Maker Pro
Maker Pro

Are white LEDs amenable to use as a strobe?

E

Eric R Snow

Jan 1, 1970
0
Would a white LED, a Luxeon Star for example, be useful as a strobe
light for checking engine timing etc.?
Thanks,
ERS
 
J

Jonathan Kirwan

Jan 1, 1970
0
Would a white LED, a Luxeon Star for example, be useful as a strobe
light for checking engine timing etc.?

LEDs are faster than bulbs, so I cannot think of a good reason why they wouldn't
make a decent timing light. I'd be surprised if someone hasn't already
capitalized on the idea!

Jon
 
R

Rich Grise

Jan 1, 1970
0
Would a white LED, a Luxeon Star for example, be useful as a strobe
light for checking engine timing etc.?

If you pulse it properly, I see no reason why not. Look up the pulse spec
on the LED. The pulse width translates into how wide of a smear the timing
mark makes. I've seen backyard mechanics use white "metal-marker" to
emphasize the timing mark on their unit. Not with an LED timing light, but
the principle is the same. The challenging part would be the pulse-forming
network to get the pulse as narrow as possible.

Then again, a one-millisecond pulse is probably not that hard to
accomplish electronically, and it is without question narrow enough to get
your timing within less than a degree.

Good Luck!
Rich
 
R

Robert Monsen

Jan 1, 1970
0
Eric said:
Would a white LED, a Luxeon Star for example, be useful as a strobe
light for checking engine timing etc.?
Thanks,
ERS

I built a strobe out of white LEDs a couple of months ago, and had fun
'stopping' an AC fan I had in the window. My kids thought it was cool,
particularly when I started using different rates, showing more fan
positions.

It worked quite well, even with cheap LEDs; I built it out of a PIC chip
and a big current mirror, with 10 or so LEDs. The brightness was
adjustable using a pot. I used a fixed 1% duty cycle (adjustable using a
constant in the software), and a variable frequency based on the A/D
converter. Simple software, took about 15 minutes to write.

I'll post the software/schematic if anybody is interested.

--
Regards,
Robert Monsen

"Your Highness, I have no need of this hypothesis."
- Pierre Laplace (1749-1827), to Napoleon,
on why his works on celestial mechanics make no mention of God.
 
D

Don Bruder

Jan 1, 1970
0
Robert Monsen said:
I built a strobe out of white LEDs a couple of months ago, and had fun
'stopping' an AC fan I had in the window. My kids thought it was cool,
particularly when I started using different rates, showing more fan
positions.

It worked quite well, even with cheap LEDs; I built it out of a PIC chip
and a big current mirror, with 10 or so LEDs. The brightness was
adjustable using a pot. I used a fixed 1% duty cycle (adjustable using a
constant in the software), and a variable frequency based on the A/D
converter. Simple software, took about 15 minutes to write.

I'll post the software/schematic if anybody is interested.

I'd throw an eyeball over that...
 
J

John Larkin

Jan 1, 1970
0
If you pulse it properly, I see no reason why not. Look up the pulse spec
on the LED. The pulse width translates into how wide of a smear the timing
mark makes. I've seen backyard mechanics use white "metal-marker" to
emphasize the timing mark on their unit. Not with an LED timing light, but
the principle is the same. The challenging part would be the pulse-forming
network to get the pulse as narrow as possible.

Then again, a one-millisecond pulse is probably not that hard to
accomplish electronically, and it is without question narrow enough to get
your timing within less than a degree.

3600 RPM (we're checking the advance here) spins 360 degrees in 16.6
milliseconds. So a millisec flash is 22 degrees. Even at 1000 rpm, 1
msec is 6 degrees.

John
 
J

John Larkin

Jan 1, 1970
0
Would a white LED, a Luxeon Star for example, be useful as a strobe
light for checking engine timing etc.?
Thanks,
ERS

I wonder if the conversion phosphors change the pulse shape; that well
might slow it down. Gotta try that some time.

John
 
J

Jonathan Kirwan

Jan 1, 1970
0
I wonder if the conversion phosphors change the pulse shape; that well
might slow it down. Gotta try that some time.

Good observation I'd forgotten to consider. I use such phosphors (not these,
but 'such') regularly. Their response times are an exponential decay with a tau
in the roughly 1ms-1.5ms range at typical temperatures (it varies from say 5-8ms
at liquid nitrogen temps to tens or hundreds of microseconds at about 400 C.)

