How to improve reflow with close small components (diy reflow oven)

[aaronhance]

I'm trying to reflow some PCBs in my diy toaster/reflow oven. I've installed a temperature control module, see below, but I think the temperature might by higher than indicated, as the silkscreen turned brown and there's a bubble looking thing in the solder mask. Another issue is some of the components have moved to the vias, 0603 and 0805, I assume this could be caused by poor/messy placement? Another issue is probably that these aren't my final components, but I wanted to test first without £25 worth of stuff, so I scraped some off an old pcb I had lying around. Kind of glad I did.

I'm using sn67 and I went to a peak of 210 indicated, took 9mins to reach, then cooled down to 170 before opening to cool faster. Thermocouple in poking in the side, horizontal center, slightly above vertical center, maybe 2 centimetres in.

Any advice before I try again, 8 fails left with test components :S.

Images attached.

[jmelson]

I have done well over 2000 boards of various types in a GE toaster oven I got from Walmart about 17 years ago.  I use an Omega ramp and soak temperature controller I got on eBay.  I poke a micro-sized thermocouple into a plated through hole in the board so the controller is controlling the actual temperature of the board.  My oven has two horizontal steel heating elements above the rack, and two more below.  I mostly use SAC305 solder. but have used leaded solder too.  The flux is really important.  For lead-free, GC10 ha proven to work extremely well, and I have had a few other brands that don't work well at all.  I set up a profile to ramp up to 180C hold for a minute, then ramp up to 236C, hold for one minute, and then ramp down.  That peak temp is for the SAC305, you can reduce the peak for leaded solder.  For leaded, a peak of about 215C or so might be good, just to be sure EVERY component has reached at least 186 C for a few seconds.  Watch closely through the window to see the solder change from grey to silver.  Shine a flashlight in at an angle so you don't get a reflection in your eyes, so you can observe the melting take place.

Your component placement before reflow is not real great.
Jon

[newto]

Whatever temperature you're measuring, it definitely isn't the temperature of the PCB, you absolutely fried that board. I've been using a second had toaster oven and a basic amazon K-type thermocouple thermometer kapton taped to a blank pcb on the side.

Even that is more than is absolutely necessary considering the size of the components you're using, just watch it for 5 minutes and turn it off and open the door when you see all the solder melt. Adding a soak phase at around 150C is better practice, but you don't have big heavy components that would take a while to bring up to temperature like inductors or mosfets with heatsinks.

If it took 9 minutes to get to temp, I'm guessing your gauge is off by like 50-100C, especially to get it that torched.

Edit: For the placement, more care is probably needed, but it wouldn't surprise me if the flux was boiling and bubbling and pushing components around. Covering the vias with solder mask is also a good idea, unless you plan to use them all as test points. You should also be able to put the USB port in for the reflow, once you figure out the temperature and timing...

[bostonman]

I have done well over 2000 boards of various types in a GE toaster oven I got from Walmart about 17 years ago.  I use an Omega ramp and soak temperature controller I got on eBay.

I've been curious about using ovens to reflow boards. Long story short, I recently acquired a new Ninja oven from someone who got it as a gift and didn't want it. My intention is to keep it should the need to reflow a board ever arise, but would I need to tap into the control panel for this temperature controller to work?

[MR]

this stencilling is a disaster... fix the stencilling / applying solder first.

[48X24X48X]

When the thermocouple is hanging in the air, nearing a setpoint when the controller pulls the handbrake, the air temperature drops easily and hence not able to reach the setpoint and experiencing an oscillation. Therefore it will try for a long time to reach the setpoint. Hence, your controller is taking 9 minutes and that too is by luck it finishes. Plug the thermocouple tip into the through hole. Also, your paste looks suspicious.

[Kasper]

I have done well over 2000 boards of various types in a GE toaster oven I got from Walmart about 17 years ago.  I use an Omega ramp and soak temperature controller I got on eBay.

I've been curious about using ovens to reflow boards. Long story short, I recently acquired a new Ninja oven from someone who got it as a gift and didn't want it. My intention is to keep it should the need to reflow a board ever arise, but would I need to tap into the control panel for this temperature controller to work?

I started with a used $20 IR oven and haven't used it enough or had enough failures to justify an upgrade.  I use a thermocouple probe on my DMM
and control the oven manually.

Print out the reflow profile of the solder.
Tape probe to similar PCB and place it near the PCB(s) being reflowed.
On until a bit below soak temp.
Off until a bit before soak complete.
On until a bit below peak temp.
Off for a bit.
Open door.

