Solid State Relay inside the right side cover since all of the heater wiring passed through this area.  I don't recall the rating but I would aim for about 20A.  That would leave a lot of head room.  Something like this should work well:

https://www.wolfautomation.com/ssr-panel-mount-ip00-24-280vac-oper-range-25a/

Note that it doesn't have a touch-safe cover.  Try to find one that does.

Here's one with a touch-safe cover but it isn't zero crossing AFAICT:

https://www.wolfautomation.com/ssr-cover-1/

Finally, I added high temperature insulation blankets to all hollow spaces of the oven.  This was absolutely necessary if I wanted to get up to temperature before Christmas.  Blankets were from McMaster-Carr but I don't remember which variant.

I poke the thermocouple into a thru-hole.  I'm interested in board temperature, not air temp.

Works well!