Hi everyone,

I'm taking up a new project on behalf of my father, an LED stroboscope. Most of my knowledge of LED's is topical, but I am aware that LED drivers exist. I know some of them have integrated control elements such as dimming. Do they offer any other advantages, like maybe turn-on time? Would they outperform say a FET with a gate driver? I realize these are broad questions, just looking for the jist.

Thanks for reading.

 

It is called PWM (Pulse-Width Modulation).
You can use a 555-timer circuit and drive the LED with a series resistor. NE555 or LM555 can supply up to 200mA.

 

For a stroboscope, because the pulses are very short, the light needs to be VERY bright because our eyes basically "average out" the light levels. That's why commercial stroboscopes use a Xenon flash like a camera. You certainly could make a stroboscope with LEDs but you will probably be disappointed with its performance. I made a stroboscope a few years ago out of an automotive timing light (which is basically already a stroboscope) that might give you some ideas. Here's a video of it:

 

You did not say the maximum number of flashes per second.
You did not say the required brightness.

 

A single 30-Amp-rated FET-Driver, such as a IXDN630MCI,
can drive ~10 or more small N-FETs simultaneously,
which can switch ~30 or more ~3-Amp LEDs,
( if supplied by an adequate Power-Supply with large Filter-Capacitors ).

The LEDs will need a large Heat-Sink.

The FET-Driver has a CMOS-Input with hysteresis,
so You can drive it with almost anything.

Be careful, this setup can cause permanent Eye damage.
.
.
.

 

just looking for the jist.

The gist is I would try a Google of "LED strobe schematic" you will get some ideas. I wouldn't worry about light brightness because you want bright, real bright. Most circuits you will find are based on MrChips reply using a simple 555 timer chip. I also agree with Strantor:

You certainly could make a stroboscope with LEDs but you will probably be disappointed with its performance.

Years ago I used quite a few Sreobotac units made by Gen Rad (General Radio) which used the old tried and true Xeon flash tubes.

Ron

 

What is the distance to the Target ?
How large is the Target ?
Indoors, with very low ambient Light ?, or Outdoors, with Sunlight ?
.
.
.

 

AAC is the best place on the internet. Thanks everyone for the feedback.

 

It is called PWM (Pulse-Width Modulation).
You can use a 555-timer circuit and drive the LED with a series resistor. NE555 or LM555 can supply up to 200mA.

Thanks for the reply. My Dad suggested I use a 555, I told him I can do it with a 555 for nickel or use a PIC for a dime. The future is now!

 

For a stroboscope, because the pulses are very short, the light needs to be VERY bright because our eyes basically "average out" the light levels. That's why commercial stroboscopes use a Xenon flash like a camera. You certainly could make a stroboscope with LEDs but you will probably be disappointed with its performance. I made a stroboscope a few years ago out of an automotive timing light (which is basically already a stroboscope) that might give you some ideas. Here's a video of it:

Cool project and good advice. The cost of a xenon bulb would tank this project, LED's my only option. I was planning on using indicator LED's but I'll go with something beefier.

 

You did not say the maximum number of flashes per second.
You did not say the required brightness.

Frankly I don't know the brightness I need. Ideally I'd like to get up to say 100k RPM, call it 1.5kHz.

 

A single 30-Amp-rated FET-Driver, such as a IXDN630MCI,
can drive ~10 or more small N-FETs simultaneously,
which can switch ~30 or more ~3-Amp LEDs,
( if supplied by an adequate Power-Supply with large Filter-Capacitors ).

The LEDs will need a large Heat-Sink.

The FET-Driver has a CMOS-Input with hysteresis,
so You can drive it with almost anything.

Be careful, this setup can cause permanent Eye damage.
.
.
.

This is more or less the solution I originally imagined. I'd like to keep the current under 5A as I am hoping to use a couple D cell NiMH batteries. Maybe an array of 10 500mA LEDs.

I think I am beginning to see the light with respect to LED drivers, specifically current based. I have a pretty mean boost IC that I planned to use to step up the cell voltage with current control by limiting resistors, but that's a lot of power down the drain. Going to look into a constant current supply and actuate with a FET. See any holes in this?

