Paul,
A matter of interest: A number of years ago, I ran a test program in
our EMC lab on LED traffic lights. We tested a number of different
brands because they'd been failing prematurely. Texas has a
requirement that the lights crowbar the 120 VAC line from the control
box when 15-25% of the LEDs in the light go bad. The failures seemed
to be related to this feature, because other states hadn't reported
the failures. It turned out there were several precipiating factors,
the most prominent of which were extreme sensitivity to RF in the
Low-Band VHF range and sensitivity to strong, transient electrical
fields.
The problem was caused by the long (4-5") gate lead of the SCR which
was used to crowbar the incoming AC and blow the on-board fuse. It
was driven by a logic chip, but it was not decoupled for RF or fast
transients. A vehicle passing under the light with an active low-band
transmitter would trigger the SCR. A nearby lightning storm could
also take out many traffic lights in a community.
Decoupling the gate line with capacitors cured the problem. A big
ferrite on the AC lead also worked well. We did not see any
out-of-spec emissions from the lamps, though. If there had been any,
TXDOT would have rejected them.
Brad
KV5V
On Sat, Apr 7, 2012 at 11:27 AM, N4XM Paul D. Schrader <n4xm@iglou.com> wrote:
> I have been complaining to ARRL and my Director for YEARS about the RFI
> from LED traffic signals.
_______________________________________________
UR RST IS ... ... ..9 QSB QSB - hw? BK
|