Want to read Slashdot from your mobile device? Point it at m.slashdot.org and keep reading!

 



Forgot your password?
typodupeerror
×
Hardware Technology

New Heat-Reduced Magnetic Solder Could Revolutionize Chip Design 103

A new heat-reduced soldering technique using magnets may lead to some revolutionary changes in the way chips are manufactured. Details are scant since the inventor seems to be playing it close to the vest for now in hopes of attracting chipmaker interest. "The result is a tin-silver alloy that contains a dispersion of iron particles tens of micrometers in diameter. When a magnetic field is applied to the solders, two things happen. First, the iron particles heat up, locally melting the solder. This localized heating, which works on the same principle as inductive stoves, remains completely contained, keeping the surrounding area cool. And second, the iron particles line up with the direction of the magnetic field, squeezing and pushing the liquid in that direction. This alignment is retained when the solder solidifies, and the well-ordered particles provide mechanical reinforcement that's greater than that afforded by a regular dispersion of particles."
This discussion has been archived. No new comments can be posted.

New Heat-Reduced Magnetic Solder Could Revolutionize Chip Design

Comments Filter:
  • A way to solder those broken lobsters! [albinoblacksheep.com]
  • Wait.... (Score:1, Insightful)

    by Moheeheeko ( 1682914 )
    ...Using magnets to solder magnetic sensetive components? Ive got this great idea for a better wood glue! just expose it to fire and it holds 10x stronger and faster!
    • FTFS:

      In contrast, the magnetic solder can be pulled up through silicon by using a relatively weak magnet

      So it depends on their definition of relatively... Is that a household kitchen magnet (which would do little damage to most magnetic sensitive components once removed from the chip)? Or is it a 0.5T magnet (that's relatively weak compared to most MRI magnets and would likely saturate most magnetic sensitive components to the point of failure)?

      • Re: (Score:3, Insightful)

        by Carnildo ( 712617 )

        So it depends on their definition of relatively... Is that a household kitchen magnet (which would do little damage to most magnetic sensitive components once removed from the chip)? Or is it a 0.5T magnet (that's relatively weak compared to most MRI magnets and would likely saturate most magnetic sensitive components to the point of failure)?

        Once the solder melts, it should be possible to shape it using a refrigerator magnet -- molten solder simply doesn't have much viscosity or structural strength. You d

      • by Futurepower(R) ( 558542 ) on Friday March 05, 2010 @04:20PM (#31375706) Homepage
        "When a magnetic field is applied to the solders, two things happen. First, the iron particles heat up, locally melting the solder. This localized heating, which works on the same principle as inductive stoves, remains completely contained, keeping the surrounding area cool."

        Quote from ScuttleMonkey, the Slashdot editor: "Details are scant..."

        Neither the Slashdot editor or the writer of the linked article understand the physics. Magnetic fields cause something to heat only if the field is rapidly changing. Then the magnetic field causes everything conductive to heat, including iron particles.
    • by Jeng ( 926980 )

      Sounds to me like the heat can be applied more precisely with this method than the current method.

      Fairly sure there is wood glue that does hold much stronger if you heat it while it sets.

    • by DeadCatX2 ( 950953 ) on Friday March 05, 2010 @04:40PM (#31375978) Journal

      When you bring a magnet near a PC, the damage is done to magnetic recording media, not the chips. Silicon is not generally sensitive to magnetic fields. This guy [magconcept.com] has managed to put a video game controllers, keyboards, and mice inside an MRI bore. If those integrated circuits can work in a 3T magnetic field, I'm pretty sure it can survive this new magnetic assembly technique.

      • Magnetic sensitive components are now usually shielded. Floppies and other magnetic media were the most susceptible so most people don't have to worry about it much now a days. Plus most shielding is just a faraday cage or iron like the outside of the computer case. HDD have the same issue and have a metal case. As well there are magnets anyways in the hard drives to move the IO head. This technology is actually exciting if you can get a consistent solder. It might even do away with microscopes for some sol
        • Re: (Score:3, Insightful)

          by DeadCatX2 ( 950953 )

          You're tellin' me! Two of the PCBs I designed use FPGAs, one of them a BGA package. It's quite the trade-off when choosing an oven profile that gets hot enough for long enough to melt the solder but not destroy the FPGA (stupid lead-free solder!). Lower temperatures would make our life a lot easier...

    • This could be absolutely awesome for surface-mounted diodes, which are quite sensitive to heat, requiring reflux soldering techniques for now.

  • Comment removed (Score:4, Insightful)

    by account_deleted ( 4530225 ) on Friday March 05, 2010 @03:27PM (#31375144)
    Comment removed based on user account deletion
    • Re: (Score:2, Funny)

      by Anonymous Coward

      How much iron are we talking about? Is this tantalum to having ferrite beads on all connections now?

