Re: [SI-LIST] : Micro Noise

About this list Date view Thread view Subject view Author view

From: Jon Keeble (jkeeble@alpha.net.au)
Date: Mon Nov 08 1999 - 00:49:34 PST


This story is about two parallel boards, one with 24 bit A2D and D2A converters (and associated op amps, differential line receivers / transmitters), the other with a DSP and other digital bits.

The problem showed up in rev4 of previously working boards.

To cut a long story short, proximity (it had to be very close) ruined the noise figures (increased the noise floor). Production found that interposing a floating metal shield fixed the problem, not a solution that would have occurred to me.

Engineering took a look, and found a wire link placed in a couple of through holes that hadn't been trimmed - the wire got to close, to pin1 of a 20 pin GAL - the clock pin on the circuitry below.

My explanation is that the floating metal shield creates a unipotential that averages out the different potentials, effectively decreasing the peak E field near the wire.

Could some similar mechanism be in play here?
    -----Original Message-----
    From: Dennis Tomlinson <det@tellabs.com>
    To: si-list@silab.eng.sun.com <si-list@silab.eng.sun.com>
    Date: Saturday, November 06, 1999 10:14 AM
    Subject: Re: [SI-LIST] : Micro Noise
    
    
    Chuck Hill wrote:
>
> Derek,
>
> I remember another fellow having this problem with an Intel 80186
> processor. He placed a piece of copper tape on the package--I don't
> remember whether he grounded it or not.
    
    This is the second or third reference on this thread to a floating shield.
    I have some concern over whether or not this would be at all effective.
    
    Wouldn't an E-field be unaffected by such a shield? I can believe that an
    H-field could be shielded in this manner, but to what avail? A piece of
    copper tape stuck to the top of a plastic IC package is in the "very" near
    field - dominated by the E and not the H. I suspect a biconical antenna placed
    three meters away would show little or no measurable difference between
    shielded and unshielded.
    
    Just my $0.02
    
    Dennis
    
>
> Charles Hill, consultant
>
> At 10:31 AM 11/3/99 -0500, Lfresearch@aol.com wrote:
> >Hi all,
> >
> >I have a problem with a microprocessor which appears to be radiating very
> >strongly from the package. I've deactivated all other circuits ( pulled
> >crystals, pins etc ) on the board. Additionally, the micro is held in reset,
> >PROM's are pulled, but it's a noisy as heck. Close field probes ( both E and
> >H field ) show the package to be the dominent source of radiation.
> >
> >The device has a maximum clock rate of 50 MHz, this is where we're using it.
> >Clock it at 32 MHz and the emissions fall about 10 dB up to about 300 MHz,
> >more above 300 MHz. Yes it's a CMOS device.......
> >
> >Am I stuck with putting a can over the chip? The product has a plastic
> >package, so there is no shielding available at all.....
> >
> >Thanks,
> >
> >Derek
> >
    
    **** To unsubscribe from si-list: send e-mail to majordomo@silab.eng.sun.com. In the BODY of message put: UNSUBSCRIBE si-list, for more help, put HELP. si-list archives are accessible at http://www.qsl.net/wb6tpu/si-list ****

**** To unsubscribe from si-list: send e-mail to majordomo@silab.eng.sun.com. In the BODY of message put: UNSUBSCRIBE si-list, for more help, put HELP. si-list archives are accessible at http://www.qsl.net/wb6tpu/si-list ****


About this list Date view Thread view Subject view Author view

This archive was generated by hypermail 2b29 : Tue Feb 29 2000 - 11:39:41 PST