Author Topic: Looking for Cheap, Ultra Fast LEDs (2-3ns)  (Read 7162 times)

0 Members and 1 Guest are viewing this topic.

Offline suppermanTopic starter

  • Regular Contributor
  • *
  • Posts: 111
  • Country: us
Looking for Cheap, Ultra Fast LEDs (2-3ns)
« on: January 24, 2020, 05:24:16 pm »
I have found it very difficult to find LEDs with ultra-fast rise and fall times. The best I can find are by Vishay.

http://www.vishay.com/docs/84581/vsly3943.pdf

They have the right price point ($0.22 in low volume - DigiKey) and I can overdrive them to 2ns rise / fall with some fancy circuitry. But they are not ideal. I would rather go towards the visual range (650-850nm). And I would love to save money on the drive circuit by having a faster LED. I could also use a larger field of view.

The problem is it is very difficult to even search for these specs...

Any tips on how to search and find LEDs by rise and fall time? Only IR LEDs seem to list it on data sheets and most websites don't let you search on it (some luck with mouser). Low current is preferred because it is cheaper to overdrive them.

any tips greatly appreciated..
 

Offline Marco

  • Super Contributor
  • ***
  • Posts: 7044
  • Country: nl
Re: Looking for Cheap, Ultra Fast LEDs (2-3ns)
« Reply #1 on: January 24, 2020, 05:44:17 pm »
Is the application really so price sensitive you can't use a laser diode?

If it has to be a LED I'd try to get the smallest blue LED you can find, probably 0201, and giving them a spin. Reducing size seems pretty much a necessity to speed up LEDs.
« Last Edit: January 24, 2020, 05:58:00 pm by Marco »
 

Offline suppermanTopic starter

  • Regular Contributor
  • *
  • Posts: 111
  • Country: us
Re: Looking for Cheap, Ultra Fast LEDs (2-3ns)
« Reply #2 on: January 24, 2020, 06:42:29 pm »
Is the application really so price sensitive you can't use a laser diode?

If it has to be a LED I'd try to get the smallest blue LED you can find, probably 0201, and giving them a spin. Reducing size seems pretty much a necessity to speed up LEDs.

I'm moving a lot of data in a unique configuration.. so I will need at least 10+ sources.. so cost per unit is a factor. Also, cost of logic goes up as well as data rates per unit go up. I would prefer to stick with LEDs.. unless you know of a laser diode for $0.50 or so..
 

Offline Marco

  • Super Contributor
  • ***
  • Posts: 7044
  • Country: nl
Re: Looking for Cheap, Ultra Fast LEDs (2-3ns)
« Reply #3 on: January 24, 2020, 10:57:56 pm »
unless you know of a laser diode for $0.50 or so..

China must have some sources, because they are selling modules with lenses for ~10 cents a piece.

PS. oh, I guess the really cheap ones can't really easily be PCB mounted. Cheapest TO18's seem more in 80 cent range.
« Last Edit: January 25, 2020, 07:07:25 am by Marco »
 

Offline moffy

  • Super Contributor
  • ***
  • Posts: 2217
  • Country: au
Re: Looking for Cheap, Ultra Fast LEDs (2-3ns)
« Reply #4 on: January 25, 2020, 11:03:00 pm »
This is a nice article about driving a blue LED with 1-2ns rise times: https://arxiv.org/ftp/arxiv/papers/1011/1011.1954.pdf
Maybe some of the Broadcom HLMP devices would be suitable.
 
The following users thanked this post: edavid

Offline suppermanTopic starter

  • Regular Contributor
  • *
  • Posts: 111
  • Country: us
Re: Looking for Cheap, Ultra Fast LEDs (2-3ns)
« Reply #5 on: January 26, 2020, 01:18:51 am »
unless you know of a laser diode for $0.50 or so..

China must have some sources, because they are selling modules with lenses for ~10 cents a piece.

PS. oh, I guess the really cheap ones can't really easily be PCB mounted. Cheapest TO18's seem more in 80 cent range.

