Ultra Short, Ultra Fast LED Flash

[Marco]

Hmmmm, my first impression is that that’s beginning to touch on the realm of  impractical.

That's only if you want to get that bright, you can make do with a lot less light. Look at the Vela One, I'd estimate they are doing closer to 5 kW, 200-300 Watts worth of COBs.

[StillTrying]

The stored energy of approximately 8 J is released in a short flash of less than 500 ns.

16 MW.

[JAndrew]

Huh....ok, so I'm not completely out to lunch... so they have something pretty neat but it's unobtainable. I think I remember seeing this a few years ago, and it looked great but it was not available then either.

So from the description they are using 9 5000lm COB LEDs over driven by 2000% to produce 1mil lumen. This is all driven off a four AA batteries.

What sort of wizardry do they have going on?

[JAndrew]

Looking around for LED COBs

https://www.ebay.com/itm/100W-Cold-White-High-Power-9000-10000LM-LED-light-Lamp-COB-Chip-35DI-/391802735671?_trksid=p2349526.m4383.l4275.c1#viTabs_0

Pretty inexpensive  for 10,000LM, but if I drive these by 1000% and get 10 of them. Then I could match the 1mil Lum. However the power requirement would be pretty high as well. Input current under normal operation is 3000mA and voltage is 32-34vdc.

This isn't something I could drive with 4 AA's.

[ogden]

Pretty inexpensive  for 10,000LM, but if I drive these by 1000% and get 10 of them. Then I could match the 1mil Lum.

10K lm? - Scam. With such specs he can put OSRAM out of business, yet they are doing well You may indeed user those to start with, but do not expect to get even close to specified light output.

Input current under normal operation is 3000mA and voltage is 32-34vdc.
This isn't something I could drive with 4 AA's.

It is easily achievable with 4AA's - in pulse, from charged flash capacitor. After all this is how Xenon flashes/strobes work for decades and those guys are doing exactly that: https://www.vela.io/vela-one-high-speed-flash

[JAndrew]

If that is the case how hard would it be to adapt a typical flash camera circuit, to drive an LED?




Sent from my iPhone using Tapatalk

[ogden]

If that is the case how hard would it be to adapt a typical flash camera circuit, to drive an LED?

In theory - yes. You may replace xenon tube with (250V...300V) LED string, modify trigger circuit. I am not sure about power capability - existing capacitor or bank will be enough or you need to add extra (capacitance). [edit] How hard - it depends on flash, it's circuit complexity.

[Marco]

The new'ish 12V COBs seem to have better construction than those older ones.

[JAndrew]

Trying not to reinvent the wheel, so I'm looking at flash circuit kits that I might just adapt to overdrive an LED.

Here's one I am looking at.

https://www.xenonflashtubes.com/flash-driver-modules/51-5v-optical-flash-trigger-photography-kit-diy_131.html#/30-flash_energy-10j_ws

I'm interested mainly in just the charge and trigger.  Looks like it can be trigger pretty simply, and It's in expensive enough that I could set up a few units to trigger simultaneously.  I suspect the 1000uF 450v isn't enough and will likely need to replace it with something larger.

I'll couple it with this 100w LED on COB that is supposedly rated for 9,000lm @ 28-36vdc@3000ma

eBay auction: #https://www.ebay.com/itm/10W-20W-30W-50W-70W-100W-High-Power-LED-Chip-COB-SMD-Bulb-DIY-Lamp-Light/263961952554?var=&hash=item3d755bc12a&enc=AQADAAAC8FjVrDbVsZ8oH%2F8PNHtt9VX4%2Fw7FZcmMuqsX8uaFEduV1JjTNcqxLENQ%2BAoiD9Vd%2Fo0t6JaYcgeiBJwDWg916eTMBW0ftEcw49m0a6WTi8bnXqMs1rT8%2FhEPw2EnUesyxZ4yaCmm3nWNnCBC77obZ03VtmDnTIEKS9Ra8EDyFGcD7BFW9%2BgJGn%2FX6rG1%2B%2F9YdGQAzZlDbBJ1KwsnmICW4EUSMkH4GvW%2BKjbiczw%2BnjyxtaHSdG9uBPK2AQHTlF9t6GJBJy19FBbuaz6d3Zneu2yCtAyqRKyR70YKKgJHYejwmZ7ID0t84YQJ8IhtgciwaCYRHvux4bD4%2B4UyzSPsDxVVVj8Hkp0nwaOKSUB383V29nAKRqyx6YYHi2wj%2FbLTYOsukafpYjDurl26JoRFx%2BnDIR12YoaoYkZvRUzQKvd%2Bg3L%2BQVGhqwQPrLGnpDiaDpxC6wnVJt95niIQ7ziF15W4hAKdaNra1Crqjdn2yEvljBVDCMP9GqL5IBxu08zjbPeeYTykPBuc3wXgVuwhXfqtV1cGCZr0MfxlOjRePdUwRXlNxOuuaH5Qps9lN1vZP4OuHqhBP%2BiaeiUc8o7NbyhS2bIpjagJX6xsc0WQRVLm4j2G6KHpCyICn7yBEwF3%2FuHWNFQPY3UhPf486TYHTxPeW5Y2dtGAI0oDlLt7Mf%2B45njYlojDp4VpRXER%2BkpRQiTJgiPzms2SLZnCOnVTVeBX0XlgFk41SY73Nujg3lAY9KkrpuAOyOtjwVJRH%2FvQR9ZwXTmswdeLyPa%2FBfUa7aznGxThaMRHONOhyZZYiotqe5%2Fjizn21%2BFlMxHaHZTiHVpgCDD1zw2pF9ZGY27SRDDllEVoa1d3Gzd0EPeV2jJ8A%2FweAWzGG5fTA3jiQEqXV42WQr9fNzu5kVWcHaEmQkKGEtChwYJm%2F13TJG7S6oJw7RbqoRUlxyPe6ZGsQnkINsEKSPQ3M%2Ft69st%2Bgs2ochwkyH0bU4ZsfMGUwzeLvMnA&checksum=26396195255498a2b9072f3042a086822bfd2b40e7f2

