By Wagner Lipnharski - UST Research Inc. - January 2003
Please DO NOT COPY the text or Pictures below - Link it.


Baking SMT on a cheap kitchen toaster oven is not difficult. You just need some patience and exploration good will.  

Each toaster oven react in a different way, each one, cheap as it is, will produce a different temperature heating and cooling profile curve.  

Soldering SMT components to a PCB requires "some" temperature profile, that is, certain temperature for certain time, then other temperature for other time, and so on.  This is NOT so much critical, as I experienced.  The soldering process happens in one way or another when the everything reaches around 450F,  but critical is the sake of the components on board.  Some components can crack immediately if the temperature goes up or down very fast, others don't show any damage immediately, but they will fail in a month or two (capacitors for example).   As a rule, average temperature profile should be:

  • Insert the board in middle of the oven.
  • Turn on the oven at 160F for 4 minutes.  This will evaporate any humidity on board and avoid little vapor explosions (that can move components on board). This will also preheat all the components, board and the solder paste.
  • Change temperature thermostat to 320F for more 2 minutes.  This will start the solder paste reaction, release the flux and evaporate all the non-necessary chemicals from the paste. At this point you can smell some nasty things, please don't breath that, allow plenty of air filtering or open window from now on.
  • Change temperature to 450F for 60 to 90 seconds.  Almost at the end of this period, the solder paste will melt, will isolate itself around the pads and component leads, all solder paste straight lines over the leads will separate as by magic - this is the flux's job.  If you into the oven that some solder paste still gray and not silver, hold the oven ON for few more seconds, until all the paste is melted.
  • Do not shake or bounce the oven, at this point the solder is melted and components can dance at easy over the board.  Some people use to nod the oven with fingers, just to create enough vibration to self-center components over the pads.  It really works, but some components can really dance rock and roll.
  • Turn off the oven, wait at least 2 minutes, then open the door a bit, just to speed up the cooling down process.  After more 2 minutes you can open the door and wait the boards to cool down to hand touch (it can take 5 minutes or more).

This cool down process is important to avoid damage to components.

We need to call the attention to the fact that a cheap toaster oven, is a little piece of garbage, the thermostat is ridiculous and it is there to control temperature in a 30 minutes average, not a 2 or 4 minutes period.   We need to understand how slow temperature travels from the interior of the oven to the thermostat that normally (on the cheap ovens) are located in the lateral compartment.  When you set the temperature to 160F, the thermostat is cold, so it turns on the thermal elements into the main chamber, and they got really RED.  The inner temperature start to travel by the inner chamber metal plates and propagate to the thermostat at the other compartment.  It can takes almost 45 seconds or more to the temperature reaches the thermostat.  In 45 seconds, a RED thermal element can increase the inner chamber temperature from 100 to 300F.  So, when the thermostat senses 160F, and turn off the thermal heaters, the temperature into the main chamber is about 350F.   In a regular cake baking or sausage roasting it does not change much the results.  At this moment, the thermal elements will stay off during several minutes, until the thermostat travels from 160 to close to 350F then back to lower than 160F when it turn on the elements again.  This process will be repeating and the cake will be ok, but not the circuit board and components.  This is a very wrong profile to bake the SMT board.

This mostly happens for low temperature, as the initial 4 minutes of 160F.  As a bypass to this problem, turn on the oven with the SMT boards inside for 160F, for no more than 90 seconds, then turn it off.  It will avoid the temperature to increase so much into the main chamber.  If you have a way to measure the inner temperature, great, if not, just guess. Keep the boards inside and wait extra 2.5 minutes.  Then turn the knob to 320F for two minutes, turn to 450F (full power in most toasters) for 60 to 90 seconds, turn all off.

Below some pictures of solder paste applied, and some results explained.
Click here for the pictures of the Toaster Oven and attached electronics.

 


FIGURE 1
Here the Kester Solder Paste Syringe 35grams.  Note the plastic needle.  You can use a regular metal injection needle size 18 or 20, serrated and filed to an appropriated size.  The plastic needle appears to be easier and do not scratch or slip over the board.  This one came along with the syringe bought from Digikey back in 1999.


FIGURE 2
Above you can compare the plastic needle size.