Thanks for the link. I did look at them.. but this is going to be a consumer product and I can't pick components without companies or data sheets... I think these lasers are mean for key chains.
 

Offline suppermanTopic starter

  • Regular Contributor
  • *
  • Posts: 111
  • Country: us
Re: Looking for Cheap, Ultra Fast LEDs (2-3ns)
« Reply #6 on: January 26, 2020, 01:24:55 am »
This is a nice article about driving a blue LED with 1-2ns rise times: https://arxiv.org/ftp/arxiv/papers/1011/1011.1954.pdf
Maybe some of the Broadcom HLMP devices would be suitable.

Thanks for that article.

I had read it during my research. They don't mention the LED (from what I remember) and it is not real world (i.e. PSpice only). In the end it is about driving the LED with an OpAmp.. and at that speed and current that is super expensive for the OpAmp. My circuit will pre-load (keep the LED off but warm) then spike it positive or negative on a change in on/off state for under 1ns.  The circuit is about $0.35 right now + LED.

Mainly just looking for actual LEDs that are fast.. in the 650-950nm range.. 800ish preferred.
 

Offline Marco

  • Super Contributor
  • ***
  • Posts: 7044
  • Country: nl
Re: Looking for Cheap, Ultra Fast LEDs (2-3ns)
« Reply #7 on: January 26, 2020, 01:52:17 am »
I think if you want visible light you're just going to have to bite the bullet and experiment, I don't think anyone has ever tried the latest smallest SMD LEDs yet and it's certainly not a datasheet feature. The 940 nm LEDs are probably used for POF, so they mention rise/fall time ... but no one is doing that with visible, so at best the datasheet might mention junction capacitance.

You might get optical speed up by applying a bit of negative voltage too.
« Last Edit: January 26, 2020, 01:55:11 am by Marco »
 

Offline Marco

  • Super Contributor
  • ***
  • Posts: 7044
  • Country: nl
Re: Looking for Cheap, Ultra Fast LEDs (2-3ns)
« Reply #8 on: January 26, 2020, 04:16:35 am »
This is a nice article about driving a blue LED with 1-2ns rise times: https://arxiv.org/ftp/arxiv/papers/1011/1011.1954.pdf
"High-Speed LED Driver for ns-Pulse Switching of High-Current LEDs" is a bit more valid, actually measures optical rise/fall with ways to speed it up.

They use a capacitor to give an initial high current kick to speed up the optical rise time and a parallel inductor to provide negative current during turn off to speed up optical fall time.
 

Offline james_s

  • Super Contributor
  • ***
  • Posts: 21611
  • Country: us
Re: Looking for Cheap, Ultra Fast LEDs (2-3ns)
« Reply #9 on: January 26, 2020, 04:43:57 am »
Does the size of the LED even matter? I thought the smallest SMT LEDs had pretty much the same size die as larger ones and ordinary 3mm and 5mm LEDs just in a smaller package?
 

Offline Marco

  • Super Contributor
  • ***
  • Posts: 7044
  • Country: nl
Re: Looking for Cheap, Ultra Fast LEDs (2-3ns)
« Reply #10 on: January 26, 2020, 05:06:25 am »
At least the bond wires will be shorter if nothing else.
 

Offline moffy

  • Super Contributor
  • ***
  • Posts: 2217
  • Country: au
Re: Looking for Cheap, Ultra Fast LEDs (2-3ns)
« Reply #11 on: January 26, 2020, 06:38:48 am »
This is a nice article about driving a blue LED with 1-2ns rise times: https://arxiv.org/ftp/arxiv/papers/1011/1011.1954.pdf
Maybe some of the Broadcom HLMP devices would be suitable.

Thanks for that article.

I had read it during my research. They don't mention the LED (from what I remember) and it is not real world (i.e. PSpice only). In the end it is about driving the LED with an OpAmp.. and at that speed and current that is super expensive for the OpAmp. My circuit will pre-load (keep the LED off but warm) then spike it positive or negative on a change in on/off state for under 1ns.  The circuit is about $0.35 right now + LED.