      -jcr

      re-engineered that for you.

    • by macraig ( 621737 )

      Maybe about the same amount found in JB Weld, only much much finer particles.

      • by toastar ( 573882 )

        sorry for the diversion, But JB Weld is awesome!

        • Re: (Score:3, Funny)

          by macraig ( 621737 )

          That's not a diversion, you're cementing the point and expressing solidarity.

        • by rts008 ( 812749 )

          I heartily second that!

          I figure with a tube of JB Weld, a roll of Duct/Duck tape, and some baling wire, the possibilities are limitless.
          You might even be able to take over the Universe!

          Why, if MacGyver had JB Weld we would not be stuck with re-runs...the plot possibilities just don't stop with JB Weld!

          Okay, I'm starting to sound like a shill(just a happy customer), so I'll quit now. :-)

    • Re: (Score:3, Informative)

      by drerwk ( 695572 )
      No. First, iron and ferrite are not always the same thing, so it might just be iron in the solder. Second, the geometry matters, and the spheres are not beads.
      • Would not all spheres exist within the set of beads. Although not all beads exist within the sphere set.
      • Re: (Score:3, Interesting)

        by Sponge Bath ( 413667 )
        I don't think the particles are spheres. The article says tens of microns in diameter, but also says the particles align to the magnetic field and never mentions spheres. I would guess the particles are cylindrical. Also pure iron has a permeability comparable to some ferrites.
        • Even a spherical piece of iron can have magnetic poles and will align along those in a magnetic field even though you can't really tell just by looking at it.

      • It responds to a magnetic field, so it'll effect the inductive characteristics of the circuit. This would seriously mess up any high-frequency circuit. Unless you can "remove" the ferromagnetic element from the solder once you're done shaping it, this is a big hurdle.
        • by drerwk ( 695572 )
          Not likely. I assume the iron is heated by induction, which would be through a strongish time varying magnetic field. The field induces eddy currents in the iron, which has some resistance. The magnetic field that does this, will not be created be the current running through the solder joint, hence no heating, so the solder will not melt. Like wise, no reason to think the inductance will be anything other than that of a short straight wire.
          • I think what he's saying is that the altered inductive characteristics would need to be accounted for in the circuit, particularly in the case of high frequency stuff like RF amplifiers etc. It could be enough of a difference to require extensive changes to circuitry, but still, those kinds of components can continue to use existing soldering methods if necessary, so no big deal.

            • by drerwk ( 695572 )
              And I am saying I would be surprised if there was any important difference. We are talking about the interconnects between stacked chips; maybe 2mm in length. The conductor is a straight line - the lowest inductance geometry you could hope for.
              Please give me any qualitative argument to support the claim that there would be a relevant change in inductance.
              • So your solid assurance of "meh, there's nothing to worry about" should be enough for engineers who design precise, complicated, and extremely expensive high frequency circuits to just throw it into production without a single worry about possible unintended side effects? That's the point here. Having iron in the joints COULD introduce some sort of unknown problem. Who knows? That's all.

                • by drerwk ( 695572 )
                  I'm thinking that the circuits will probably behave according to Maxwell. The statement

                  It responds to a magnetic field, so it'll effect the inductive characteristics of the circuit. This would seriously mess up any high-frequency circuit.

                  seems rather stronger than your "who knows".
                  So again - bring some physics, or other qualitative argument.

  • by jollyreaper ( 513215 ) on Friday March 05, 2010 @03:31PM (#31375190)

    Will Microsoft install a magnetic field generator in the next Xbox to ensure the solder fails there, too?

  • Accidental reflow? (Score:4, Interesting)

    by pavon ( 30274 ) on Friday March 05, 2010 @03:32PM (#31375204)

    From the article:

    A new type of solder can be melted and shaped in three dimensions under the force of a weak magnetic field

    How weak are we talking about here? I wouldn't want my chips to become desoldered just because they were exposed to an electromagnetic field. The article didn't mention any thing about that.

    • Just don't spec this on any components laid out next to mains surge suppression inductors... ;)
    • How weak are we talking about here? I wouldn't want my chips to become desoldered just because they were exposed to an electromagnetic field. The article didn't mention any thing about that.

      It works through induction heating [wikipedia.org], which requires a time-varying magnetic field. It's perfectly safe to use refrigerator magnets to stick things to your computer case, but you might have trouble using these new chips in the computer's speakers.

      • Re: (Score:3, Interesting)

        by X0563511 ( 793323 )

        Hmm. How would the solder bond to connections though? From what I'm getting from this, the solder (and just the solder) would heat up. Basically you'd get a blob of molten solder refusing to bond to anything because of surface tension?

        (100% guaranteed cold solder joints, basically)

        You could pre-tin the connections, but then wouldn't the solder just remelt (while the metal plating remains cold) and "ball" off?