Jon
 
J

John Larkin

Jan 1, 1970
0
Good observation I'd forgotten to consider. I use such phosphors (not these,
but 'such') regularly. Their response times are an exponential decay with a tau
in the roughly 1ms-1.5ms range at typical temperatures (it varies from say 5-8ms
at liquid nitrogen temps to tens or hundreds of microseconds at about 400 C.)

Jon

What do they use for white LEDs... yag powder or something. Phosphors
are interesting; tau's go from under 1 ns to hours.

John
 
R

Rich Grise

Jan 1, 1970
0
3600 RPM (we're checking the advance here) spins 360 degrees in 16.6
milliseconds. So a millisec flash is 22 degrees. Even at 1000 rpm, 1
msec is 6 degrees.

Well, I'll be.

I stand corrected, and will present myself as an object lesson in the
risks of seat-of-the-pants design. :)

Then again, one of the most effective ways to get accurate information on
USENET is to post something obviously wrong. You _will_ be corrected. ;-)

Thanks!
Rich
 
J

Jonathan Kirwan

Jan 1, 1970
0
What do they use for white LEDs... yag powder or something. Phosphors
are interesting; tau's go from under 1 ns to hours.

I don't know, exactly. But the stuff I use often enough comes from GE and
places that make light bulbs for the compact fluorescents and other fluorescent
bulbs used most everywhere. Cheap as dirt and of such extremely high purity and
quality, too. I'm sure that someone spent some time on selecting the right mix
for making white LEDs, but I'll wager it's not a complete home-brew. We do
that, too, and it's difficult to maintain consistency and quality if you are
making small batches -- or large ones, for that matter. So, if they were
'smart' about it, they'd piggy back on what's already being done in large scale
fashion. If so, chances are it's about like what I see. But that's only a wild
ass guess, to be sure.

Jon
 
W

Watson A.Name - \Watt Sun, the Dark Remover\

Jan 1, 1970
0
Robert Monsen said:

It's a big waste of transistors. Replace each transistor on each LED
with a short. The 10 ohm resistors will drop 1 volt at 100 mA, and the
LED forward V drop will be about 4V at that current, so the total will
be about 5V. The LEDs will easily handle that much current at such a
low duty cycle. And the peak light output will be much higher and
brighter. And that's what you're looking for: a lotta light for a short
time, which is what a strobe puts out.

Dr. Flash, Harold Edgerton, used to set a sheet of newspaper on fire
with his strobes. Yeah, not too shabby! They used them in airplanes to
take pics of cities.


 
J

John Larkin

Jan 1, 1970
0
Dr. Flash, Harold Edgerton, used to set a sheet of newspaper on fire
with his strobes. Yeah, not too shabby! They used them in airplanes to
take pics of cities.


There's a newish book of his pics, "Stopping Time." Beautiful stuff.

John

who has a krytron on the shelf over there.
 
R

Robert Monsen

Jan 1, 1970
0
Watson said:
It's a big waste of transistors. Replace each transistor on each LED
with a short. The 10 ohm resistors will drop 1 volt at 100 mA, and the
LED forward V drop will be about 4V at that current, so the total will
be about 5V. The LEDs will easily handle that much current at such a
low duty cycle. And the peak light output will be much higher and
brighter. And that's what you're looking for: a lotta light for a short
time, which is what a strobe puts out.

Thanks for the analysis. Your comments are as sagacious as usual.

The transistors are used to control the flash from the PIC. Without the
transistors, you don't have a flash. I guess you could use a big power
transistor to control the 10 LEDs, but this was cheaper for me, since I
had them laying around. They can easily handle the 20mA to 100mA per flash.

The current through the transistors during a flash is controlled by the
pot, which enables the brightness to be changed between very dim to
fairly bright without toasting the LEDs.

I bought a bag of those PNP transistor on ebay recently, so I have lots
of them. They were a 3 cents apiece, but I'm guessing you could get them
cheaper. They are similar in rating to 2N4403s.
Dr. Flash, Harold Edgerton, used to set a sheet of newspaper on fire
with his strobes. Yeah, not too shabby! They used them in airplanes to
take pics of cities.

There was an exhibit of his work at the SF MOMA a decade ago. I wonder
if it's still there?

--
Regards,
Robert Monsen

"Your Highness, I have no need of this hypothesis."
- Pierre Laplace (1749-1827), to Napoleon,
on why his works on celestial mechanics make no mention of God.
 
