Printed Circuit Board Assembly & PCB Design SMT Electronics Assembly Manufacturing Forum

Printed Circuit Board Assembly & PCB Design Forum

SMT electronics assembly manufacturing forum.


Internal Oven Calibration

Views: 3733

I am putting together an internal oven calibration procedure... - Feb 10, 2006 by RFredericks  

#39612

Internal Oven Calibration | 10 February, 2006

I am putting together an internal oven calibration procedure. I am interested in how stringent your oven calibration process is. Really what I am putting out is a verification process and a calibration process for the event that verification should fail. We lie in the hi-rel market, but are not attempting to match mil-specs. We, for now, just need to make sure that our temperature controllers are on point. Key things I am interested in are: 1. No. of testing points within a chamber 2. What procedure differences can I consider when calibrating set temp. ovens, adhesive curing ovens, and ovens used for burn in and temperature cycling. 3. What frequency of verification is used? I appreciate any advice that you can give. Thanks! -Russ

reply »

#39613

Internal Oven Calibration | 10 February, 2006

Russ, We use a unit called a Datapaq surveyor to monitor our ovens. Its basically a unit that a data logger rides through the oven in. It has 3 thermal couples across the top and 3 across the bottom. It takes measurements the full width of the process chamber. This unit also verifies that your belt speed has not changed.

We monitor ramp, soak, and reflow.I won't go into a complete explanation, but the software basically gives you a green light, yellow light, or a red light depending on how you set your parameters around your "golden run".

We make 1 run through each oven weekly.

The only reason I am talking about Datapaq is because it is the only profiling equipment we have used in mainline production at this point.

I have also heard that KIC works well too.

reply »

#39615

Internal Oven Calibration | 10 February, 2006

We do the same as SR Tech but add in a check of the thermocouples the oven uses to monitor it's own temp.

Annually I attach a thermocouple from a calibrated thermometer to the thermocouple in the oven. I compare the readings. We specify a plus minus 3 degrees C tolerance. If this is out we calibrate the ovens temp reading.

Jerry

reply »


RDR

#39617

Internal Oven Calibration | 10 February, 2006

We do something quite simple that may or may not be adequate but I cannot think of why it would not be adequate. If anybody can please inform.

We have a large board about the samer width as our oven. We have dedicated thermocouples on this board along with a dedicated profile. We ran the board and recorded profile. now we just run this board under that profile and overlay the results. To date I haven't had a calibration failure only blowers or heating elements.

reply »

Chunks

#39619

Internal Oven Calibration | 10 February, 2006

Establish a profile board. It should be representative of what you normally run on a daily basis, size and parts wise. Attach dedicated thermalcouples so they are permanent and will not move or come loose. High temp solder is best. Establish a profile representative of your current profile.

Assure your oven is running correctly and readings are correct. Parallel porting off of each thermalcouple works best. Parts from Omega work well, and they even make face plates for these things to mount too. Gives your oven a very clean look. http://www.omega.com/ppt/pptsc.asp?ref=ICIN&Nav=tema04 Use a calibrated therocouple device that can be calibrated annually.

Run profile board monthly and record profile for each (electronic and hard copy is best). Run this board thru your ovens monthly and keep data for each oven seperate.

If there is any descrpancy with your monthly profiles, use the parallel ports off the thermal couples and your calibrated thermcouple device. They should read very close to each other (maybe 2 degrees F). If not, you know you have something wrong with the heating elements.

An over looked area is your conveyor speed. For this you need a calibrated ruler (yeah I know, but it has to be calibrated) and a calibrated stop watch. When you run your profile board, your oven profile should always have the same conveyor speed. Figure out how long a link in the chain will take (in seconds) to travel say, 12 inches. Now use your calibrated ruler and stop watch to make sure it's close.

Sound like a lot of work, but every auditor that's ever seen my ovens set-up this way just drool all over it. Plus it gives you the confidence to know your stuff is calibrated to the best of your ability.

reply »

KEN

#39626

Internal Oven Calibration | 11 February, 2006

One factor everyone overlooks is that furnace and external T/C wires are (generally) not calibrated.

Unless you are sending your wires to a lab for NIST characterization your T/C wires are NOT calibrated. They (new wires)operate within a specified tolerance band and are often assumed to be "absolute". They are not absolute, nor is thermal profiling an oven via standard methods described here. What you are doing is characterizing a system with an uncertainty factor. Without defining this uncertainty factor you do not know where you truly are on the abloslute temperature spectrum.

reply »

#39630

Internal Oven Calibration | 13 February, 2006

Hi,

Does it really matter what temp the oven runs at as long as the PCB's are profiled correctly? I mean we run boards through all the time to check profiles, so as long as that's ok, would it matter what the oven though the temps where?