 

The gist is I would try a Google of "LED strobe schematic" you will get some ideas. I wouldn't worry about light brightness because you want bright, real bright. Most circuits you will find are based on MrChips reply using a simple 555 timer chip. I also agree with Strantor:

Years ago I used quite a few Sreobotac units made by Gen Rad (General Radio) which used the old tried and true Xeon flash tubes.

Ron

Control will be handled with a microcontroller as there are a few other bells and whistles. As for the LED's, maybe not the best solution for performance but the simplest in terms of power and control. Originally planned to use indicator LED's but obviously they will not suffice.

 

1.5kHz is slow enough.
a battery voltage is not regulated and will keep dropping as it is used.

 

What is the distance to the Target ?
How large is the Target ?
Indoors, with very low ambient Light ?, or Outdoors, with Sunlight ?
.
.
.

Distance will probably be about 3' max, with the target being a surface of up to 4sqft. Indoors with moderate lighting. Industrial setting.

 

"" a surface of up to 4sqft. ""
What does this mean ???

At "~3-feet",
with plenty of Florescent-Lighting around,
You will need some serious Light-Output,
as I described above, ( ~100-Watts or so ),
and a high-contrast Target-Paint-Scheme,
or Reflective-Tape.

A 555 as a Controller is a rather "clunky" solution.
A proper adjustable Pulse-Generator-Circuit will be far more satisfactory.

Do You need to accurately measure RPM ?
.
.
.

 

Years ago I used quite a few Sreobotac units made by Gen Rad (General Radio) which used the old tried and true Xeon flash tubes.

The best device ever! If you know how to read it properly. During turbine generator testing and overspeed-trip adjusting on (undisclosed ship) at (undisclosed shipyard), a mistake was made using a strobotach. A stobotach will make a machine shaft look stationary when the flash rate matches the rotation speed. Or the flash rate is twice the speed. Or half the speed. There are simple techniques to verify you're reading the actual speed and not a harmonic or sub-harmonic of rotation speed.

On the particular turbine under test, it was found that the overspeed trip was way out of adjustment, tripping at about half the expected speed. This particular style of overspeed trip mechanism uses shims to set the speed, with about 1/8" of shimming typically required to dial the speed in.

Shims were added. More shims were added. Over an inch of shims were added and still the overspend was tripping too soon.

(Anybody who was paying attention would have figured out the problem well before this point).

Everything came to a screeching halt when the generator rotor expanding so much it contacted the stator and put the brakes on. The 150 pound overspeed trip mechanism broke its shaft and was launched at nearly 20,000 rpm. Amazingly, no one was injured as it made a couple laps around the engine room.

Detailed analysis agreed that the turbine was running at nearly twice its normal operating speed when it failed – pretty tough machine! All because the guy reading the strobotach didn't know how to read it properly. And all the machinists didn't raise a red flag when things just didn't make sense.

 

Cool project and good advice. The cost of a xenon bulb would tank this project, LED's my only option. I was planning on using indicator LED's but I'll go with something beefier.

The kind of LEDs you would need, I think would cost more than this: https://a.co/d/fYjbp9p

 

The best device ever! If you know how to read it properly. During turbine generator testing and overspeed-trip adjusting on (undisclosed ship) at (undisclosed shipyard), a mistake was made using a strobotach. A stobotach will make a machine shaft look stationary when the flash rate matches the rotation speed. Or the flash rate is twice the speed. Or half the speed. There are simple techniques to verify you're reading the actual speed and not a harmonic or sub-harmonic of rotation speed.

On the particular turbine under test, it was found that the overspeed trip was way out of adjustment, tripping at about half the expected speed. This particular style of overspeed trip mechanism uses shims to set the speed, with about 1/8" of shimming typically required to dial the speed in.

Shims were added. More shims were added. Over an inch of shims were added and still the overspend was tripping too soon.

(Anybody who was paying attention would have figured out the problem well before this point).

