Connect with us

Inductor 'spoiled' by utrasonic cleaning at 65C?

Discussion in 'Electronic Design' started by George Herold, Aug 1, 2013.

Scroll to continue with content
  1. Hi all, So we are using the following inductor in a series RLC circuit.
    (C = 0.01 uF ceramic (npo) f~6 kHz, Q >= 30.)
    The inductor must be a little ‘on the edge’. When we got the first batch in there were 3 out of 100 that had terrible Q and low inductance (as measured on a SRS lcr meter.) (Q measured at 10 kHz.) In the process of finding the bad eggs all the inductors were tested. Now another batch of boards has come for testing and for the first 3 all the inductors are bad. Inductors are hand soldered and then go into an ultrasonic cleaner. (I’m not sure about exact times and temperatures, but maybe 65 C for maybe 1/2 hour?) Could the solvent, ultra-sonics or temperature cause the failure? Anyway to bring the inductors back to life?


    George H.
    (I’m going to try and kill a few in the bath this morning.)
  2. Glenn

    Glenn Guest

    Hi George

    I think the problem is that you have:

    * a very thin cobber wire spooled up.

    * under a plastic sleeve

    When water or solvent get in under the sleve it get sucked in because of
    the capillary effect of the pores between the windings. If it is water
    you use, you might have shorted water windings or just water with bad

    The inductor ought to be moisture open so the moisture or solvent can
    escape during the drying fase - or the windings must be soaked in
    lacquer or enameled so a another fluid can not get in any more.

    The ultrasonic treatment might damage the isolation? I assume the
    ultrasonic treatment produce vacuum/air/solvent gas bubbles, that
    implode - and in this process produce very hot spots on the surface. The
    are treating the electronics with cavitation implosions! No surface
    (isolation) can endure that?:
    Quote: "...
    Such cavitation is often employed in ultrasonic cleaning baths and can
    also be observed in pumps, propellers, etc.
    However, it is sometimes useful and does not cause damage when the
    bubbles collapse away from machinery, such as in supercavitation.
    Quote: "...
    The agitation produces high forces on contaminants adhering to
    substrates like metals, plastics, glass, rubber, and ceramics. This
    action also penetrates blind holes, cracks, and recesses. The intention
    is to thoroughly remove all traces of contamination tightly adhering or
    embedded onto solid surfaces.
    Most hard, non-absorbent materials (metals, plastics, etc.) not
    chemically attacked by the cleaning fluid are suitable for ultrasonic
    cleaning. Ideal candidates for ultrasonic cleaning include small
    electronic parts [!?], cables, rods, wires and detailed items, as well
    as objects made of glass, plastic, aluminum or ceramic.[8]

  3. Joerg

    Joerg Guest

    Check a good and a bad one carefuly under a microscope. Cut and gently
    peel off the yellow sleeve, then unwind the cooper. Look for hair cracks
    in the core. Also check the wire and look for embrittlement of the
    enamel because that could cause two turns to short.
  4. Geesh, now you tell me :^)
    We now have to do this Rhos stuff to ship to Europe, and we recently switched solders and added this ultrasonic cleaning step to all pcb's.

    George H.
  5. Thanks Joerg, (see my reply to Glenn) It appears the solvent is doing something.. I'll try peeling back the yellow cover and see what that does.

    Short answer is not more of this solvent for inductors!

    George H.
  6. Joerg

    Joerg Guest

    Looks like at least my news server didn't pick up that reply. But yeah,
    if the solvent is attacking the enamel that would not be good.

    Or a different, less aggressive one.
  7. Hmm, I don't see my reply to Glenn either. So let me recap.
    My first test was just to stick the inductor into the solvent. It failed with one dunk. I also made a mistake in my first post, the inductance *increases* and the Q goes down. (Resonant freq from 6.13kHz to 5.43 kHz so L~86 mH.)

    I also found that some of the 'bad' inductors that I pulled several weeks ago are now OK.

    So I peeled the yellow heat shrink off of a bad one. No change. Then washed with hot water. a bit better.. then dried with heat gun. Good as new.

    George H.
  8. Joerg

    Joerg Guest

    Maybe the solvent seeps underneath the windings and on the first three
    didn't have a chance to come back out?
  9. I'll try to find the article, but yes Ultrasonic cleaning is bad for
    components because of the stresses involved. It can weaken bonding
    wires and cause stress fractures. You dont see the failures right
    away but over time they become apparent.

    De-ionized washing or a good old vapor degreasing is better.
    For the DIY ers Ensolve works good.

  10. John S

    John S Guest

    Water has a dielectric constant of about 80. So, there should be some
    shift in Fo due to the increase in stray C. However, I can't justify
    your change with just the dielectric constant.

    John S
  11. legg

    legg Guest

    I believe you may be misapplying the part. It's not intended for
    High-Q use - it's a loose tolerance hash choke employing an iron dust
    core. The lossy-er the better, in that application. The Q measurement
    itself is a fairly good indication of functional integrity, however.

    The low inductance can be caused by internally shorted turns and
    shorts through the core - a product of vibration, bumps, abnormal
    winding tension, poor layering technique, poor core surface finishing,
    heat cycling and lousy magnet wire.

    If bad parts were found in the first batch, prior to processing, this
    was/is a sign to reject the whole batch. There are plenty of other
    sources that won't exhibit this failure mode.

  12. Thanks for all the responses, (Jon, legg, Martin, John S.)

    John I think you nailed it. The problem was the water. (Apparently the solvent is water with some special 'soap' added.) So the water gets between the turns and increases the turn to turn capacitance. (At least that's myworking hypothesis.) There's a piece of data I didn't really notice and that is with a 'wet' inductor the resonance not only shifts to lower frequency and amplitude, but the peak of the resonance happens at a phase shift that is less than 90 degrees from the applied voltage.

    legg, I didn't have any part in selecting the inductor. (My only contribution was to suggest a nice npo ceramic cap rather than a film one.)

    George H.
  13. Fred Abse

    Fred Abse Guest

    Philips did some research on board defluxing in the 1970s.

    Your process is what they deemed best, albeit with now-banned 1.1.3 TCTFE
    and isopropanol (50% w/w).
  14. Fred Abse

    Fred Abse Guest

    What solvent?
  15. It's mostly water. The container says elma tec clean A1

    George H.
Ask a Question
Want to reply to this thread or ask your own question?
You'll need to choose a username for the site, which only take a couple of moments (here). After that, you can post your question and our members will help you out.
Electronics Point Logo
Continue to site
Quote of the day