Calibration of electronic measurement instruments is a necessary process, even though most electronic equipment is very stable and somewhat “resistant” to the effects of environment and changes due to aging.

Q: So why calibrate if my MOLE is “in spec” every time I send it in for calibration?

Because calibration is not so much an adjustment process but rather a proofing process that shows, over time, that your MOLE has been in calibration and thus should remain in calibration, because you have a track record to prove it. Documented history of a MOLE’s performance is the only way to claim your MOLE is in calibration at any given instant.

Most good labs will tell you that when your MOLE is calibrated, it is compared to standards , typically standards that have traceability to NIST, and if it is shown to be measuring within its specified accuracy they will not make any attempt to “adjust” it. Only if it is “on the edge,” which usually means it is getting to the last 10% to 20% of the specified accuracy limit, will they make any adjustments. Your MOLE may still be “in spec” and thus “in calibration,” when the lab received it, but getting close, so they will adjust it closer to the middle of the spec. range.

If it is out-of-spec when received by the lab, then a red flag goes up and calls into question every measurement made since the last calibration! The lab will tell you how far out of spec it is, and you can decide if its measurements during that time affect the quality of the measurements made more than can be tolerated, or if they are “close enough” to still be acceptable.

Q: So, when should the MOLE be calibrated?

The number one best time to calibrate the MOLE is on a regular time-based interval, which is recommended once a year. However, there are other events which may cause you to want to seek calibration at other times of the year, such as:

1. When the MOLE is subjected to rough treatment like a fall to the floor,
2. When your MOLE is accidently “over heated” ,
3. When you are starting a new product introduction and you are characterizing an oven and new assembly to find the right recipe,
4. When a new customer’s contract stipulates you use equipment that has been recently calibrated,
5. When your in-house quality program requires a calibration interval.

Getting your MOLE calibrated is easy and we want to make sure you are always making the highest quality temperature measurements.

1. “I’m from the Metrology lab and it’s time for the annual calibration of your reflow oven.” We’ve got an app for that.  After you’ve finished with the oven’s calibration procedure, you can run the MegaRIDER-20 with a Process Test Pallet to see if the machine is uniform across the conveyor width and has the same heating capacity as it did the last calibration or maintenance.

2. “I’m the Manufacturing Engineer and our QC Department wants me to show that this oven is in control.” We’ve got an app for that. You probably need more information than the once a year Metrology profile can provide. So weekly you can an OvenRIDER and see that every zone in the oven is performing the same using X-Bar R charts to prove it.

3. “I’m on the New Product introduction team and I need a good recipe to solder a new board without killing the parts.” We’ve got an app for that. The Super M.O.L.E.® Gold thermal profiler will let you connect T/Cs to the board to see exactly what’s going on, thermally, on the areas where you and the designer have the most concern. Use the Prediction tools in the new MAP software to lock in the perfect recipe.

4. “I’m a Line Technician and I have to know my reflow oven is ready to run product without all the wires and circuit board stuff.” We’ve got an app for that. OvenCHECKER is one pallet loaded with the most powerful profiler on the market today. It takes no more time to run than the first production board and it lets you know if the reflow oven is ready or not. No downloading, no comparing numbers on a chart, just Go, or No-Go.

The ECD V-MOLE with patented one button “OK” profile evaluation

(The OK Button is also available on the 20-channel MEGAM.O.L.E.™ thermal profiler, and OvenCHECKER™ )

Taking only seconds, the MOLE can tell you and your oven operators if the profile just measured is in or out of specification with the universally understood Green for good (GO!) or Red for bad (STOP!).

You get to set the specification limits for any or all of the four most popular profile parameters:

• Ramp Slope
  
• Time Between temperatures
  
• Time above Liquidous
  
• Peak Temperature
  

…and you can choose which of the MOLE’s input channels to include, up to 20 channels on the MEGAM.O.L.E.™, and three on the V-M.O.L.E.™ thermal profilers.

The Specification Table in MAP Software

Using the MAP™ Profiling software, enter your specification limits for the four profile parameters in the Upper and Lower Limits table in the “Target-10 OK” tab. These values will automatically be sent to the MOLE profiler when you use the Verify Process Wizard to confirm that a previously characterized oven recipe is still performing within specification.