W

Watson A.Name - \Watt Sun, the Dark Remover\

Jan 1, 1970
0
Robert Monsen said:
Thanks for the analysis. Your comments are as sagacious as usual.

The transistors are used to control the flash from the PIC. Without the
transistors, you don't have a flash. I guess you could use a big power
transistor to control the 10 LEDs, but this was cheaper for me, since I
had them laying around. They can easily handle the 20mA to 100mA per flash.

The current through the transistors during a flash is controlled by the
pot, which enables the brightness to be changed between very dim to
fairly bright without toasting the LEDs.

I bought a bag of those PNP transistor on ebay recently, so I have lots
of them. They were a 3 cents apiece, but I'm guessing you could get them
cheaper. They are similar in rating to 2N4403s.

Are you sure that the numbers on those transistors are 2N3566? My old
Motorola manual shows the 2N3566 as being a silicon NPN, similar to the
MPS6514.

Yeah, I forgot to mention that removing the ten transistors requires
adding a single high current transistor to switch all the LEDs on and
off. A suitable NPN choice might be the 2SD965 or NTE11, which is
capable of switching up to 5 amps, in just a little TO-92 package. The
complement to that is the NTE12, but I don't know what the 2N or 2S
equivalent is. You can often find the 2SD965s in disposable cameras,
which often use them in the flash charging unit. I made the mistake of
not discharging the 150uF 330VDC capacitor, and got a really good 'bite'
from it when I touched it! That was even tho the camera had been off
and not used for more than a week.

BTW, some guy put an experiment on his website where he took a
dispoosable camera flash and fed it with a power supply, and triggered
it with repetitive pulses. The rep rate was much faster than normal,
and the strobe got the reflector so hot that it started to melt. He
came to the conclusion that if you try to convert a disposable camera
flash to a strobe, the charging capacitor should be a much smaller
value, just a few uF or so.

For those just tuning in, the schematis is at
http://home.comcast.net/~rcmonsen/strobe/circuit.gif
 
C

colin

Jan 1, 1970
0
Eric R Snow said:
Would a white LED, a Luxeon Star for example, be useful as a strobe
light for checking engine timing etc.?
Thanks,
ERS

ive made very simple led strobes ages ago using the old red leds and a 555
timer, the discharge pin can sink quite a lot of curent and drive a led
directly, or an emiter folower boosts it even more, the LED curent resets
the voltage on the timing capacitor, peak curent is limited by the 555 and
or LED resistance etc, the average LED curent is controled by suplying a
constant curent to the whole circuit, leds can be put in series if you have
a high enough coltage supply but the lower threshold voltage (1/3 supply)
needs to be higher than the voltage drop of the leds of course, or u can use
a level shifter.

its not teribly good with old leds but newer ones are so very much brighter
i havnt tried with those but might do out of interest now.

it all depends how dark your environment is, ive found cheap neon engine
strobes to be virtualy usless unless u can use them in a garage wich u can
make dark or use at night of course, or unless your timing mark is in a very
convenient place.

Colin =^.^=
 
E

Eric R Snow

Jan 1, 1970
0
I built a strobe out of white LEDs a couple of months ago, and had fun
'stopping' an AC fan I had in the window. My kids thought it was cool,
particularly when I started using different rates, showing more fan
positions.

It worked quite well, even with cheap LEDs; I built it out of a PIC chip
and a big current mirror, with 10 or so LEDs. The brightness was
adjustable using a pot. I used a fixed 1% duty cycle (adjustable using a
constant in the software), and a variable frequency based on the A/D
converter. Simple software, took about 15 minutes to write.

I'll post the software/schematic if anybody is interested.
Greetings Robert,
I for one, am interested. Thanks.
Eric R Snow
 
R

Robert Monsen

Jan 1, 1970
0
Watson said:
Are you sure that the numbers on those transistors are 2N3566? My old
Motorola manual shows the 2N3566 as being a silicon NPN, similar to the
MPS6514.

When I typed in the model, I apparently transposed the first two digits;
its 2N5366. http://www.fairchildsemi.com/ds/2N/2N5366.pdf

--
Regards,
Robert Monsen

"Your Highness, I have no need of this hypothesis."
- Pierre Laplace (1749-1827), to Napoleon,
on why his works on celestial mechanics make no mention of God.
 
Top