From there I should be able to use the photo transistor in the circuit to measure the flash duration. At this point I'm into the project for under $40 and should be able to prove out the concept

[Marco]

Just drive them with a MOSFET, the capacitive discharge method is for xenon tubes ... it's a bit silly for a LED.

[ogden]

I'm interested mainly in just the charge and trigger.

Only thing you need from that flash is charge stored in the capacitor. Xenon flash trigger is no use anyway, LED needs high voltage transistor driven from high power gate driver (refer to 4ns LED flash circuit posted here). You may even leave that flash thing unmodified and operable if needed, just tap capacitor for your LED project. COB's you mention looks good as well (for starters), hopefully their bonding wires will not explode when overdriven. Flash voltage is specified 310V, so you need ~4 COB LED's in series for 200% power, then you "improve" it further until it explodes.

Be careful. Hopefully you know how to handle high voltages, 1000uF@300V (~50 joules of energy) is deadly.

Just drive them with a MOSFET, the capacitive discharge method is for xenon tubes ... it's a bit silly for a LED.

Right. LED shall be connected to capacitor and after 500ns transistor shall be switched off. You do not leave LED after trigger connected to cap forever, such way it is impossible to get short pulses.

[Marco]

That's why the 12V COBs are nice, if you use the 70W version which Big Clive reviewed then 36V should get you roundabout 20x the current. Nice low voltage.

You might need one driver+MOSFET per COB, but the voltage stay safe.

[JAndrew]

I wanna make sure I'm understanding the idea.

We use the charging circuit, but the rest of the circuit isn't needed since we're not driving a Xenon flash. Instead we tap in a MOSFET rated for medium voltage, and trigger the MOSFET (in this case we have a logic level MOSFET driving the larger MOSFET) to flash the LEDs.

(refer to 4ns LED flash circuit posted here

Which one was this one exactly?

Hopefully you know how to handle high voltages, 1000uF@300V (~50 joules of energy) is deadly.

I was hoping to stay away from high energy sources. It looks like it's going to be unavoidable. I'll be brushing up on the safety aspects of this project before embarking on it.

[Marco]

You need high energy, but you don't need high voltage.

[ogden]

I wanna make sure I'm understanding the idea.

We use the charging circuit, but the rest of the circuit isn't needed since we're not driving a Xenon flash. Instead we tap in a MOSFET rated for medium voltage, and trigger the MOSFET (in this case we have a logic level MOSFET driving the larger MOSFET) to flash the LEDs.

You use MOSFET driver IC which have high current push-pull drive capability. Single mosfet driving other mosfet's gate will not be fast enough. You need both - *fast* turn-on and turn-off of the main MOSFET. Sorry, I did mean 4us flash, not 4ns. There's circuit in this post:

https://www.eevblog.com/forum/beginners/4-microsecond-high-power-pulses-through-led/msg1417075/#msg1417075

I was hoping to stay away from high energy sources. It looks like it's going to be unavoidable.

High voltage is avoidable, you just need bigger wires, capacitors and transistors because currents are higher As Marco suggested - to avoid high voltage you may use 12V COB's in parallel, then you mau get away with safe <50VDC voltage. Which way to go - up to you. We (forum members) are not your parents.

[JAndrew]

I appreciate all the input.

https://www.eevblog.com/forum/beginners/4-microsecond-high-power-pulses-through-led/msg1417075/#msg1417075

Definitely a worthwhile read through. The schematics he posts at the end to show his completed projected aren't in a format that I am used to reading.

