RE: LED Road Test - Induction Heating Reflow Soldering - Blog 16 Jan12

DAB — Wed, 13 Jan 2016 20:52:08 GMT

Hi Douglas,

There are several issues with induction heating various materials.

You need iron rich material to heat properly. Stainless steel does not do a good job as the crystal structure inhibits the magnetic fields.

Aluminum is diamagnetic, which means it will do some interesting things in magnetic fields.

Overall, I think if you get an old cast iron skillet, you can probably do a good job of controlling the temperature with your induction stove.

DAB

RE: LED Road Test - Induction Heating Reflow Soldering - Blog 16 Jan12

shabaz — Wed, 13 Jan 2016 10:53:07 GMT

Hi Doug,

You're on fire (not literally ever I hope [View:/resized-image/__size/16x16/__key/commentfiles/f7d226abd59f475c9d224a79e3f0ec07-306b2e92-3ec7-4f9d-868f-3f22b6dd51be/5482.contentimage_5F00_1.png:16:16] ) with all these experiments!

[mention:f80b53cee57c44bc9d7c577d07d7c791:e9ed411860ed4f2ba0265705b8793d05] did some experiments a few months ago using the LDC1000 device, and that may help determining the eddy current effect. The information is here: Ti LDC1000EVM A Question on Inductance

(in the post and the discussion comments).

At that time, it was shown that the LDC1000 measured value in the presence of steel was due to eddy currents more than permeability, and this was because of the high frequency that was used by the LDC1000 device.

So, by using the LDC1000 with different thickness metals, it is possible to see (with a extremely high resolution - 24-bit) if thickness is impacting the measurement any longer after a certain thickness is reached.

In your case, I guess that thickness is the top metal part of your induction heater but then you've got the additional thickness of the aluminium plate and then the MCPCB.

Also, there is a formula here for a depth which appears to be about 1.8mm for the worked example there (some particular steel). At that depth, eddy currents are almost a third (37%) of whatever they are

on the surface. And another 1.8mm would reduce that to almost a third (37%) of that, and so on. So with a thick plate, the eddy currents will have diminished. For different metals, there is a very nice metals conductivity list here.

Regarding relative permeability, I'm not sure what aluminium would be - apparently relative permeability changes a lot with temperature, so may be close to 1 (no idea) at the soldering temperature. Anyway

it works in your favour for most of the time, since a higher (cold) value would reduce depth since it is part of the denominator in the formula.

Google books found 'Practical Induction Heat Treating' by Richard E. Haimbaugh and it had the diagram below which shows it for different metals and different frequencies - quite interesting!

Maybe an interesting method might be some sort of shaped metal, e.g. some metal base with protruding cylinders of LED width, long enough to conduct heat but not have any significant eddy currents at the far end.

Could be a cool way (again not literally!) of supplying heat to the LEDs without requiring to heat a massive block of metal basically. But I don't really know much (anything) about thermal design to be honest.

[View:/resized-image/__size/620x377/__key/commentfiles/f7d226abd59f475c9d224a79e3f0ec07-306b2e92-3ec7-4f9d-868f-3f22b6dd51be/3348.contentimage_5F00_180593.png:620:377]

All this talk of Eddy currents..

[View:https://www.youtube.com/watch?v=vtPk5IUbdH0:520:333]

RE: LED Road Test - Induction Heating Reflow Soldering - Blog 16 Jan12

jw0752 — Wed, 13 Jan 2016 05:32:16 GMT

Hi Doug,

I have to admit that I do not know much about inductive heating. Very interesting to see your experiment. What is the frequency of the field that is generated by the stove?

John

RE: LED Road Test - Induction Heating Reflow Soldering - Blog 16 Jan12

Robert Peter Oakes — Wed, 13 Jan 2016 05:30:44 GMT

Would there not be an issue with the induction field acting upon the metal inside the LED as well causeing or potentially causing damage there, I know it worked in the video but this seems to be a potential area for problems ??