Just a quick note on the use of capacitors as voltage droppers -
Yes, high frequency disturbances do get through the series cap and upset
things - that's why capacitive dividers are much better. They also have the
advantage of lowering the source impedance somewhat and the shunt capacitor
doesn't need particularly  high voltage ratings.
You still need some form of spike supression however

Fair comment re the parallel operation - but do leds fail open or short? -
A combination of series and parallel leds would seem safest & be the best
way of driving a number off a convienient voltage.
But - the leds in the centre of the mass will run hotter, have a lower
forward voltage & hog the current - so do you add power - wasting dropping
resistors as well, or just try and organise the strings so parallel leds
are at musch the same temperature?


(144 pence to the pound ? - it was 240 where I live(d)).

Richard P









Alan Shinn Wrote

> >Date:    Mon, 9 Jul 2001 20:13:51 +0300
> >From:    "Peter L. Peres" <plp@ACTCOM.CO.IL>
> >Subject: Re: [EE]: Bulb Life
>
> > Peter L. Peres wrote:
> >>
> >> > LED traffic lights
> >>
> >> Of course everyone realises that in the event of a major surge or EMP
> >> all the traffic lights will die. With or without switchers. (LEDs
> >> don't like reverse biasing).
> >
> >> Interesting! Would you suggest a reverse diode across the LED array?
> > Or better still a big zener or specialty surge diode, this would
> > protect for both forward and backward surges, considering there should
> > be a fair series resistance anyway? -Roman
> --
>
> If you put half of your LED bank in parallel but with reverse polarity
> (anti-parallel?) to the other half you get full wave power use and no
> reverse polarity problems. You could also feed a non-anti-parallel
> system with a full wave bridge rectifier if you want to buy an extra
> part. Now use a series cap for non-dissipative current limiting and you
> might have a survivable system (with the bridge system the cap is on an
> input leg).
>
At last.  I have been watching this and waiting for someone to come up with
the back-to-back solution. I will now throw in my groats worth.

1. Back each individual LED with another one. Do not parallel long chains
of
LED's. Why Not ? Hint : Consider the failure of a single LED.

2. Do not use a capacitor for current limiting. Why not ? Using Capacitive
reactance is very attractive (and effective). However, what tends to be
ignored is that interference travels through a series capacitive dropper
like a hot knife travels through Butter. So with a capacitive current
limiter your spike protection has to be considerably more robust than that
for a resistive (or better still an inductive) current limiter....Discuss

3. The Groat is a redundant English Coin which even I do not remember as a
lad, but I do remember the Farthing. Ah, happy days when there was 144
Pence
to the Pound.

Regards

Chris Carr

--
http://www.piclist.com hint: PICList Posts must start with ONE topic:
[PIC]:,[SX]:,[AVR]: ->uP ONLY! [EE]:,[OT]: ->Other [BUY]:,[AD]: ->Ads

--
http://www.piclist.com hint: PICList Posts must start with ONE topic:
[PIC]:,[SX]:,[AVR]: ->uP ONLY! [EE]:,[OT]: ->Other [BUY]:,[AD]: ->Ads