EDIT:

I've studied his schematics and everything he's doing makes sense. I am assuming the length of the pulse is being controlled by the software on the MC.

I get it now on the MOSFET Driver, I haven't used one before, but it makes sense. All looks very doable. I wish he posted an update on the how the project turned out. I checked out his YouTube channel but that video looked to be the last he had posted on the topic.

[Zero999]

I'd like to try and take photos like this on my own. I could go out and purchase the commercial unit for $1000+ or I figured I'd see if I could build it using RGB LED's. The reason I'm using RGB LED's is the phosphorus layer on a "White" LED Glows for a unspecified amount of time after the blue LED shines on it. This may extend the length of the flash beyond my target of 500ns.

You mean phosphor, not phosphorous, which is flammable.

One problem I've had with over-driving LEDs is colour shift, especially with green LEDs. I haven't noticed it with red and didn't test blue, but green LEDs shift towards the blue end of the spectrum, overdriven. This effect is clearly visible, when I tried drivinga green LED at ten times the rated current, but at a low duty cycle its colour drifted from a bright green to turquoise, almost cyan hue. It wasn't a problem with any of the phosphor LEDs I tried.

Some phosphor LEDs have a significant afterglow, others don't. The green phosphor LEDs I tested were very fast. No measurable afterglow, with instant on and off times. The yellow ones were the worst, with a slow on time and lots of afterglow. If I get the time, I'll update with part numbers and an oscillogram.

See the paper linked below about LED pulsed power. At the start of my tests I used a similar circuit, with the MOSFET driving the low side and the current sensor in series with the source, but I abandoned it in favour of using transformers for both driving the LED and current measurement, as the galvanically isolating the test equipment gave better, more reliable readings.
https://core.ac.uk/download/pdf/30991250.pdf

[Marco]

I've studied his schematics and everything he's doing makes sense. I am assuming the length of the pulse is being controlled by the software on the MC.

Microcontroller control only really makes sense if the micro triggers both the chain of events you want to photograph as well as the flash.

If you are say pulling the trigger on a gun with a string it makes much more sense to just keep everything analogue and trigger on sound or a light gate, no micro necessary.

[Sceadwian]

Is this even possible with white LEDs? I thought the phosphor delay was WAY more than the nanosecond range?

[Zero999]

Is this even possible with white LEDs? I thought the phosphor delay was WAY more than the nanosecond range?

It depends on how fast and the LED. I posted an oscillogram of the CMA3090 being pulsed at 1µs here. As you can see there some delay in the on and off time, but not much. As I said in my previous post, some phosphor LEDs are worse, others are better.

What's odd about phosphor based LEDs is, although the initial decay is fairly fast (even the slow yellow one I tested had a time constant in the µs range) a very faint afterglow persists for a long time after the LED is turned off. This is no problem for a photoflash application, because it's several orders of magnitude too small to cause a problem: it's just about detectable, with dark adapted eyes, in a very dark room. I've seen it for myself. I ruled out persistence of vision and charged capacitors by closing my eyes, exposing the LED to a backlight (UVA) whilst my eyes were still closed, turning off the UV source and looking at the faint afterglow from the LED. I have only done this with white LEDs and don't have the equipment to make quantitive tests, otherwise I would.

[JAndrew]

I've done this before. The requirement was for a 1µs. Phosphor LEDs are variable. Some are fast and others are slow.

Attached is a schematic of the test set-up I used to drive LEDs with 1µs pulses of hundreds of amps, hence the low value current limiting resistor. It's possible to overdrive LEDs by a factor of 10, for short pulse lengths. Note transformers are used to drive the MOSFET and measure the current, to avoid large currents passing through the grounds of the test equipment.

I'm working on digesting all the idea's that have been suggested here. Zero999, I'm just making sure my flea brain is understanding the diagram,  you use an inductor to drive the LEDs rather then using a Flash Capacitor.

I appreciate the lead on the LED you were testing. Vela mentions they spent a lot of time trying to find the right LED's for the job. I suspect this gets back to the issue of how long the phosphor stays excited. I'm fairly confident they are using white LED's, the images of the unit show the LED's and the yellowish tint characteristic of white LED's is clearly visible.  This would simplify the circuit, as we wouldn't need to worry about white balancing the RGB LED's. Since LED's MFG's aren't typically concerned about the switching time of a White LED, I can image there is a lot of variation, possibly even from lot to lot.

I've studied his schematics and everything he's doing makes sense. I am assuming the length of the pulse is being controlled by the software on the MC.

Microcontroller control only really makes sense if the micro triggers both the chain of events you want to photograph as well as the flash.

If you are say pulling the trigger on a gun with a string it makes much more sense to just keep everything analogue and trigger on sound or a light gate, no micro necessary.