Mainly just looking for actual LEDs that are fast.. in the 650-950nm range.. 800ish preferred.

They do verify their PSICE model with actual measurements, at least voltage rise and fall. The opamp is just providing a very fast, high current pulse (there is no feedback except for voltage gain). You just need a push/pull output stage to source/sink the current. Lots of ways to do that cheaply.
The LED is mentioned in the appendix: The datasheet for the HLMP-CB15-P0000. It is referenced along with the datasheet for the opamp. Unfortunately it is obsolete but you should be able to find an equivalent. :)
 

Offline ali_asadzadeh

  • Super Contributor
  • ***
  • Posts: 1967
  • Country: ca
Re: Looking for Cheap, Ultra Fast LEDs (2-3ns)
« Reply #12 on: January 26, 2020, 06:47:09 am »
Interesting topic, what do you want to do with these leds? If you found good sources in china please share :-+
ASiDesigner, Stands for Application specific intelligent devices
I'm a Digital Expert from 8-bits to 64-bits
 

Offline Marco

  • Super Contributor
  • ***
  • Posts: 7044
  • Country: nl
Re: Looking for Cheap, Ultra Fast LEDs (2-3ns)
« Reply #13 on: January 26, 2020, 06:53:53 am »
They do verify their PSICE model with actual measurements, at least voltage rise and fall.

The problem is that until the entire junction is filled/swept-out the current doesn't equal the optical output, so electrical response isn't all that relevant.
« Last Edit: January 26, 2020, 07:01:38 am by Marco »
 

Offline OwO

  • Super Contributor
  • ***
  • Posts: 1250
  • Country: cn
  • RF Engineer.
Re: Looking for Cheap, Ultra Fast LEDs (2-3ns)
« Reply #14 on: January 26, 2020, 07:05:35 am »
Ordinary blue LEDs (usually GaN based) will all go to a few hundred MHz easily from my tests.  What you want to do is bias the LED to half the operating current and measure the small signal impedance, like so:

Then tune out the reactive portion and drive it with a wideband MMIC.
Email: OwOwOwOwO123@outlook.com
 
The following users thanked this post: I wanted a rude username

Offline OwO

  • Super Contributor
  • ***
  • Posts: 1250
  • Country: cn
  • RF Engineer.
Re: Looking for Cheap, Ultra Fast LEDs (2-3ns)
« Reply #15 on: January 26, 2020, 07:13:22 am »
Those rise and fall times in that datasheet are complete bollocks. There is no test circuit, nothing about the drive conditions, nothing about how I_F is defined. For all you know they could be driving with a hard low-impedance voltage pulse source and only counting final steady state current (which experience says they probably are). It's the same story with MOSFET rise/fall times. A better way to characterize speed is to drive with a defined impedance source (e.g. 50ohms), defined power level and plot the optical response vs frequency. Then you can look at the -3dB bandwidth and say how fast the LED is.
Email: OwOwOwOwO123@outlook.com
 
The following users thanked this post: hans, Someone, daqq

Offline suppermanTopic starter

  • Regular Contributor
  • *
  • Posts: 111
  • Country: us
Re: Looking for Cheap, Ultra Fast LEDs (2-3ns)
« Reply #16 on: January 26, 2020, 03:52:42 pm »
Wow, a lot to catch up on guys.. thanks..

I have my own circuit.. cost me about $0.35 right now + LED. It pre-biases the LED (keeps it off but warm). Gives a very large positive or negative (adjustable) pule on high / low transitions wile keeping currents low on continuous on stages.
I measure voltage and current to verify. I have a very fast photo-diode to see optical response. There is plenty of room for error here.. but I'm not a test lab.. just looking for a readable signal.
For the most part the rise time on spec sheets is well defined.. they typically list the drive current and since it is 10-90% in most cases.. they don't worry about pre-biasing it. Usually a fast Mosfet acting as a switch.

I'm using it to send data across about a 1 cm air gap.. and I use around 10+ of them to get the data I need. Not easy.. since we are talking about serious data over very cheap components.

