[phpBB Debug] PHP Warning: in file /mnt/forum-shared/ext/sitesplat/BBCore/event/listener.php on line 70: in_array() expects parameter 2 to be array, boolean given
Reflow Oven - Tune - SparkFun Electronics

SparkFun Forums 

Where electronics enthusiasts find answers.

General project discussion / help
Did you make a robotic coffee pot which implements HTCPCP and decafs unauthorized users? Show it off here!
By tlogic
#180356
Hello all,

I'm new in SMD technology and i would like to share with you my issue.

I bought an oven toaster in order to convert it to a SMD Reflow Oven having in mind to use an arduino board with a K-type thermocouple as the PID controller.

As a first step i modified the oven adding an SSR to control the heating elements. I connected a K-type thermocouple to an MAX31855 in order to read the temperature and i added some extra gold thermal tape in the inside of the oven to increase the heat reflection. The code i used is the rocketscream reflow example modified according to my needs.

Configuration values of the code:
Code: Select all
// ***** CONSTANTS *****
#define TEMPERATURE_ROOM 50
#define TEMPERATURE_SOAK_MIN 150
#define TEMPERATURE_SOAK_MAX 200
#define TEMPERATURE_REFLOW_MAX 245
#define TEMPERATURE_COOL_MIN 100
#define SENSOR_SAMPLING_TIME 1000
#define SOAK_TEMPERATURE_STEP 5
#define SOAK_MICRO_PERIOD 9000
#define DEBOUNCE_PERIOD_MIN 50

// ***** PID PARAMETERS *****
// ***** PRE-HEAT STAGE *****
#define PID_KP_PREHEAT 100
#define PID_KI_PREHEAT 0.5
#define PID_KD_PREHEAT 20
// ***** SOAKING STAGE *****
#define PID_KP_SOAK 300
#define PID_KI_SOAK 0.05
#define PID_KD_SOAK 250
// ***** REFLOW STAGE *****
#define PID_KP_REFLOW 300
#define PID_KI_REFLOW 0.5//0.05
#define PID_KD_REFLOW 350

#define PID_SAMPLE_TIME 1000
Here is a photo of my oven.
2.jpg
The solder paste i use is sn62pb36ag2

This is the thermal curve output of the oven
4.jpg
and this is the result of the reflow
3.jpg
As you can see the reflow result isn't as it was expected to be.

1) Is this a problem of the reflow curve?
2) How can i tune the PID in order to achieve the ideal curve?
3) Shouldn't this curve reflow the solder paste?
You do not have the required permissions to view the files attached to this post.
By waltr
#180403
250°C (I assume its C) is NOT hot enough to melt the solder.

Also, the temperature peaks then drops. A "soak" time is required.
Look up what a proper temperature profile looks like. Many manufactures have re-flow profiles in the data sheets.
By marcovansteen
#180703
Hey, nice project. I'm about to try something similar soon.
The profile looks OK to me. The measured curve shows no overshoot or instability, so the PID_K* values should be usable.
The cool down rate is a bit slow. As soon as the solder is solid, you can open the oven's door, to cool down a faster. This may reduce oxidation. But I don't think that that's the real issue.

@Waltr: this is lowtemp solder paste, melting temp is around 180°C, see the datasheet.

I'ld suspect:
- the solder paste did not melt (it should be shiny after soldering, shouldn't it?). Heat with a solder iron and check if it starts to flow. If it flows, the actual pcb temp was much lower than your thermocouple indicates. Make sure the thermocouple is thermally coupled to the pcb; you may also want to check and calibrate the temp measurement (with ice and boiling water?)
- something wrong with the paste: too much paste, paste not well mixed (were these the first drops from a tube?), paste too old, paste wasn't stored in the fridge, etc.
By tlogic
#180768
Hello marcovansteen,

Thank you for your reply.

- No it is not shiny after the reflow cycle. It looks as it is dry with mat colouring. If you scratch it with a screw drive it is reveals an under layer of melted shiny solderpaste. The thermocouple is attached at the top of the pcb and it touches the surface of it. I haven't calibrate the thermocouple but in comparison with a multimeter i have the temperatures readings looks to be similar.
- The paste is a fresh new http://termopasty.pl/produkty/easy-prin ... b36ag2-en/
The amount of the paste could be too much at the first samples but then i tried using the stencil so that looks not to be the issue as the problem still appears. I also tried different paste samples so there is no mixture issue. Finally i tried to add some more flux in the paste and mixed it well with the result of a small difference in the melting point.

Here is a sample of the latest curves i have used.
Screen Shot 2015-03-29 at 15.49.37.png
With visual inspection, it looks the solderpaste starts melting at 180C and then for some reason to stop. I have configured the oven to stay at 200 for more seconds in order to see if this is the issue but no result.
You do not have the required permissions to view the files attached to this post.
By marcovansteen
#180798
Good checks. So the paste did melt and the paste should be OK. With the stencil, did the solder still flow outside the pads? Was the connectivity OK?
The mat colouring could be due to the flux residues. It is a no-clean flux. Maybe it needs more heat to really disappear? What happens if you heat it with a hot soldering iron (at 350 °C or so)? Maybe the people at termopasty.pl can help you out on that.