    • Re: (Score:3, Interesting)

      by moteyalpha ( 1228680 )
      What about -intentional- reflow. It would be useless for Mil Spec certainly and I would guess that war walking could take on a new level of meaning. It would be interesting if they used multipurpose dies that could be connected and recycled to a new configuration. Oh yes, liquid metal terminator brains. Mmm brains.
  • iron, huh? (Score:4, Insightful)

    by metamechanical ( 545566 ) on Friday March 05, 2010 @03:32PM (#31375206)

    The result is a tin-silver alloy that contains a dispersion of iron particles tens of micrometers in diameter.

    Not saying it can't work, as the above is light on precise chemistry, but in an alloy like this, you're bound to have atoms floating around... say, to the surface of the deposition... where it will oxidize. And something like OSP [wikipedia.org] (which yes, wouldn't bond to SnAg) only lasts so long in storage... Don't we already have ENOUGH problems with solder joint oxidation? I look forward to seeing how this issue is addressed.


  • What happens if the finished product is exposed to a strong magnetic field? Does it all fall apart?
  • I would prefer the iron particles to be exactly 10 microns in size, and not 10 micrometers, as it makes the process a whole lot cleaner....

  • magnetic field? (Score:3, Insightful)

    by Lord Ender ( 156273 ) on Friday March 05, 2010 @03:40PM (#31375296) Homepage

    When a magnetic field is applied to the solders [they heat up]

    Not to be pedantic, but this is the Internet... They are applying magnetic flux to their solder, not just a magnetic field. A field doesn't impart any energy.

    • Re: (Score:2, Funny)

      by HalifaxRage ( 640242 )
      Will the PCB include a capacitor for this flux?
    • Re: (Score:3, Informative)

      When a magnetic field is applied to the solders [they heat up]

      Not to be pedantic, but this is the Internet... They are applying magnetic flux to their solder, not just a magnetic field. A field doesn't impart any energy.

      Perhaps true, but keep in mind that in the context of Solder, "Flux" [wikipedia.org] has an entirely different meaning.

    • Re: (Score:3, Informative)

      by zippthorne ( 748122 )

      Neither does magnetic flux. Magnetic flux is a local property of the static field and not the same thing as EM flux. It was a good pun, though.

    • Re: (Score:3, Informative)

      by TeknoHog ( 164938 )

      Magnetic flux is simply the area integral of magnetic field strength. You can't have one without the other. (For the pedantic, the integrand here is actually magnetic flux density, which is proportional to magnetic field strength.)

      However, for inductive heating you need a time-varying magnetic flux, which in this case is probably achieved by varying the field strength. The field can still be aligned on an axis, so the method works anyway.

    • Re: (Score:3, Informative)

      You can't apply a magnetic field to a material where a field didn't previously act without having the field intensity fluctuate (flux).

  • JB Weld (Score:5, Interesting)

    by macraig ( 621737 ) <mark.a.craig@gmCOMMAail.com minus punct> on Friday March 05, 2010 @03:46PM (#31375356)

    JB Weld contains so much iron particulate in suspension that it responds to a magnetic field. If it weren't for the fact that the particles are so much larger and get drawn out of suspension and toward the magnet, it might be possible to speed-cure the stuff with this same trick.

    • Re: (Score:2, Funny)

      by Anonymous Coward

      If only there was some way to alternate the magnetic field...

      But that's crazy Star Trek talk

      • by macraig ( 621737 )

        Yeah, it is the stuff of Star Trek.. I'll use the deflector array! (Thanks for pointing that out.)

  • Solder is already conductive, so the eddy current losses won't be localized in the iron particles. Further, copper traces are even more conductive.

    This must be based on the hysteresis losses in the iron B-H curves. That means he's probably got a very high frequency magnetic field generator that he's using to heat up the iron. Seems like a simple principle.

    That said, I still don't want iron filings in my solder!

    • by pz ( 113803 )

      Solder is already conductive, so the eddy current losses won't be localized in the iron particles. Further, copper traces are even more conductive.

      This must be based on the hysteresis losses in the iron B-H curves. That means he's probably got a very high frequency magnetic field generator that he's using to heat up the iron. Seems like a simple principle.

      Could also be from friction heating as the iron particles get moved around within the solder paste.

