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

Printed Circuit Board Assembly & PCB Design Forum

SMT electronics assembly manufacturing forum.

attaching thermocouples


attaching thermocouples | 16 August, 2000

O.K., I give up. I am through trying to solder thermocouples on to 20 mil pitch QFPs. Whether it's my lousy technique (fairly likely), the wrong solder (Kester calls it thermocouple solder, and who am I to's a Sn10Pb88Ag2 alloy) or just bad kharma coming back to haunt me, I don't care any more.

Can anyone suggest a reasonable substitute, like a good thermally conductive and electrically insulative adhesive? I found an epoxy, but really would prefer something removable if at all possible. Is Loctite 384 (used for attaching heatsinks, and removable) an option? Or do I have to spring for Temprobes?



reply »


Re: attaching thermocouples | 16 August, 2000

Hi Steve You really didn't tell us what is going wrong. Not enough heat to melt this hi temp solder (290C/554F)? Too much heat and burning up the pad/board? Thermo falling off? (remove all the original solder) You won't find Loctite 384 all that removable in this application and anyway the need is for the thermo to be in contact with the intended item rather than suspended in a blob of epoxy. Because of the cost I have not tried the temprobes but some here have had misgivings about their contact reliability. Try going to the library section of the link below and reading the paper on thermo attachment. Good luck John Thorup

reply »


Re: attaching thermocouples | 16 August, 2000

Steve: Two things: 1 Chrys "The Thermo Princess" Shea wrote a great thread on attaching thermocouples that's in the SMTnet archives 2 Dymax (and probably others) makes a uV cureable thermal adhesive that's not a ugly to use as soldering thermal couples

reply »



Re: attaching thermocouples | 17 August, 2000

Steve, I'm using 5 Sn/1.5 Ag/93.5 Pb Solder wire with a melting point of 300 deg. C to connect my thermocouples on smd components for thermal profiling. I'm using a soldering iron with an adjustable temp. to be able to melt my solder wire. Temp is about 320 degrees to 335 deg. C and I'm soldering the thermocouple without any problem. I've used some thermal glue before including thermal adhesive to connect my thermocouples to my profile boards but I still find the High Temp Solder wire to be the best way (also cheaper) to use when connecting thermocouples onto the components. I've read hips of articles about the disadvantages of using high temp. solder wire when thermal profiling and have done experiments on some recommended ways of properly connecting thermocouples onto components of the profile board and found not much difference about temperature readings. That's why I'm saying that using high temp solder wire is still effective and economical.

regards, Dreamy

reply »

Brian W.


Re: attaching thermocouples | 18 August, 2000

I have been using aluminum tape to attach thermocouples lately. I got some when I ordered new thermocouples from KIC. I ran some comparisons using high temp solder and the tape, with no difference. You cut a small square of the tape (approx 1/8" square), apply the tape, and put kapton tape over it. The secure the wires to the board using kapton tape. The only problem I have ever had is when the profiler got snagged in the oven, the thermocouples pulled out of the tape. It's removable, quick, and easy.

Good Luck, Brian W.

reply »


OK, now I think I've got it...... | 18 August, 2000

Thanks for all your replies, folks. yes, I've read Chry's detailed description more than once (I too wish I could SEE it done rather than read about it), but even in following it I ran into problems. I was making it much more difficult on myself than necessary, probably, by following someone's recommendation that the tc needs to be between the lead and the land to give accurate results. I've decided that if KIC can recommend that profiling be done on unpopulated boards (how valuble is THAT??) I can certainly attach my tc's to the top of the foot of the lead, or as close to the land as possible on a J-lead or leadless component. Problem pretty much solved.

Thanks again.

reply »

Greg Jones


A Note from KIC Thermal | 21 August, 2000

In Steve's last posting he wrote (in part): "I've decided that if KIC can recommend that profiling be done on unpopulated boards (how valuble is THAT??)"

I contacted Steve off forum to discuss this. In the paper referenced by John Thorup (above), we did use unpopulated boards to compare the accuracy and reliability of various methods of attaching thermocouples to PCBS.

We do not reccomend profiling with unpopulated boards for product or process setup, nor do we reccomend profiling for production with fixtures or "golden" boards. The best test vehicle for a given profile is the product itself, and we reccomend the use of AL tape as a method that is simple, accurate, and nondestructive.

Steve's questions raise interesting issues about the accuracy of TC's attached to the surface of the PCB vs. attached to the leads...there's probably another conference paper in that.


reply »

SMT feeders

pressure curing ovens