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

Printed Circuit Board Assembly & PCB Design Forum

SMT electronics assembly manufacturing forum.


Criteria for thermocouple wire attachment

Views: 177796

Jack

#46322

Criteria for thermocouple wire attachment | 20 December, 2006

Hi all,

Is there anyone know the correct way to obtain the accurate temperature profiling for oven reflow? Are below claims true?

(1) Instead of full components on the PCBA (Motherboard), the profile taken from PCBA with few components can be considered accurate. In other words, reflow profile from incomplete PCBA(Some SMT components are not mounted) is the same as reflow profile from PCBA with full complete component mounted. (2) This component(DPAK type) has heat sink at the bottom and is located on the big copper layer area (Light geen solder mask color indicate copper layer underneath). Question: In this case, attach thermocouple wire to the component lead is sufficient to obtain accurate temperature profiling.

These 2 claims sounds weird to me but I need confirmation from expert here.

Lastly, wish every single SMTnet user "A Merry X'mas & Happy New Year"!!

reply »

#46324

Criteria for thermocouple wire attachment | 21 December, 2006

Good morning,

(1) This depends on component type. In general it would be better to have fully populated board. Or at least you should mount components with different thermal mass. You should place TC�s at three points at least: component with high thermal mass (transformers, metallic connectors, D-PAKs and so on), components with low thermal mass (chips) and heat sensitive components. If you assembly is populated with BGAs put the TC under BGA as well. Profile after that. Assumption that you will get accurate readings without BGAs, transformers and so on is wrong. Assumption that you will get accurate readings without some chips seems to be right. (2) You should attach it to heat sink of course. You won�t get accurate readings if you attach TC to DPAK�s lead different from heat sink.

PS Happy New Year :)

BR, Pavel

reply »

Robot X-Pert

#46327

Criteria for thermocouple wire attachment | 21 December, 2006

You guys treat this profiling nonsense like rocket science.

Here's how easy it is to "reflow" and "profile".

http://www.seattlerobotics.org/encoder/200006/oven_art.htm

reply »

#46331

Criteria for thermocouple wire attachment | 21 December, 2006

1. Yes you can get away with a partial popped board. BUT, it really depends on your board (see 2nd paragraph). Your main rule of thumb when determining where to place t-couples is your board will heat from the outer edges and heat up towards the middle. So since small parts heat up fast and big parts take longer, you want to measure small parts on the outer edges and heavy heat sinking parts in the middle. Thus you can see if you over heat any small parts while at the same time make sure you are reflowing the heavier parts in the middle. So in theory, you could just measure these parts described and get a pretty good profile

Paragraph 2 >>>> Now you must take into account your total board mass as well. If you have tons of heat sinks, or parts on the other side that can affect your profile, you should have them on the board when profiling. That�s where the "engineering" part in all this comes in.

2. When profiling something like a D-Pak, you may want to drill a hole form the opposite side thru the board so the t-couple hits the ground pad od the D-Pak. Use adhesive to fill the hole and keep the t-couple in place. This will give a very accurate reading of what a real board will see. If you cannot drill holes, then yes, attaching the t-couple to the lead is OK. Realize that this temp will be hotter than what the real ground pad will see.

reply »

Jack

#46345

Criteria for thermocouple wire attachment | 21 December, 2006

Hi Pavel, Real Chunks,

Thanks for your valuable inputs..Selecting the correct location for board profiling is critical in determining the solder joint quality. Am I right to say TC wire attachment locations priority should be as below:

(1) Thermal mass of component. (2) Proximity of component to components of high thermal mass. (3) Proximity to edge of assembly. (4) PCB pad with big copper layer underneath(Usually, light green solder mask indicates copper layer underneath).

Hi Real Chunks,

On item #2, usually, if we were to attach TC wire to D-PAK lead, what is the temperature different compare to the real ground pad?

As you know, voids phenomenon is very likely to happen in LEAD FREE process, especially part with heat sink on the bottom. Hence, in order to reduce such defect, I believe accurate profiling is paramount important. Usualy, we just play around the pre-heat and soaking zone to reduce such defect. What is your profile recommendation to reduce voids at DPAK components in motherboard assembly?

Cheers, Jack

reply »

#46346

Criteria for thermocouple wire attachment | 21 December, 2006

#46396

Criteria for thermocouple wire attachment | 27 December, 2006

Hi Jack,

I've seen around 7 degrees diff in max temp. The real factor here is time to reach peak and time at peak.

reply »

SWAG

#46397

Criteria for thermocouple wire attachment | 27 December, 2006

Look just above the toaster oven in the last picture - these guys have a chipshooter, too!

reply »

#46398

Criteria for thermocouple wire attachment | 27 December, 2006

Cal

#46404

Criteria for thermocouple wire attachment | 28 December, 2006

Skillet Soldering...http://www.sparkfun.com/tutorial/ReflowToaster/reflow-hotplate.htm

What next...George Foreman Reflow???

reply »

#46408

Criteria for thermocouple wire attachment | 28 December, 2006

How cute is that?

We and Earl Sir Moonbeam were using those electric fry pan 'hot plate' things years ago. Look here: http://www.smtnet.com/forums/Index.cfm?CFApp=1&Message_ID=10155

Haters, turn down the volume.

reply »

#46554

Criteria for thermocouple wire attachment | 9 January, 2007

Hello everyone,

any estimation how many (minimum)thermocouples on board, how to calculate this ?

regards

reply »

#46562

Criteria for thermocouple wire attachment | 9 January, 2007

You could probably make do with two as long as you attach to the smallest part in a sparsely populated area and a large heavy part in a densely populated area to cover both ends of the spectrum. When I had a MOLE I used all 6. Now I use the 3 the oven gives me.

reply »

#46583

Criteria for thermocouple wire attachment | 10 January, 2007

I always try to profile using a fully populated board and soldering the thermocouples. This is the only true way of getting an actual temperature of the joints. But, with that being said, its not always feasible to use this.

There have been a couple papers written on the subject.

Thanks, Mario Scalzo mscalzo@indium.com

reply »

asbok...

#46606

Criteria for thermocouple wire attachment | 11 January, 2007

Why don't you just use a Super MOLE oven rider? And besides, profiling will depends on the paste you are using, isn't that right?

reply »

One stop service for all SMT and PCB needs

fluid dispensers