If you're doing a bunch and want to fine tune: record the temp as you go and watch out for overshoot.

[Pinkus]

Well this looks to me like the use of a cheap/useless or expired paste. Try a fresh paste, which is also known for its good soldering performance in DIY ovens. Avoid the stuff on Ali!

[jmelson]

I have done well over 2000 boards of various types in a GE toaster oven I got from Walmart about 17 years ago.  I use an Omega ramp and soak temperature controller I got on eBay.

I've been curious about using ovens to reflow boards. Long story short, I recently acquired a new Ninja oven from someone who got it as a gift and didn't want it. My intention is to keep it should the need to reflow a board ever arise, but would I need to tap into the control panel for this temperature controller to work?

Yes, I cut a wire and spliced a solid state relay in series with the existing thermostat.  If you don't want to do that, you could just make a cord with a plug, SSR and socket, and then plug the oven into that.
Jon

[jmelson]

When the thermocouple is hanging in the air, nearing a setpoint when the controller pulls the handbrake, the air temperature drops easily and hence not able to reach the setpoint and experiencing an oscillation. Therefore it will try for a long time to reach the setpoint. Hence, your controller is taking 9 minutes and that too is by luck it finishes. Plug the thermocouple tip into the through hole. Also, your paste looks suspicious.

Yes, first try with my toaster oven, I put the thermocouple in the air, and the board got severely toasted!  The tiny thermocouple doesn't absorb IR like the board does.
Jon

[jmelson]

Well this looks to me like the use of a cheap/useless or expired paste. Try a fresh paste, which is also known for its good soldering performance in DIY ovens. Avoid the stuff on Ali!

Yes, good paste is essential.  Did you use a stencil?  Or did you apply blobs of paste by hand?  Looks like maybe too much paste in some places, and water absorbed in the paste splattering and causing components to move.
Jon

[Doctorandus_P]

What brand of solder do you use?
Apparently some of the cheap brands are completely unusable.

The rate at which the temperature rises is also an important factor. If it's too quick, you get uneven temperatures and bad soldering. If it's too slow, then the flux may evaporate before the solder melts, and it does not flow properly.

[Kasper]

What brand of solder do you use?
Apparently some of the cheap brands are completely unusable.

The rate at which the temperature rises is also an important factor. If it's too quick, you get uneven temperatures and bad soldering. If it's too slow, then the flux may evaporate before the solder melts, and it does not flow properly.

Cheap stuff works well for me, even a year or so beyond expiry.  I'm just building little prototypes and testing them briefly, no vibration testing.

Chip Quick: NC191SNL15T5

https://www.digikey.ca/en/products/detail/chip-quik-inc/NC191SNL15T5/11480392

[aaronhance]

So, it looks like I did everything wrong.   

I used "MECHANIC XG-50 Liquid Soldering Solder Paste 42g SN63/Pb37".

[Kasper]

So, it looks like I did everything wrong.   

I used "MECHANIC XG-50 Liquid Soldering Solder Paste 42g SN63/Pb37".

Now that's out of the way, you can do some things right!

[aaronhance]

So, it looks like I did everything wrong.   

I used "MECHANIC XG-50 Liquid Soldering Solder Paste 42g SN63/Pb37".

Now that's out of the way, you can do some things right!

This time it went soo much better. The resistors were so cheap I picked up 100x of them, so now they've arrived I'm using them, that eliminates an unknown. I also attached the thermo couple to the pcb, and turned it off when I saw shiny balls. But I am still dreading the 7 pin IC.

[JURP]

1) Too much paste
2) Bad thermocouple housing. How many has a response time constant? 0.5, 1, 2 minutes? Or even more?

[woody]

Why not use a 'bare' thermocouple? Then you can place the little bead (where the two metals are fused) right on the PCB or even better, in a hole. I used to put a plated-thru hole in the middle of my designs just for this purpose. It might be that a bare thermocouple reacts faster?

Before I changed to vapor phase I used a modified toaster oven for lots of prototypes without any problems. Always using lead-free solder. Biggest problem was that I had to babysit it to open the oven door right after the 10 seconds of reflow; the cooling phase temperatures would otherwise not be near where they should be. Without manual intervention the thing stays way to hot for way too long a time.

[bob8819]

It is better to use an air fryer, its temperature is more uniform

[48X24X48X]

Just left the stencil thickness to be fixed. Just reduce the thickness, then the excess solder paste will not be there.

[Kjelt]

It is better to use an air fryer, its temperature is more uniform

Interesting idea, just thinking is the airflow not way too much, blowing parts away on the pcb?