Easily LED

RE: Easily LED

alexdbird — Sun, 03 May 2020 07:28:35 GMT

The reason I think measurements are so variable is the same reason the longer leads don’t work. The output of the LED is very high impedance. You can solve this to some extent with a resistor, but as the LED also has a very low output that loses some sensitivity. For the best results you need to use an op-amp. Powering an op-amp is a PITA though, so try a single rail one like an MCP-601 that can run on the same supply as the Arduino. You can also change the Vref on some chips to make the analogue inputs more sensitive.
Sorry I’m not awake enough to actually explain the impedance thing!

RE: Easily LED

jw0752 — Sun, 19 Apr 2020 22:20:11 GMT

Hi Paj,

I enjoyed the trip you took us on through your experiments. While everything you learned could have been googled and read it was great to see you actually put the wisdom of the net to the test in your experiments. It is amazing how often new things are learned. One thing that I learned a while ago and that surprised me is that even with our precise manufacturing procedures most semi conductor, including LEDs, come out of the factory with slightly different specs. This is probably why you were reading different voltages on your LEDs of the same color. If one wants two identical components it is necessary to search through a bunch to find two that are within the tolerance of similarity to each other that the designer requires.

Thanks for your post.

John

RE: Easily LED

jc2048 — Fri, 17 Apr 2020 11:50:50 GMT

I was curious about your reverse measurements. It's not something that you normally ever look at if you design with LEDs, you just go by what the datasheet tells you (which is usually 5V).

I've tried it with a few LEDs, but unfortunately I don't know what they are as they are from old bags that have been lying around for years (once upon a time, LEDs were part of my day job and these are some from spares that were kept on a storeroom shelf and never used). This is with a test current of about 20mA, but it's all a bit approximate. This isn't a piece of professional test equipment; it's just a lash-up with an Arduino, so take the results with a little skepticism.

Here's a red, clear, high brighness, narrow viewing angle

[View:/resized-image/__size/480x234/__key/commentfiles/f7d226abd59f475c9d224a79e3f0ec07-25fdfc61-5907-48ae-b4f2-36c9ab662349/contentimage_5F00_206729.png:480:234]

Here's a green, fairly high brightness

[View:/resized-image/__size/480x234/__key/commentfiles/f7d226abd59f475c9d224a79e3f0ec07-25fdfc61-5907-48ae-b4f2-36c9ab662349/contentimage_5F00_206730.png:480:234]

Finally here's one I do know the part number of, it's a Kingbright L-7104ID red LED (GaAsP/GaP), which is impressive.

[View:/resized-image/__size/480x234/__key/commentfiles/f7d226abd59f475c9d224a79e3f0ec07-25fdfc61-5907-48ae-b4f2-36c9ab662349/contentimage_5F00_206731.png:480:234]

That last one is getting very 'iffy'. This is my test circuit without an LED

[View:/resized-image/__size/480x234/__key/commentfiles/f7d226abd59f475c9d224a79e3f0ec07-25fdfc61-5907-48ae-b4f2-36c9ab662349/contentimage_5F00_206732.png:480:234]

so it only reaches -180V anyway and at -170V it's running out of 'steam' [current really] and it's probably just increased leakage that's folding it over and not an actual, sustained avalache.

All the above were happy to repetitively avalanch twice a second, but I also tried an old Nichia high-brightness true-blue and that failed almost immediately after showing a breakdown at about -70V. When they first came out they had to be treated carefully as static-prone components, so I'm not surprised it died quickly.

The low datasheet figure might be down to what can be tolerated before it starts to compromise other things like the lifetime, rather than anything to do with where it breaks down and what damage might then ensue.

As [mention:87db22544d8844169f5ec3635dfac969:e9ed411860ed4f2ba0265705b8793d05] points out, there's a relationship between the forward current and the forward voltage that you need to understand. If your power supply works constant-current to a reasonable accuracy, you could try increasing the current from about 1mA in small steps and seeing what the forward voltage is for each step. If you plot it on a graph, current against voltage, you'll soon see why applying 5V directly across the diode is problematic.

RE: Easily LED

Fred27 — Thu, 16 Apr 2020 10:27:18 GMT

What's critical for an LED is controlling the current, not the voltage. Search for "LED current limiting resistor" and you'll find some useful information. This might be a good place to start.

RE: Easily LED

dubbie — Thu, 16 Apr 2020 09:00:35 GMT

Paj,

Some interesting results. I haven't tried videoing a LED failure so that was something new.

I did like the pan and tilt mechanism - I have wanted one for some time but never got around to making one. I will definitely have to have a go at yours, it looks useful. If making 3D printed tubes for the sun tracker you might find that the 3D printed material is just too light transparent. You might need to insert tubes of some alternative material that is not transparent. I would be interested in seeing more results of the sun tracker as it is another thing that I am interested in making.

Dubbie

RE: Easily LED

Jan Cumps — Tue, 14 Apr 2020 19:41:54 GMT

So from that I can conclude , for my LEDs at least, that they would act as a diode for quite a high current, certainly more than I ever play around with. I guess other LEDs will have different characteristics.

Datasheets state max 4 - 5 V reverse voltage for common LEDs

edit: Jon Clift just posted a two-post series on using LEDs as light sensors: Experimenting With Using an LED as a Light Sensor

His circuit is sensitive to noise too.