I always have to remind myself that light travels about one foot at 1ns.. to get some perspective..

Also - running this with an FPGA which is limited to about 600 Mbit per pin so taking it down to say a single laser would add all sorts of cost to serialize multiple channels.. etc.. so I think I'm happy with LEDs giving me around 400 Mbit... I think it should be possible..


 

Offline dmendesf

  • Frequent Contributor
  • **
  • Posts: 342
  • Country: br
Re: Looking for Cheap, Ultra Fast LEDs (2-3ns)
« Reply #17 on: January 26, 2020, 05:39:18 pm »
Is this an isolated scope/probe? I can't imagine something else that would need isolation and transfer gigabytes per second...
 

Offline suppermanTopic starter

  • Regular Contributor
  • *
  • Posts: 111
  • Country: us
Re: Looking for Cheap, Ultra Fast LEDs (2-3ns)
« Reply #18 on: January 26, 2020, 09:16:01 pm »
Is this an isolated scope/probe? I can't imagine something else that would need isolation and transfer gigabytes per second...

No. Consumer. Video... can't say more than that...
 

Offline Marco

  • Super Contributor
  • ***
  • Posts: 7044
  • Country: nl
Re: Looking for Cheap, Ultra Fast LEDs (2-3ns)
« Reply #19 on: January 26, 2020, 10:08:30 pm »
Did you see the speedup circuit in the paper? That should fit in your budget and might eek out some extra performance from indicator LEDs if necessary.
 
The following users thanked this post: MagicSmoker

Offline StillTrying

  • Super Contributor
  • ***
  • Posts: 2850
  • Country: se
  • Country: Broken Britain
Re: Looking for Cheap, Ultra Fast LEDs (2-3ns)
« Reply #20 on: January 26, 2020, 11:22:50 pm »
so I think I'm happy with LEDs giving me around 400 Mbit... I think it should be possible..

I think you could do a 4Mb/s channel for $0.50, but I think there's no chance of 400Mb/s.
.  That took much longer than I thought it would.
 

Offline ali_asadzadeh

  • Super Contributor
  • ***
  • Posts: 1967
  • Country: ca
Re: Looking for Cheap, Ultra Fast LEDs (2-3ns)
« Reply #21 on: January 27, 2020, 06:18:42 am »
What's the receiving circuit? are they LEDs too?
ASiDesigner, Stands for Application specific intelligent devices
I'm a Digital Expert from 8-bits to 64-bits
 

Offline Marco

  • Super Contributor
  • ***
  • Posts: 7044
  • Country: nl
Re: Looking for Cheap, Ultra Fast LEDs (2-3ns)
« Reply #22 on: January 27, 2020, 06:32:57 am »
Presumably SFH2701.
 

Offline TheUnnamedNewbie

  • Super Contributor
  • ***
  • Posts: 1211
  • Country: 00
  • mmwave RFIC/antenna designer
Re: Looking for Cheap, Ultra Fast LEDs (2-3ns)
« Reply #23 on: January 27, 2020, 03:19:36 pm »
This sounds a lot like an x-y problem.

Incidentally, getting lots (>>gigabit) data from A-to-B for cheap is the field I'm working in, so perhaps I could help with some outside-of-the-box ideas if you can give more information/limits you are facing. What volume are you looking at?

Depending on what output type you have, you might be able to do something with larger symbol size? instead of OOK, consider 4 or even 8 ASK to lower symbolrate? 
« Last Edit: January 27, 2020, 03:22:03 pm by TheUnnamedNewbie »
The best part about magic is when it stops being magic and becomes science instead

"There was no road, but the people walked on it, and the road came to be, and the people followed it, for the road took the path of least resistance"
 
The following users thanked this post: tooki

Offline suppermanTopic starter

  • Regular Contributor
  • *
  • Posts: 111
  • Country: us
Re: Looking for Cheap, Ultra Fast LEDs (2-3ns)
« Reply #24 on: January 27, 2020, 04:39:49 pm »
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf