1
\$\begingroup\$

I made a homemade reflow oven and used it for the first time to solder a board. As you can see in the image below, I botched the job and burned the solder connections (not surprisingly, the board does not work). I'm trying to diagnose what went wrong in order to solder this correctly the next time.

My solder temperature profile is (using Sn63/Pb37 ChipQuik solder paste):

  • Preheat zone (25-100C, 0-75s);
  • Thermal soak zone (100-150C, 75-165s);
  • Reflow zone ramp (150-183C, 165-195s);
  • Reflow zone (183-235C, 195-255s);
  • Cooling zone (235-25C, 255-376s).

However, my oven exceeded the max temperature and got up to 247C before I opened the door.

The obvious first mistake I made is to let the temperature get above 235C. However, I wonder if I might have made other mistakes as well. For instance, it looks like there might be some mild solder-bridging from solder-ball formation (note that the top left two pins should be bridged). Does that mean that my ramp rate is too high, or I didn't hold the temperature at about 150C for long enough? The other mistake I may have made is I only took the solder paste out of the fridge for about 1 hour before soldering.

Do these diagnoses sound accurate? Anything else that comes to mind? I hope the image is sufficiently clear, but if not let me know what is unclear and I will see if I can take a better picture to satisfy those criteria.

enter image description here


UPDATE 1

Here's the same component after cleaning. I'm having some trouble getting more of the flux off than this using a toothbrush and isopropyl alcohol. The I'm also still getting the error "Device not responding to setup address." in the Linux kernel ring buffer.

enter image description here


UPDATE 2

I resoldered the USB micro receptacle and the board now works. None of the other components (including the one shown) required resoldering.

\$\endgroup\$
  • 3
    \$\begingroup\$ A solder bridge is a short circuit. Not sure how it can be 'minor'. Perhaps you can check whether any pins are actually shorted before assuming that you have cooked components. It's hard to see if you have any bridges with the flux there. Is is worth cleaning it? I also noticed there is no solder mask between pins. \$\endgroup\$ – Steve Hubbard Nov 2 '18 at 5:29
  • 3
    \$\begingroup\$ Not perfect, but looks okay to me... What kind of oven did you make, and how does it control the temperature? 235 vs 247 is on the margin of error, it can't be the main problem. No solder mask looks like the major issue here. \$\endgroup\$ – Ale..chenski Nov 2 '18 at 5:53
  • 1
    \$\begingroup\$ I also built a reflow oven once and encountered similar issues. How do you messure the temperature? I first used a too massive thermocouple, which delayed me the temperature and so the actual temperature was maybe 50 degrees too high \$\endgroup\$ – HansPeterLoft Nov 2 '18 at 7:34
  • 2
    \$\begingroup\$ The solder connections don't look "burned" to me, but there is a beefy amount of flux on them. \$\endgroup\$ – Scott Seidman Nov 2 '18 at 12:22
  • 1
    \$\begingroup\$ Try an old toothbrush with the alcohol to get that flux off. No clean is notorious for causing problems with analog circuits but your micro may not care about a bit of leakage. \$\endgroup\$ – Spehro Pefhany Nov 2 '18 at 13:26
1
\$\begingroup\$

Temps are way too high and times are too long.

I manually reflow also using Kester EP256, https://www.cmlsupply.com/kester-ep256-lead-solder-paste-63-37-150g-jar/

I have a multimeter with thermocouple probe, Extech EX330. http://www.testequipmentdepot.com/extech/multimeters/compact-digital-multimeters/autoranging-digital-multimeter-600v-10a-non-contact-voltage-detector-temp-function-ex330.htm?ref=gbase&gclid=EAIaIQobChMI6-uRsOq13gIVlUoNCh0r_wsqEAkYAyABEgJG1_D_BwE Purchased at my 'local' electronics supply store for around $40. I've gotten a lot of use out of that probe. I place the probe to sit just above the boards, the door closing holds it in place. The toaster oven is an ancient Sears toaster oven with 4 heating elements, and a tray to place the boards on.

Start the heat: turn temp to high, back it off when it hits 145C, the oven will continue heating a little, let it settle around 150C. Let it sit there 90 seconds.

Start the heat again: turn temp to high, back it off when it hits 185C, work the heat on & off, keeping it above 185C and below 205C, for 90 seconds. 205C is important, quite a few parts show 205C as their never exceed temperature.

Turn the heat off, open the door a little to let the heat start escaping slowly. When the temp drops below 180C, smoothly open the door fully. Let the tray cool some more, then remove the tray and let it cool fully. I let mine sit on a stove burner to finish cooling, they can take high heat.

\$\endgroup\$
  • \$\begingroup\$ I will try that out next time. Any idea why the profile on the solder paste datasheet used higher temps/longer times? \$\endgroup\$ – MattHusz Nov 2 '18 at 15:11
  • \$\begingroup\$ I think you misread the times. 30 sec to 120 sec = 90 sec, same for the time at high temp. 235C seems high for melting, when parts data sheets are showing 205C as never exceed temperature. \$\endgroup\$ – CrossRoads Nov 2 '18 at 18:41
1
\$\begingroup\$

You need to check you EDA package or library part for the chip. The solder mask should go around each pin on the chip entirely. There will be settings that pull the solder mask back a certain amount. This may be a global setting that affects every component on the board or it may be the specific library part for the chip.

The board looks like it's from OSHpark and they are capable of doing so with no problem. Note that in the following image the solder mask is right up to the pads. The solder mask helps to reduce or eliminate solder bridges between the pins.

enter image description here

\$\endgroup\$
  • \$\begingroup\$ So you suggest having a solder mask clearance of 0? Isn't there a risk of the pads and solder mask not lining up exactly? For instance, on the OSHPark site they seem to recommend a solder mask clearance of 3mil. \$\endgroup\$ – MattHusz Nov 2 '18 at 16:15
  • 1
    \$\begingroup\$ @MattHusz Read the following article from OSHPark docs.oshpark.com/tips+tricks/stop-mask-expansion \$\endgroup\$ – vini_i Nov 2 '18 at 16:55

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.