Re: !!!Magnet Wire!!! TI-H: URGENT regarding EL backlight

[Prev][Next][Index][Thread]

Re: !!!Magnet Wire!!! TI-H: URGENT regarding EL backlight


>From: Firepower5@aol.com
>In a message dated 8/4/98 4:38:25 AM EDT, larry1492@juno.com writes:
>> Magnet wire AKA Winding wire is designed for making magnetic
>> fields when wound around a iron core.
>
>Also, just to interject, you don't need to do it around an iron 
>core.  As far as I know, anything will work.

Yep, core material is selected based on application.

Anyway, now that I can post, let me expound on the El backlighting
EMI problem.

I personally have installed 5 El kits in TI-85 calculators, using
the method detailed on my web page, with slight variations in switch
installation.  The wire used in the installation was 18AWG solid core
singel enamel wire, AKA 'magnet' wire (manufacture available to any
intrested parties).  In each installation the inverter ground wire 
was between 1 and 1.5 inches long, and the supply wire was between
2 and 3 inches.  The wires running from the inverter output to the El
lamp were between 4 and 5 inches, twisted together at about 3-5 turns
per inch, then encased in heat shrink tubing (to prevent any
possablity of a short to the calculator PCB).

In each case, the only problem evident was a dimming of the LCD when
the inverter was on.  This is apparently due to the current draw of
the inverter (~35ma).  Fresh batteries eliminate the problem.

The only trouble reports I have received involved apparent EMI from
the inverter and a drastic calculator slow-down when the invert was
on.  

A definate cause of the slow-down problem has yet to be isolated,
but the problem is only evident with strong batteries, and is solved
by placing a small resistor in the power lead to the inverter.

The EMI problem is evidenced by rolling bars in the LCD or trash
showing on the LCD.  Shielding of the inverter and lamp wires with
a grounded foil shield attenuates the EMI, but there is no official
solution yet.  Sensitivity to EMI from the inverter may differ between
calculator models, which would explain the apparent randomness of
what calculators the problem effects.  Other factors may include the
type of wire used (solid core or multistrand), length of the lamp
supply wires, supply wire twisting, and inverter location.  All 5
kits that I have installed located the inverter at the bottom of the
calc, near the positive battery connector.  At least two of the
calculators exibiting the EMI problem had the inverter located near
the LCD, with very short lamp supply wires.  However, at least one
other calc with the inverter in the same location and using multi-
strand wire was not effected.

In one case, a TI-86 that was severly effected by the EMI from the
inveter has stopped working.  I do not know if the failure is related
to the EMI, an installation problem (likely not), or if it was totally
unrelated to the kit.  My personal calculator, the first to use the
backlight kit, has operated trouble free for well over 100 hours.

The only problem with the installation was due to carelessness,
where I did not properly insulate the wires (as can be seen in the
images on the installation web page).  After about 20 hours of
operation, the calculator was dropped and the inverter output wires
shorted together, causing the inverter to fail.  The output end of
the inverter became hot enough to melt the plastic casing on the
inverter.  The calculator was in no way damaged, the problem was
detected before anything burst into flames ;)
Any other inverters that fail in this manner will be replaced free
of charge (damaged inverter must be returned to me).

In any case, I no longer supply wire with the kits unless requested
to do so.  Standard order includes the lamp, the inverter, and a
length of heat shrink tubing with which to encase the lamp power
leads.  Installers who are concerned that solid core wire will cause
excessive radiated EMI are free to use multistrand or Litz wire.

For those calculators that experiance severe EMI from the kit,
it may be wise to discontinue use of the kit until the EMI problem
is solved, in case the TI-86 failure mentioned above was indeed
caused by the EMI (which would have to be pretty darn bad to fry a
chip).

DK
----------------------------------
David Knaack <dknaack@hotmail.com>
Electroluminescent Backlighting for
TI Calculators and other devices:
http://www.GeoCities.com/SiliconValley/Lab/5725/


______________________________________________________
Get Your Private, Free Email at http://www.hotmail.com
Follow-Ups: