I have a variable strobe which uses two bright leds in series driven by an oscillator and a BC547. This is quite effective and will strobe at just over 60 Hz. Could anyone give me an idea of the maximum on/off strobe frequency possible from LEDs? I do notice that this strobe smears the disk sector much more than a flash strobe so that would perhaps point to a not very sharp switching of the LED but I dont know if this is a shortcoming of the driving circuit or the LED characteristics, which might in turn show a definite limit to practical RPM possible to detect even if the oscillator ran faster. Would HiBright single colour LED be better in this application, perhaps the phosphors in white LED have a slow rise/fall character? Charlie
Remote controls for TVs etc. usually operate at 38kHz. These are IR LEDs, but the principle should be the same. I have seen music transmitted through a red LED, and I have transmitted 9600bps unmodulated data through a red LED, so 60Hz is probably well within a LEDs capabilities.
Smearing is probably caused by the LEDs on-time being too long. You need shorter blinks. Of course, this will result in less visual brightness, so you will have to compensate by using a brighter LED.
Very much faster than 60Hz. I've got a small circuit nearby operating them around 80kHz, easily.
It's really hard to say. A strobe is often a capacitive discharge triggered by an SCR mechanism and may have a very small duty cycle. I've no idea what your LED duty cycle is or the rates you are using for each during the above observation. But your issue likely is about duty cycle here, my guess.
I don't know much about phosphors used with white LEDs. I'm kind of familiar with some phosphors that I use, though, which are used for temperature measurement and are selected for longer tau periods and single exponential decay modes. I think Don knows quite a bit about it, though. Perhaps he will correct some of what I'll write, now.
The term 'phosphor' is old and well predates modern knowledge about mechanisms. In broad usage, the term covers several different mechanisms that happen to have some similarities to an observer. Mostly, just about anything illustrating a Stokes Shift qualifies (shift to lower wavelength.) Just to make it more confusing, anti-Stokes Shifts are probably included. So 'phosphor' means just about anything than can cause any noticeable change in wavelength.
LEDs probably use faster phosphors that depend on fluorescence. Fluorescence is caused by absorption (which is VERY FAST -- on the order of 10^-14 second, or so) -- and then often some vibrational relaxation (thermal relaxation), which can take place on the order of 10^-12 seconds, or perhaps an order of magnitude slower or faster in solids and liquids. Fluorescence involves both, plus some time for emission probability to play out (which is also fast), so I usually just keep in mind that fluorescence is a roughly nanosecond phenomenon. (Singlet state, all totalled, works out to on order of 10^-8 second.) About 20ns covers most cases likely to be seen, though some are ten times longer.
Phosphorescence is on the order of 1/10th of a millisecond to tens of milliseconds... and rarely into seconds. It's much slower, because it depends on intersystem crossing between singlet and triplet states, which are very much less probable and therefore appear to be slower when they do happen. Triplet state lifetimes are on the order of 10^-4 second, or so. Because of how long it takes, in liquids these states usually get de-energized as heat through internal conversion and aren't visible. But in solids, like ceramics, they can be quite visible. I just don't think LEDs depend on phosphorescent modes, though.
There's another method, delayed fluorescence, with times that fall between the two, perhaps on the order of 10's to 100's of microseconds, though I think various dopants greatly affect all that and widen the range quite a bit. Anyway, I don't know much about it, except that it seems to involve two triplet states (biphotonic.)
If I'm right that LEDs probably depend on fast fluorescence, the tens of nanoseconds required by them shouldn't be an issue for you.
LEDs (especially the cheap ones) are a lot dimmer than xenon flash tubes, to compensate for this the LEDs in your strobe unit are driven with a longer on-time than the flash tube would have
as a result the disk moves further whilst the led is lit.
that could be a factor too, but if present would show up as blue/yellow fringing on the disk. I think flourescent dyes are used instead of phosphors, and I've never noticed a phosphor effect with them.
Yup, I use red LED's to check the speed of ~MHz photodiode TIA's.
I hit the LED with a step and look at the PD response. The IR LEDs are a lot slower. There must be some physics/ solid state reason for this but I don't know what it is.
Next time I'm testing PD's I'll look at them with some white LEDs and see if there is any sign of a longer turn-off time.
To the OP how much current are you sending through the LEDs? As others have said it sounds like you have a long pulse time. More current for a shorter time may help.
You most likely have a saturation problem in the LED from your driving circuit, the turn off time is more noticeable at higher rates. Also, you maybe not be sinking the off cycle which will cause a little off delay.
I suppose one could use a feed back PD (Photo Diode) to throttle the current feed when the LED is actually generating light. This will result in a faster turn off because you won't have the LED in or over saturation.
Not that i've ever had a need to resolve an issue of this type, it does sound logical.
Saturation? What an odd notion that seems to me? I think you might have a slower turn off in white LEDs due to the phosphor, the answer then would be to go monochrome LEDs. No phosphor lifetime or afterglow or whatever the effect is called, to cause a smear. Use RGB LED if you need white?
On Sep 2, 12:19=A0am, Charlie+ wrote: > I have a variable strobe which uses two bright leds in series driven by an > oscillator and a BC547. =A0This is quite effective and will strobe at just > over 60 Hz. =A0Could anyone give me an idea of the maximum on/off strobe > frequency possible from LEDs? > I do notice that this strobe smears the disk sector much more than a flash > strobe so that would perhaps =A0point to a not very sharp switching of the LED > but I dont know if this is a shortcoming of the driving circuit or the LED > characteristics, which might in turn show a definite limit to practical RPM > possible to detect even if the oscillator ran faster. =A0Would HiBright single > colour LED be better in this application, perhaps the phosphors in white LED > have a slow rise/fall character? > Charlie
When I was designing the photo sensor preamps for an HDTV telecine, I built a driver to work with HP super bright LEDs that had 10 nS turn on / off times used as a preamp test for the photo sensors. The LED light source driven by a Burr Brown transconductance amp was a dB down at 30 MHz and quite linear from off to 10 mA drive. A 30 kHz (line rate) linearity ramp was in fact linear. Garden variety LEDs from Radio Shack aren't anywhere near as fast but could do audio.
ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here.
All logos and trade names are the property of their respective owners.