MAP™ Profiling Software Target-10 OK Profile tab

Once programmed, the MOLE profiler can be used many times (up to 96 times) to Verify your oven is producing the same profile, without reconnecting to your computer. Simply run the profile and press the “OK Button” on the MOLE. No more running back to the PC software to download to see the results. One push of the OK button, and you get your answer…Go, or No-Go. It’s that simple!

*U.S. Patent Number 7653502.

You might ask, why should I perform an Oven HealthCHECK? In other words, why should I run a rather sophisticated measurement system through my oven to produce a rather nice looking 3-D plot of the cross belt temperature uniformity?

Figure 1: An example 3-D plot of the oven with very good cross belt uniformity, < 3ºC

I may not like the answer because I might find out my oven has a problem that I cannot fix? Sounds like a “head in the sand” sort of excuse to me.

I would ask a different question: “Why do you calibrate your bench test instruments?” Is it to find out that the instrument is out of spec? Heavens no!! That would be a nightmare because it would call into question everything that instrument was used to test since the last time it was calibrated. So why do you calibrate if the results could be so disastrous? Simple, it allows you to show and document that the instrument is and always has been “within specification.” So when the auditor asks, how do know your instruments are in calibration, you pull out the Certificates of Calibration.

Figure: What’s the first thing you check after you receive your Certificates of Calibration? The “As Received: Within Tolerance”

ECD’s Oven HealthCHECK is designed to certify and document your oven’s performance or health. It is a “calibration” done on a regular interval, say once a year, to show that the oven performance is within specification and not changing over time. It can also provide a baseline level of performance around which you can compare into the future. Further, if you have several ovens, the HealthCHECK can show you which ovens are best for applications where oven uniformity is critical.

Back this up with much simpler and more frequent verification profiles of your oven using OvenRIDER or OvenCHECKER, where simple software-generated Xbar-R control charts show daily indications of a thermal process that is “in control,” and you will no longer have to steer the auditor around your reflow oven. You can proudly show that you know your oven’s performance level and that it is consistent because you have taken the steps to measure your oven’s health as part of your Thermal Quality Management program. Such a program should be marketed, since it shows you commitment to understanding your oven’s thermal nature and you have the data to show it. This sure beats the “head in the sand” quality program which may characterize your competition.

1. (Max preheat) How hot the board is just before it hits the wave
2. (Dwell time) how long do you spend in the wave
3. (Contact temp) Temperature of the solder at the contact surface with the board
4. Conveyor speed

All the rest of the many measurable parameters are secondary to these in my opinion. Let’s talk a little about each of these as measured by the WaveRIDER SPC software:

Figure 1: WaveRIDER SPC Software

Max Preheat temp (Preheat Max Temp) is really the temperature of the board (solder joints) just before it hits the wave so you can understand the “shock” the solder joint will get when contact with the wave forces it to melt the solder in a quick few seconds. This temp may NOT be the max preheat temp applied because most wave machines have a large gap between the pre-heater and the solder wave so the board will loose some of the heat during that space/time.

The Dwell time is how long the solder will be above liquidous, all said and done, and this time needs to be consistent across the width of the wave (AKA: parallelism), but you can see consistency across the wave just as easy by setting a min/max spec for the three dwell times. How long you must dwell is a function your solder joint requirements and components ability to withstand the wave temp. Most component specs say they can take 10 seconds or less and most solder joints need only a second to form. The longer you spend above liquidous the more time you give to form the brittle “intermetallic” alloy between the not molten metals. So the only reason you dwell at all is to get the heat to travel up the lead to the board top (on through-hole parts), which may take a few seconds depending on the thermal conductivity of the board and its thickness. A pot of liquid solder has plenty of heat to force into a board, so it depends mostly on the thermal conductivity of the board material itself. You will have to watch/trial your assembly (board) to see just how long that takes so you can set a minimum dwell time. The Dwell time is usually about 2.5 seconds for most 0.062″ (1.6mm) thick boards and this can vary +/- 10% to 15%.