FIGURE 3
Here the idea of how much solder paste is applied to individual component pads, most of those are size 1206. At the right you can see some long strip of solder paste applied directly over all the pads of  SOIC ICs (50 mils pitch.   The solder paste reduces volume to less than 1/3 after melting and soldered.  When possible I just hold the syringe pressed a bit, and pull the board to lay down the long strip, sometimes it seems to create a better shape and uniform thickness of solder paste.


FIGURE 4
Here a better view of the continuous strip of solder paste all over the SOIC pads.  When melting, the solder paste will separate itself and accumulate at each pad and component lead, eliminating the bridges between the leads.  If too much solder paste is applied, it WILL NOT move enough, and some bridges will be there. If too few solder paste is applied, one pad can suck the solder from its neighbor pad, and some leads and pads can end up without soldering at all.

Below pictures after the toaster oven baking process.


FIGURE 5
Here a clear example of an bridge caused by excess of solder paste. The solder had no other place to go and ended up connecting two pins.  A simple touch of a clean hot soldering iron tip will clear the bridge. You can use some nice desoldering tool, but most of the time it is not necessary, just the iron tip can do the job.


FIGURE 6
Here two bridges at the same component.  Clearly shows that excess of solder paste was applied here.


FIGURE 7
Here we can see a different problem.  Missing solder paste at this lead.  Is it soldered or not? Nobody knows, when in doubt, just use the soldering iron small tip to get some solder from its neighbor lead (you can see the excess of solder at the corner lead) and move to this one.  This problem is expected.  Note that this lead is connected to the GROUND plane.  What means that?  Means that this pad and its component lead was a little bit cooler than other pins when the solder paste melted, so, the flux just sucked the solder paste from this pad to hot neighbor pads, leaving this lead and pad with very small or nothing solder paste.  This "failure" will always happens here, except if "more" solder paste is applied in this pad.  Probably it could be minimized if the ramp profile from 320F to 450F takes place.  Probably the fast ramp does not heat uniform the board and that particular pad+lead will be few degrees cooler than others.


FIGURE 8
Here you can see the same situation. The bottom chip lacks solder at the corner lead+pad, since probably they were cooler than the inner leads during the melting process.  The chip heats more than the board, the corner leads probably will be cooler and solder paste melting will be stolen by the adjacent leads+pads.

Note the upper chip, the situation is worse, the lead+pad are at a corner, cooler, and worse, they are connected to a ground copper wide area, cooler yet.  But both leads were soldered with zero ohms contact.

It is not totally repeatable, a little bit more of solder paste fixes it.  Check FIGURE 5 above, you see a solder bridge, but not lacking of solder at the corner leads


FIGURE 9
Here you can see two chips with corner leads+pads with reduction of solder.  In both, the soldering was ok, but it could not.  I already experienced some totally isolated lead to pad, requiring manual post soldering.

To avoid those corner problems, I started to make the solder paste strip longer, a little bit beyond the corner lead+pad, as you can see at FIGURE 4 at this page.  It means that when the melting process starts, the adjacent pin will steal some solder paste, but even so, there will be more than enough solder paste to solder the corner lead.  This gave me a little better success, and peace of mind.

 

As you can see, there are more catches in this technique than you can think about.  Only dedication and observation can lead you to a result close to the professional ones.

After the solder process, the board requires an alcohol 91 wash and brush, following by hot water and soap brush to finish cleaning and removal of the extra garbage from the solder paste.  5 minutes into another warm box at 120F will finish dry the board perfectly. The board ends up clean and shinning.

If you have anything to comment or improve this page (mostly the English grammar), please contact me at wagner@ustr.net

The above text and pictures are property of Wagner Lipnharski or UST Research Inc,  they are not free to copy.  If you want, you can link this page to your own pages, but please do not copy them.  I intend to keep updating and improving this page, so copying is not smart.  Linking will keep the readers with the up-to-date information.  This was done with my time and dedication, just to help other people that can be wanting to do home made SMT baking.


Sent by Peka Siiskonen

My experimenting w. SMD:

http://www.taprk.org/img/merg_decod12a_bottom.jpg
http://www.taprk.org/img/merg_decod12a_top1.jpg
http://www.taprk.org/img/merg_decod12a_top2.jpg

Bottom side (first link!) soldered w. DIY shop solder, the top part 1st
phase (top1) with water based solder paste, the 2nd phase (top2) capacitos
w. proper solderpaste, crystal w. with 0.23mm solder and the PIC with the
pre-solder of the board only!

Last edit Feb/03/2003