Regards,

Grant

reply »

Chunks

#39632

Internal Oven Calibration | 13 February, 2006

That's why I parallel ported the T/C's. Now I can use a calibrated hand help thermocouple and plug into the T/Cs on the oven. If they're within about 3 degrees of each other, then they're considered within our specification. Paper work on the hand help thermocouple can be trced to an outside lab we use to calibrate such devices.

reply »

Chunks

#39633

Internal Oven Calibration | 13 February, 2006

Hi Grant,

Most inspectors want your oven temps close to what you normally run your product. I've had to show both glue and past profiles on every oven before. Then they want to see if you are matching the profile specified by the manufacturer. If you're not, you got some explaining to do.

reply »

#39637

Internal Oven Calibration | 13 February, 2006

I believe I understand what Ken is talking about. Most "calibrating" is done but applying a known voltage to the thermocouple reader. But what if the thermocouple wires are not producing the expected voltage?

I bet your paperwork is for just the reader not the system. To do what I would consider a genuine calibration you would measure a known temperature with the system, not measure voltage on part of the system.

Now having said that I would show any auditer the paperwork without explaining that it is not a direct measurement. I may be crazy but I'm not stupid.

Ken; If I misunderstood you, please let me know.

reply »

#39640

Internal Oven Calibration | 13 February, 2006

Thanks everyone, I have found your postings very helpful for reflow ovens and intend to review my reflow calibration and validation, but in this particular instance I should have been more specific. We use the non conveyorized type of oven (Tenney, Blue M, the entire Lunaire spectrum) for burn in and adhesive curing among other testing and assembly ovens. I was curious about calibration/validation of these types of ovens. Thanks again, -Russ

reply »

#39641

Internal Oven Calibration | 13 February, 2006

I have been through this before. At a med. device company they required every device that performed a measurement to be calibrated. O-scopes, calipers, 6-inch steel rules, etc. And thermocouples.

While I can't see what difference it makes to the product as long as you have adequate profiling procedures and you're using a calibrated profiler. The thing is, if an auditor comes by and asks you why you're running your reflow zone at 300 top and 245 bottom to get a 218� peak and you tell him/her it's because the oven thermocouples aren't very accurate, you may have some 'splainin' to do.

We calibrated ours once per year. Had a guy come in house to do it, and it was expensive. If you don't have nosey clueless auditors that are just looking for something to write up, your product won't suffer.

reply »

Chunks

#39650

Internal Oven Calibration | 13 February, 2006

Same way. Either plug into existing T/C or use a hand held that's been calibrated. We have the same Blue M set up too.

reply »

#39652

Internal Oven Calibration | 13 February, 2006

ECD's OvenRIDER is the only instrument I know that can measure thermal transfer efficiency of a an oven. I suppose it would work for environmental chambers too. This instrument has thermocouples to measure ambient temp of oven zones and well as hi-mass sensors to measure thermal transfer efficiency. This measures temp ramp up and cooling ramp down.

They have demo sftware you can downlonad and run.

http://www.ecd.com/emfg/instruments/ovenrider/

reply »

#39655

Internal Oven Calibration | 13 February, 2006

Ah yes, environmental chamber. Jargon updated, thanks!

reply »

KEN

#39677

Internal Oven Calibration | 13 February, 2006

You are correct. Regardless of what any auditor is asking what we're dealing with here is sound science and good Process Engineering. Standard methods for temp profiling are not absolute.

If you are using a Temp simulator you are only verifying the accuracy of the A-D converter of the measurement system. Everone forgets the TC wires in the funace and the wires used to develop the profile can (and will) produce an uncertainty factor.

Typical (T/C grade) wires are 1.25% or 2.2 Deg. Celsius (which ever is greater). This for (new) K-type wire.

Furnace manufacturers often substitute extension grade wires rather than higher accuracy (T/C) wires for longer runs. This is fine if you use software correction tables. Unfortunatly, these are rarely verified on a regualr basis and will drift over time.

reply »

#39815

Internal Oven Calibration | 20 February, 2006

Pete, I also emailed you about this. Do you think this OvenRIDER is something that can be made? Those "high mass sensors" you speak of...what does that consist of? Is this something you can buy from Omega?

reply »

#39817

Internal Oven Calibration | 20 February, 2006

Samir, I didn't get the e-mail unless I thought it was spam and erased it. The high-mass sensors are custom made for ECD. The round sensors on the OvenRIDER are the "process" or high-mass sensors. They are designed to never reach zone temnperature. They are measureing thermal xsfr.

reply »

Chunks

#39878

Internal Oven Calibration | 22 February, 2006

I agree with Steve and would hate to have Ken audit me.

"Mrs. Chunks, sure your profile looks like the manufacturers recommended spec, your oven is calibrated to the best you can do or afford, your profiler is also calibrated to the best of your capable abilities, and YES your product is coming out defect free, meets IPC class 3 standards, Trace labs says your intermetalics look better than anyone else�s, BUT those darn thermocouple wires look like they could use some calibration! Oh you're using a chain conveyor too! Oh my too bad. That means your conveyor speed is off since you never measured your link-to-link wear over the last year and correlated that to the heat coefficient of expansion over each zone and compared that to the over all relative extent it has on your chain length? I do believe we're going to site you failure to comply!"

;^)

reply »

PCB contract manufacturer

SMT in-printer dispensing