As far as triggers go, I was planning on using an IR optical sensor. I've put one together before, and have had fairly good success with it. As the bullet passes by the "shadow" creates a signal. The trouble with using this directly is that the duration of this pulse is dependent on bullet velocity and bullet length. Fast short bullets would create pulses that would likely be far less then 500ns, while long slow bullets would create a pulse that would last far longer then 500ns. (These type of sensors are commonly used in high end chronographs)

So we would want the MC to trigger on the pulse, with a programmable delay (just a few microseconds) before sending the trigger for the flash.  By having a short delay, we can dial in the position of the bullet in the frame without physically moving the set up. I know I've tried to use an Ardiuno to create a short 500ns square wave but I seem to remember that it struggled. I don't think the code was fast enough to switch the I/O pin that quickly.  It will be another problem I'll need to work through.


To Marco's comment of using low voltage and high amps, I was thinking on that last night. A 10s (37v) Lipo battery rated at 4000mAh could provide a pretty good jolt of energy across an LED. With the duration being so low, I don't think there would be any lasting damage to the Lipo and it would be good for quite some time. There may be other reasons that this wouldn't work, but I was brainstorming.

[ogden]

I've studied his schematics and everything he's doing makes sense. I am assuming the length of the pulse is being controlled by the software on the MC.

Microcontroller control only really makes sense if the micro triggers both the chain of events you want to photograph as well as the flash.

Right. Attiny and software definitely is not up-to task for external trigger application. It possibly could be done in *hardware* of other MCU - like advanced timer of stm32 clocked at 72MHz. Using such one can make adjustable delay and pulse length with 14ns error. If programmable delay is not needed then just pulse length has 14ns error.

[Sceadwian]

Yeah I missed some of those earlier posts thanks, some good info there. I've noticed the phosphor afterglow myself apparently it's normal. The action/reaction of phosphors has never been anything I"ve looked into, get into chemistry and some atomic physics that have a bit more nuance than I care to get into Thanks for the graphs though! Very helpful.

Is this even possible with white LEDs? I thought the phosphor delay was WAY more than the nanosecond range?

It depends on how fast and the LED. I posted an oscillogram of the CMA3090 being pulsed at 1µs here. As you can see there some delay in the on and off time, but not much. As I said in my previous post, some phosphor LEDs are worse, others are better.

What's odd about phosphor based LEDs is, although the initial decay is fairly fast (even the slow yellow one I tested had a time constant in the µs range) a very faint afterglow persists for a long time after the LED is turned off. This is no problem for a photoflash application, because it's several orders of magnitude too small to cause a problem: it's just about detectable, with dark adapted eyes, in a very dark room. I've seen it for myself. I ruled out persistence of vision and charged capacitors by closing my eyes, exposing the LED to a backlight (UVA) whilst my eyes were still closed, turning off the UV source and looking at the faint afterglow from the LED. I have only done this with white LEDs and don't have the equipment to make quantitive tests, otherwise I would.

[Marco]

So we would want the MC to trigger on the pulse, with a programmable delay (just a few microseconds) before sending the trigger for the flash.

Do an analogue delay, a current source charges up a capacitor, you compare it against an adjustable threshold with a fast comparator.

I was thinking on that last night. A 10s (37v) Lipo battery rated at 4000mAh could provide a pretty good jolt of energy across an LED.

Nope.

The amount of energy you're burning is tiny, so you just use electrolytic capacitors (they actually have pretty low inductance). It's advantageous to keep the powersupply which charges those capacitors low current, because if something does go wrong by any chance that limits the continuous current which can flow so everything can't blow up. Unlike what happens if you short that LiPo stack.

None of this is entirely trivial, it's quite a complex project. I think I'd design it with a comparator on the input with adjustable threshold and feedback to the input with a diode (so once triggered it stays on, until you arm it again with a physical button, add a signal LED so you can see if it's armed). This will turn your trigger signal into a step signal. The step signal goes into your pulse generator with adjustable delay and pulse duration, a fairly complex circuit ... too complex to describe in text. That then goes to the drivers/LEDs/MOSFETs. You need 2 boost converters. One boost converter for the MOSFET drivers, one boost converter to charge the flash capacitors. Also one linear regulator for the trigger section.

All in all, a lot of components.

[JAndrew]

None of this is entirely trivial, it's quite a complex project.

I definitely see that. I've learned quite a bit in the last 48-72hrs just going through everyone's responses. Honestly its likely a little above my current level of understanding. If I am going to make this work I'm going to have to spend some time on smaller and slower circuits. Particularly looking at/recreating the examples of what has already been done.

I appreciate everyone's input, I definitely have a better sense of where to start, and how to eventually get there.

To be clear, you can still take some pretty awesome pictures with flash duration of 2-4 microseconds. I just set a target, based on the specs of the equipment I had used prior.