Everything came to a screeching halt when the generator rotor expanding so much it contacted the stator and put the brakes on. The 150 pound overspeed trip mechanism broke its shaft and was launched at nearly 20,000 rpm. Amazingly, no one was injured as it made a couple laps around the engine room.

Detailed analysis agreed that the turbine was running at nearly twice its normal operating speed when it failed – pretty tough machine! All because the guy reading the strobotach didn't know how to read it properly. And all the machinists didn't raise a red flag when things just didn't make sense.

Yeah, I played a similar game. Generator and generator control unit for the F15 screaming eagle fighters. The generator was spinning at 10,000 RPM off a constant speed drive. Initial testing was done in a bunker because you don't want to be anywhere near one that goes out of balance and literally disintegrates. Those Gen Rad 1531 units were all over the place and some of the best strobes I ever worked with. Gen Rad made a few flavors of them.

Ron

 

Just some thoughts (from an old article from the time when white LEDs were getting bright enough for some serious uses).

Finally, white LED's are bright enough to use in a practical stroboscope. This circuit can operate as a bench-top stroboscope that, in conjunction with an oscilloscope or frequency meter and bench top power supply can accurately measure rotational speeds, or it can be operated hand-held from a single 9V battery, using a marked scale to estimate rotational speeds. Its a nice iinstrument to have on or around the workbench. Besides that, its fun to use.


(Above) The stroboscope freezes a piece of white tape on a fan blade that is rotating at 2616 RPM. Power is supplied and the strobe frequency is monitored via the connector in the lower right corner.

Overview
Its bright enough to use without dimming the room lights and with the aid of a frequency meter or oscilloscope, can measure rotational frequency to within a few degrees per second. This LED stroboscope is a fairly simple circuit for the similarly simple task of measuring the speed of things that rotate. I wondered how fast some DC fans I had on the bench were, and tried a quick-and-dirty LED driver, driven by pulses from a function generator. I was pleased to see that, in contrast to my earlier experience years before, the white LEDs I have now, BUXC333 from BestHongKong.com, is bright enough to use with the room lights on. I decided to go ahead and build up a small stroboscope with one of them.

The circuit and an optional 9 volt radio battery are housed in a small plastic box. A range switch selects the RPM range, and a knob adjust the oscillator frequency. A white LED is mounted at the end of about 60 cm of wire, so it can be aimed at whatever it is that I want to illuminate with it, without affecting the orientation of the box.

When operating in the hand-held mode from its internal battery, the pot is adjusted so that the LED flashes once per rotation of whatever is being measured, and the revolutions per minute are estimated by reading from the hand calibrated dial.

A headphone jack mounted in the box provides connections for an external power supply, such as a bench supply or a "wall wart" power supply and an pulse output so that the LED pulse frequency can be accurately measured with a frequency meter or oscilloscope.

A note of caution: Be careful not to load the pulse output, since its not buffered and loading it can stop the oscillator.


Circuit

((Above) The comparator on the left is the oscillator. The comparator on the right makes a
constant duty cycle pulse by comparing the sawtooth generated across the oscillator
capacitor, Ct, to the DC level set by the 50k pot.


The circuit is composed of an oscillator, a pulse slicer, and a gated constant current source. All of the circuits except the anode of the LED receives power from an LM78L05 5 volt regulator.

The comparitor connected to pins 1,2, and 3 is connected as a 50% duty cycle oscillator. This circuit's oscillation frequency is = 1/(1.4 R C) where R is the sum of of the 10k fixed resistor and the 100k pot, and C is the capacitance from pin 2 of the comparitor to ground. Multiplying the frequency in Hertz by 60 gives RPM (Revolutions Per Minute), so the formula becomes RPM= 43/(R C). Using this formula, you can choose a value for the capacitor that will center the oscillator's range around the frequency you are most interested in.

The second comparitor, the one using pins 5,6, and 7, slices the sawtooth that appears across the oscillator capacitor. Whenever the voltage of the sawtooth exceeds the voltage on the wiper of the 50k pot, the comparitor's output is high. When the comparitor's output is high, the constant current source made of the two transistors is on, and the LED emits light.


