T i m wrote:
On Wed, 14 Oct 2020 08:26:12 -0400, Paul
wrote:
snip
So, whilst it is all up and working, I think I'd like to look for an
LED lamp replacement, basically to save power but am having difficulty
pinning down what that might be.
It's a 15mm diameter, non-offset pin, twin contact (that's the bit
that seems to make it more complicated), bayonet lamp and I'd like an
LED that has 'all round' illumination like the original incandescent
lamp, so you get a nice general light, rather than a beam.
I think I have found a suitable lamp:
LED Lamp S8 BAY15D (MCS8Y15D16NS30DCW) ... but they seem quite
expensive and often from
snip
https://en.wikipedia.org/wiki/Bayonet_mount#Light_bulbs
I looked though this one:
https://en.wikipedia.org/wiki/List_o...ght_bulb_types
But it still wasn't clear if there was a code that described that base
(with two connectors and one filament) exactly?
If you need a better price, pump that part number into Google.
I know in the past, those LED replacement bulbs had ridiculous
pricing.
Yeah, I did that but didn't find anything (other than the one I linked
to) and as you say, often quite expensive. Well, I found one that was
affordable but the P&P from the USA wasn't. ;-(
Try a local auto parts store, and check the
"ricer aisle" for LED bulbs like that.
Trying to avoid bricks-n-mortar places atm (but I would normal do as
you suggest). ;-)
The datasheet I got for your part number, isn't exactly very good.
But that's all I can find for it.
http://www.farnell.com/datasheets/1564831.pdf
Noted.
- the LED might overheat, if there is a globe cover over
the whole assembly.
I was thinking that most 'car lamps' are running in a similar sized
enclosure?
For the number of service hours this
application will receive, it won't matter.
Hopefully.
- the light output might not be uniform with respect
to direction.
No, that was one of the 'worries', compared with the frosted
incandescent 'globe' that is in there now.
More light from the side than the end.
That shouldn't be a problem (and might even be an advantage), given
there is a basic reflector in there.
In a car light assembly, the reflector and diffuser
fix this.
This one has a reflector (of sorts) but no diffuser, relying on the
lamp in the std design.
- the lamps with LED filament chains might be no more
robust than an incandescent, when you drop your
inspection light. The product above appears to use
discrete LEDs which are likely sitting on a flat
substrate.
Agreed.
A proper datasheet for a product like that, should
stretch to about 30 pages in the PDF.
Well, I just need to know a few bits of info and you would have
thought 'fundamental' points, re how it was wired might be a one of
them? ;-)
Something that has since come to mind though ... a 'what if' ... ;-)
The lamp holder is fairly substantial metal (not needed as such as the
connection is just via the base connections) and on an even more
substantial metal disk that sits in a groove in the clamshell
'handle'.
The two contacts to the lamp base are made by nicely engineered screw
terminals ... so, if I was to use an easier-to-find dual filament lamp
(say a stop tail) then I could re-wire the unit so that the ground is
on the lamp base (as normal by drilling and tapping for a small
machine screw or a small nut / bolt though the flange with a
pear-shaped-tag) and re-wire the lamp holder to give the 5W on all the
time (when connected to 12V) and then the 21W on (as well) when it's
turned on at the rocker switch.
Or ... replace the SPST switch with a SPCO, so I can swap between lamp
sides (21/5W), if that might make the lamp last longer if both 'sides'
aren't on at once (you would normally only have the side lights on
continuously and the stop occasionally) or if I can get a SPCO with
centre off, have a filament on each side of the switch and could also
turn it off with the switch (a bit like a dual heat hot air gun)?
You could go further and mount some SM LEDS / Cree modules directly to
the disk but I'd like to keep it as standard as possible. ;-)
Cheers, T i m
It's possible the two pins are (+) and the shell is (-).
You have to current steer, or current share, or do something
with the two pins, to drive the chains of LEDs and
give a pleasing light pattern.
One mystery with some of this stuff, is what is the
semiconductor other than the LEDs and resistors for ?
The only thing I could figure, was a zero-bias FET
as a constant current source. But that never makes sense
because these zero bias FETS, the current flow was
way too low for usage with LED arrays. You can do
zero bias FETS with a single LED, as that's enough
for a pleasing light level with a cheap T-1 3/4 LED.
https://www.electroschematics.com/wi...age-range-led/
Your lamp product appears to have a 10V (four white LED)
to 30V range. I don't know how that's possible, while
at the same time claiming to be consuming 5W for the
lamp function. If there was any sort of current
limiting linear function in there, it would be
getting hot.
The 16 LEDs could be split as four groups of four LEDs.
And the four LEDs in a group, could have 10V across them.
OK, now, how do we connect those four chains of LEDs
together ? Four chains in parallel ? Or some other
arrangement ? And operating off a car battery, that's
not really a well regulated source as such. If you
were to jam the field winding, the automotive bus
voltage goes up to 18V. So any LED lamp has to withstand
at least 18V. The other specs for automotive situations
involve pulse patterns (load dumps) at much higher
voltages.
If the lamp was larger, we could assume there was an
SMPS in the base. That would solve a lot of problems.
Paul