  • This sounds like this would use a very high intensity high frequency magnetic field tuned to make the iron particles resonate and heat the material. I suspect it would be like normal solder with some iron particles. As a bonus, small iron particles would act a a dispersion hardener in the alloy. It's also quite possible that the particles are actually nanoparticles of iron rather than micrometer sized ones as this size range works better as a dispersion hardener.
  • This may work in testing, but does the field that is required to make the solder melt going to have impact on the very delicate circuits in a CPU or other chip. That is why the details are closed, because there are other problems.
  • Anyone have any ideas as to what impact this may have on the formation of "tin whiskers"? Slashdot article on 6/15/08 re: Tin Whiskers [slashdot.org]
    • by mcgrew ( 92797 ) *

      You need tin to have Tin Whiskers. 100% tin ensures tin whiskers, which is why lead was added to the tin to make solder; enough lead and the whiskers go away. If you RTFS you'll see that there's no tin in this solder; it's silver solder, not tin solder.

      • If you RTFS you'll see that there's no tin in this solder; it's silver solder, not tin solder.

        From the summary: "The result is a tin-silver alloy..."
        Sounds like tin-silver solder to me, which is common these days.

      • Lead was added to tin to make a good solder long before sizes were small enough to make whiskers a problem. Lead-tin solders are close to the eutectic point, which makes them much easier to melt than either pure lead or pure tin.
    • Tin whiskers only form in the presence of pure tin, as in a tinned PC board trace or component lead. Combine the tin with lead or silver or anything else, and the pure tin crystal structure won't happen, hence no whisker formation. Tin whiskers (dendrites) form when a layer of pure tin is mechanically stressed; the tin recrystallizes as dendrites in response to the strain on the crystal lattice. The recent increase in tin whiskering is due in part to hazmat-reduction regulations that discourage the use of l

    • Why would you want a tin of whiskas on your circuit board? :o http://www.whiskas.com.au/WhiskasAU/en-AU/food/can/default.htm [whiskas.com.au]
  • I read that "New Heat-Reduced Magnetic Soldier..."
  • In soldering 101 I learned that if you put hot solder on cold metal, the solder doesn't really wet the joint, and you will get early or immediate failure of conductivity. You can even accidentally create a diode. So why will this work?
    • by Enleth ( 947766 )

      It is.

      Either it's a bad case of lousy, pseudo-scientific journalism, or the people who invented this were not as bright as it seems. The invention itself is interesting and undoubtedly will be found useful for some very specific uses (molding microscopic parts, maybe?), but soldering electronics is not going to be one of them.

      You just can't put hot solder on a cold contact or PCB pad and expect it to work. It will not. Soldering requires heating the surfaces being joined as well as the solder, to the same o

      • by Khyber ( 864651 )

        "Soldering requires heating the surfaces being joined as well as the solder"

        Nope. Surface-mounting doesn't require this - there's a contact pad and you use reflux soldering or forced hot air to melt a solder paste. Things tend to slip right into place nicely and neatly thanks t surface tension.

        • by Enleth ( 947766 )

          As far as I know, the contact pads are still being heated because the processes involved in reflow and hot air soldering generate much "excess" (actually, not) heat around the solder paste itself and heat those pads, as well as the rest of the whole board, to a uniform temperature.

          Besides, I can recall a technology that involved microwaving a board with a water-based solder paste on it and AFAIR it was said to be highly unreliable oexcept in strictly controlled conditions because of the pads not getting hot

    • Re: (Score:3, Informative)

      by mikael ( 484 )

      It's for use with assembly line surface mount technology [wikipedia.org]

      The boards are then conveyed into the reflow soldering oven. They first enter a pre-heat zone, where the temperature of the board and all the components is gradually, uniformly raised. The boards then enter a zone where the temperature is high enough to melt the solder particles in the solder paste, bonding the component leads to the pads on the circuit board. The surface tension of the molten solder helps keep the components in place, and if the sold

  • Is involved with the metcal [metcal.com] hand solder station, with the twist that there are magnetic particles in the solder.

  • Well ok, if the iron particles heat up that's fine it melts the solder. But after that I can't see them adding anything in the way of strength since last I knew, a solder joint was a Covalent Bond, not a mechanical one.

    Anyone know if you can get a covalent bond between iron and tin/silver? The whole connection is based on the covalent bond between copper and silver/lead/tin. Not being a chemist I am not sure if these things matter or not.

    Now the other part is, how does one keep said iron particles from ge

  • Sounds like just the technology I need to speed up assembly of the new MRI machine I'm building... D'oh!
  • by NetNed ( 955141 )

    But does it beat THIS!? [asseenontv.com].

  • The kdawson scripts posts the vapourware snake oil shill press releases. What's the ScuttleMonkey script doing muscling in?
  • AFAIR, a major problem before phase-change memory [wikipedia.org] can become a flash memory replacement is its sensitivity to heat and the resulting modifications that producers would have to introduce to their manufacturing processes (e.g. putting data on the memory chips after, not before assembling).

    Would this technology lift this requirement from them by lowering the tempeartures involved in the soldering process?

Every nonzero finite dimensional inner product space has an orthonormal basis. It makes sense, when you don't think about it.

Working...