A 2N4401 or similar should work as well as a 2N2222.
Both have the same pinouts though the 2N2222 can handle more current.

Whenever there is current available through the 3.3k resistor, the upper 2N2222's current is limited to that which provides enough voltage across the 10 Ohm emitter resistor to turn on the lower 2N2222. This voltage is about 600 millivolts, so current through this resistor, which is approximately the same as that of the LED is set to about 60 milliamps. That's twice the datasheet's 30 milliamps maximum current, but well less than the 100 mailliamps peak pulse current. To reduce the current, make the emitter resistor larger. To increase the current, make the resistor smaller.

Assembly and Calibration



(Above) all of the circuitry, plus the optional 9 volt batter fits snugly in a 5.5 cm x 6.5 cm plastic box. The red/black wire exiting the hole in the box cover connects the LED to the circuitry.

I had a pretty good idea of what I was going to use as a circuit, so I went ahead and modified the box - cutting the holes and mounting the frequency control pot, the on/off switch, range switch, and I/O jack, and cutting the circuit board to size before prototyping the circuit. If I had built the circuit first, I think I would have used a larger box. The small parts were mounted on a small piece of pad-per-hole phenolic circuit board and the leads were bent toward one another and soldered together. This layout is not critical, but you should keep the leads to pins 2 and 3 of the comparitor as short as practical since noise pickup on these leads can affect the oscillator.

After you've got it all connected together test the circuit using an inexpensive red LED rather than an expensive white LED because if the constant current source doesn't work, or if the duty cycle pot is turned the wrong way, you run the risk of "frying" the LED.

Set the duty cycle pot by aiming the LED at a rotating target, preferably one with a timing mark of some kind on it, like the small fan blade with a piece of white tape shown in the photograph at the top of this page, and adjust the duty cycle pot for the best tradeoff between apparent brightness and sharpness. A large duty cycle will smear the image. A very short duty cycle will look dim. The tradeoff is yours to make.

You can also use an oscilloscope to set duty cycle by probing pin 7 of the comparitor and adjusting the positive going pule to the desired duty cycle. I recommend 10% as a starting point.

You can calibrate the frequency adjust knob if you want. I put a piece of white cardboard on the front of the box and drew a scale on it with a ball-point pen. Using a frequency meter to monitor the oscillator output of the headphone jack, I made marks at dial positions corresponding to 500, 750, and 1000 RPM on the slow scale, and at 300, 3000, 4000, 5000, and 6,000 RPM on the fast scale. With this scale, I can estimate revolutions per minute when operating in the battery powered hand held mode.


UV Strobe
Given that this circuit can drive LEDs with a forward voltage drop of up to about 1 volt less than the battery voltage, it will also drive an Ultra Violet (UV) LED. An ultraviolet stobeoscope in conjunction with florescent paint might be more useful in a room with the lights on than a while LED stroboscope.

I tried something similar using my one and only UV LED and luminescent paint, and found it to work in principle. I painted a dot of luminescent paint onto an electric fan and turned it on. The paint glows yellowish-green for a long time after being exposed to the UV. When I have the UV strobe timing set correctly, a bright yellowis-green ring appears, tracing the orbit of the luminescent dot. The ring glows and fades as the phase of the fan drifts with respect to the phase of the LED strobe. With this setup, one could find the rotational frequency of the target by adjusting the strobe frequency for maximum brightness of the glowing ring.

Printed Circuit Board Version
Guillermo Cardarelli, of Argentina, made a fine looking strobe with the beautiful printed circuit board in the pictures below.


The component side of Guillermo's board is shown above.


Above is the copper side of Guillermo's board.


Guillermo used external pots to control both the frequency and the duty cycle. Other wires
lead to the on/off switch, the fast/slow switch, and a connector for the 9 volt battery.


Guillermo's stobe (above) used the same circuit as the origingal single LED stoboscope,
but it drives seven LEDs in parallel. This gave him the same amount of light as the single LED strobe,
but with the advantage of the illumination being more even.