Figure 2: Wave crest pattern

Pot temp at the solder contact surface (Contact temp) does vary depending on the load (your board) places on the surface of the solder wave. It is not the same as the solder pot temp, since that temp is controlled by a Thermocouple deep in the core of the solder pot and not the surface where the board skims across. This temperature should be maintained at least 20ºC or more above liquidous and no more then the maximum temperature allowed by the component specifications.

The conveyor speed is measured by dividing 9.75″ (distance between C and Speed sensors on the WaveRIDER pallet, see Figure 4) by the time measured between the C and Speed sensors hitting the main wave. It is easy to see why this is an important measure since ALL other values will be altered if the conveyor speed changes. Most solder machines can maintain speeds within +/-0.1 ft/minute (+/-1.2 in/minute, +/-3cm/minute)

Immersion depth is only important to prevent solder from spilling over the top of the board during wave contact. The absolute setting is not critical provided it is NOT spilling over the board and its depth is affording the dwell time one needs. However, it is easier to adjust the conveyor speed to control dwell time then the board’s immersion depth and risk a spill over. This is usually controlled by pump RPMs and how full the pot is of solder. If these change over time, the dwell times slowly change and thus is be one of the reasons for dwell changes. However, neither can be directly measured by WaveRIDER which has no chance to access the pump’s RPMs or the amount of solder in the pot.

Conclusion:
Focus your efforts on the critical WaverRIDER measurements by setting reasonable specifications. The many other things the WaveRIDER measures are not as critical, and can be used to diagnose a problem if there is one rather then used to prove there is NO problem. Use the WaveRIDER SPC software to set Spec limits and plot X-Bar R charts for these critical parameters. (Figure 3)

Figure 3: X-Bar R charts

Figure 4: Keep the WaveRIDER pallet contacts clean using a fine wire brush. This will reduce oxide build up on the sensors which can cause the Dwell times vary beyond normal.

Reach For the Sun

Solar Cell manufacturing has been around for a long time; however the materials and process have changed drastically in the past few years and will continue to evolve as the technology and need for Renewable Energy grows. The costs of manufacturing and the risks associated with the ever changing processes can strain  the ability to maintain yields and improve quality. Proving these new manufacturing processes, then achieving the repeatability and yield needed for production have always been a challenge.

ECD’s suite of profiling tools allow detailed Characterization of these advanced processes  during the R&D phase.  During production ramp-up good thermal thermal profile data drives yield improvement leading to maximum profitability.   In  production these same tools can quickly Verify that the optimum manufacturing  process is being maintained.

Thin Film Solar

ECD has many types of customers in the Solar Cell manufacturing industry.  They include start-ups, university research departments and production facilities, located around the globe. 

The application areas that we have been able to identify are among the following:

Silicon Metallization:
   Thermal profiling is used to optimize the drying, rapid firing and following cooling process in the oven.

Silicon Diffusion:
   Thermal profiling is used to optimize the heating, high temperature diffusion and cooling process in the oven.

Thin film Solar Cells:
Done on glass and other substrates, this process is similar to Diffusion process, but at lower temperature.

Profiling Equipment Requirements:
Minimum profiler thickness is important – Many of the ovens designed for solar manufacturing provide little vertical clearance for the profilers.   Thermal barrier requirements vary and in some cases, time and temperature do not allow pass-thru profiling.  In most cases 3 channels of thermal data are sufficient for process verification as all areas of the silicon heat similarly.   Thermocouple attachment can also be tedious, with mechanical pressure being the most common contact method.

We look forward to the continued growth and success of the Solar industry and we would like to invite all solar industry participants to work with us and discover how we can help you reach your performance goals.

A good Thermal Quality Program (ThQM) demands consistent oven verification to show that the reflow oven is reproducing the same thermal environment as it has in the past. The OvenRIDER is a good way to verify oven performance. How do you make best use of these tools, both the OvenRIDER pallet and the OvenRIDER SPC (ORSPC) software?

There are five basic steps:

• Set up the Oven, Workbook and MOLE
  
• Model the oven
  
• Collect a base set of OvenRIDER runs
  
• Set the spec limits
  
• Monitor of verification

Let’s get right into it:

Set up the Oven, Workbook and MOLE

1. Place a magnet at the beginning of zone one. The OvenRIDER came with at least three magnets. This is not because we think you only have three zones, but because we figure you may have more then one oven, and you only need one magnet per oven to mark the start of zone one and let the OvenRIDER measure conveyor speed for you. The software detects this magnet in the profile and uses it automatically place the start of the oven model on the profile’s time axis. It does not matter where you put this magnet as long as it is on or before the beginning of zone one. You can even put it outside of the oven so it can be removed when you are not using OvenRIDER. Just be sure you always put it back in the exact same spot.
2. Set the oven’s recipe you plan to use for this OvenRIDER run. It can be the same recipe you are now using to solder boards. In fact, you may want to start an OvenRIDER collection of runs for all your different recipes. This does need to be a big one time effort for all your recipes. You can do this slowly for each recipe one at a time as you do line change over. In time you will have enough runs for each recipe to begin useful oven Verification.



3. Start the OvenRIDER software and select the Workbook you wish to save these recipe runs. If this is the first time you opened the software, a sample work book will be opened. Go to the file menu and close this Workbook. Then open the file menu and start a New Workbook or open an existing Workbook you previously created. The Workbooks can help you group like OvenRIDER runs. Having a work book for each oven recipe is the best for Oven Verification. You can name Workbook the same as the recipe name to help keep it straight.�
    
4. To make sure the MOLE is ready, connect it to the PC via the communication cable and run the Configuration Wizard in the OvenRIDER menu. This will find the MOLE and set it’s clock and other settings.

 



 

5. Load the MOLE into the OvenRIDER barrier and connect the configuration plug

 



 

6. Make sure the oven is up to temperature and ready to receive product. Also make sure the conveyor width is set the OvenRIDER’s width.
7. Start the MOLE, close and lock the lid, and load the OvenRIDER pallet onto the oven conveyor.
8. When it gets through the oven, open the barrier and stop the MOLE by pressing and HOLDING the button until the Status LED turns off by itself.

Model the oven

Oven modeling is critical. This tells the software how big each oven zone is and where each zone is on the thermal profile so it can correctly show which zone had what effect on that portion of the profile.

Above is a typical Oven RIDER thermal profile with an oven model (vertical dashed lines) in place. Each zone influenced the “shape” of the thermal profile of the OvenRIDER’s sensors as it passed through oven. The amount of influence depends on the oven recipe temperature set point for each zone, the conveyor speed and the convection rate (fan or air speed). The temperature set points and conveyor speed has a straight forward and expected effect. The convection rate is a little different. Some ovens allow this to be changed and may be measured in several ways like: velocity, pressure, percent, Hz or RPMs. In some ovens, the convection rate is not adjustable or has two or three settings like: Low, Medium, and High. In either case, convection rate changes are one of the leading causes of profile changes when neither the conveyor speed nor zone temperature set points have been altered. So understanding what part of the profile each zone influenced is critical to pin pointing where problems in the oven have occurred.

Here is the best way to model an oven, any oven:

1. After the profile run, connect the MOLE to the PC via the communication cable and press the Read OvenRIDER button.

 

2. The software will as you how much data you want to use from the previous run, if there is a previous run. This time, the first time don’t select anything by un-checking all three.



3. The OvenRIDER profile will look something like the profile below. Because you installed a magnet, the start of zone 1 is in place on the profile, and that’s it. It is ready to Model the oven.

 

4. Start the Manual Zone Definition by selecting it on the Manual Zone menu.
5. Select the number of zones check boxes on the let that your oven has. Name them if you like. Leave the values it defaults for Zone Lengths and Units. We will take care of that next. Note it has the measured conveyor speed already calculated and entered in as the conveyor speed value.



6. Enter the Zone Temperature Set Points in the Top column. Note the Bottom values copy from the Top values as you enter them since most oven use the same Top and Bottom values. Enter the bottom values, if different, after you enter the Top values. When done, click OK.
7. You will now see all of your zone boundaries located in the default locations, but not in the right locations. Note the small boxes at the top of each boundary. This is a handle for you to grab and move the boundaries to the right locations. The right locations are in that little dip that happens between zones you can see in the Ambient sensors, the Red, Blue, and Green profile lines (Channels 1, 2, and 3). So start moving the zone boundaries into place. Remember, don’t move the start of the first zone, this was set for you by the software when it detected of the magnet.



8. Your profile and oven model should look like mine below. This is the “thermal” Model of your oven. It’s already saved with this profile, but let’s save it for future use with other recipes used in this same oven.



9. Go to the Manual Zones menu and select Setup Zones.



10. This will re-open the Oven Zone Setup dialog box. Hit the save button and name your oven model. I recommend using the Oven or line name. This model can be used for other OvenRIDER profile runs using the same model of oven, but with different oven recipes later on. Use the Load button to recall this oven model on future runs.

Collect a base set of OvenRIDER runs

You now have a good oven model and your first run at this recipe. You need to collect at least three runs at this same recipe before enough data is collected to meaningful, statistically. This can be all at once, but this take a big chunk of time from production, and most don’t have that kind of time. I recommend you take an OvenRIDER run at the beginning of each day or shift, while you are running the same recipe, for several days. You will have 3 or more runs in less then a week.

If you don’t run the same recipe for more then a day, you can set an “Oven RIDER Recipe” near the recipe setting you use most often. Start the day with that recipe, take the run, and then move to the “real” recipe you plan to use that day.

If you want to do three or more Oven RIDER runs all at once, just be sure to cool the MOLE and the Pallet to room temperature between runs. Do NOT short cut this critical cooling process. This will save the MOLE from possible over heating and assure the pallet temperatures are consistent run to run. 20 minutes on a nice fan will be enough in most situations.

Once you have several runs, you Spreadsheet will look something like this:

Make use of the user columns, the Green ones. This is where your can name the column anything you like. At least make a column for Recipe, and maybe the Machine or line name. I also have a Part Number column to identify the part or assembly soldered using this recipe.

Set the spec limits

With a good set of OvenRIDER runs collected under the same oven recipe, at least three, but five or even 10 is better, you can now set limits on the average of these runs. Do the following:

1. Select the Spreadsheet tab and filter out or “hide” the runs that are NOT part of the base OvenRIDER runs. If you named each base run like I did in the example, this is easy. Select the filter dropdown arrow in row 4 of the column where you named the base runs, n my case column E, and select the name that you used to name the base runs. I used “Base Runs,” for simplicity. This hides (temporally) all the other runs. If you don’t have any other runs yet, then you don’t need to do this. The statistical data at the bottom of each column is now calculated only from those runs still visible.



2. Now select the Admin tab and click the drop down arrows in the LSL and USL columns. Select the -10% and + 10%. This will set as upper and lower spec limits +/- 10% of the Average values for your base line runs showing on the Spreadsheet tab for ALL the values OvenRIDER measures. These will be good starting values to see how your oven is doing.





3. Optional – You may want to select +/- 5% later on after you have a few more runs, if you wish to run tighter specs. Or, you may want to change the conveyor speed spec limits to a little tighter because you know you machine can do better then 5% or 10%. Finally you may want to remove some of the specs because you are not concerned with those measured values. To start with, it will not hurt to keep all the values.

Monitor of verification

Now each time you run the OvenRIDER at this recipe, you can see in a flash if the ALL of the measured values are within the specs you set. Simply select the OvenRIDERData tab and give the values a look.

If any of them are Red (above spec) or Blue (below spec), then there may be a problem in this portion of the oven. The values of most concern are the Average Temperatures and Process Delta per zone. If these begin to fall out of spec, then you know there may be a problem in that zones ability to heat.

All black, as in this run, the oven is ready to roll.

Any parameter out of spec should be examined to see if it’s a concern or not.

If you wish to get deeper in the SPC control you can create. Xbar-R Charts for any of these values as well. Simply drag and drop any measured value in the column to the left into one of the SPCA, B, C… boxes. Check the box to add that SPC group of charts to the tabs pages below.

Select the tab to see the X-bar R charts. Here are many production runs of the oven ruder captured of time.
These chart use all the typical SPC rules to determine process control.