Breadboarding is a great way to get started with electronics, and with the wide availability of those little wire jumpers, it’s never been easier – until you hit roadblocks due to poor connections and parasitic capacitance futzing with your signals. However, in today’s current climate, the latest and greatest modules are too often available only in SMD packages, and while breakout boards can help, it’s probably overcomplicating things a bit when it comes to SMD LEDs. It’s all good, though – [Simon Merrett] has a workaround, as part of his Yapolamp project.
[Simon] first took a flat strip of steel, and placed two neodymium magnets on top. The assembly was then wrapped in electrical tape for insulation, and two contacts were created with copper tape. The LEDs were then placed across the two contacts and wires were attached to join them to the breadboard. The 5630 LEDs [Simon] must contain some sort of ferrous material, because they were attracted to the magnets and sat neatly in place.
It’s a neat hack that would be particularly useful if you needed to quickly swap out LEDs, and saves them from damage by soldering. Meanwhile, check out this SMD LED matrix from 2009.
This looks like a great way to check polarity and “is it broken?”.
did this really need an article? its just tape… rant.end();
… and magnets. Yes, just those things. I submitted because it might be helpful for someone else. If not for you, that’s ok.
I for one wouldn’t have dreamt of SMD LEDs being magnetic. Now I know it’s worth to try…
So thanks!
These ones are only because of the excessive amount of metal in their construction. Most SMD LEDs seem to have more non-ferrous bond wire than base plate metal (slightly ferrous)
But…..magnets!
i like your idea…
Great idea but LED’s in parallel? Good for testing one at a time, or winner takes all.
In the case of Yapolamp and the TritiLED it takes inspiration from, we’re driving the LEDs at much lower current than their datasheet nominal rating, as their peak efficiency is not up near those ratings. We’re also using a “less than one” duty cycle, so I don’t think thermal runaway or other current-related failure modes are likely, if that’s what you are hinting at, in the case of the current being concentrated in fewer LEDs as and when some of them fail.
Specifically for Yapolamp, I’m using multiple smaller LEDs as a proxy for a “distributed die” – I want to spread the light source over an area so that if (when) the user shines it in their eye at close proximity, there’s less energy entering the eye and less resultant damage. That’s also the reasoning behind the selection of lower frequency light than white LEDs based on blue-emitting dies.
It’s a great idea to apply a magnet to the LED tester! You can take a piece of LED tape, solder the LED and use it here. Or a piece of flexible kaptone plume with copper tracks from old devices
You can still drive the LEDs with low current with a series connection, it is the right way to do it.
Please could you elaborate on what makes series right, as opposed to parallel, in this case?
I was under the impression that the driving charge coming from the inductor would need to be double the forward voltage of the LEDs to drive two and I’d like to drive four or five. Wouldn’t that mean the drive circuitry would have to provide a much higher voltage to drive in series?
Tolerances.
Look to https://en.wikipedia.org/wiki/Light-emitting_diode#/media/File:Diode-IV-Curve.svg – What happens if one of the LEDs has a slightly lower Vd? It get’s most of the Current.
Thank you for the explanation and link, but I’m not sure the problem applies to the Yapolamp; if the total current through all the parallel LEDs is less than rated current for a single LED, what’s the concern with one LED receiving most of that low current, if it happens to be at one extreme of the tolerances?
Furthermore, from an application perspective, with a series arrangement the failure of a single LED means no more light can be produced, whereas the failure of a single LED in a parallel arrangement still allows all remaining LEDs to provide light.
Smerret79, the problem is you get into a feedback cycle. LED 1 draws a little more current than LED 2, so #1 heats up and #2 cools off a little, which causes even more current to flow through #1, which heats it more, which causes it to draw more current… And the cycle continues until LED 1 is drawing twice the rated current and burns up. If heat raised the resistance we wouldn’t have this problem because the current would stabilize from negative feedback. In reality we gave positive feedback so it burns the LED up.
Smerret79, the problem is you get into a feedback cycle. LED 1 draws a little more current than LED 2, so #1 heats up and #2 cools off a little, which causes even more current to flow through #1, which heats it more, which causes it to draw more current… And the cycle continues until LED 1 is drawing twice the rated current and burns up. If heat raised the resistance we wouldn’t have this problem because the current would stabilize from negative feedback. In reality we gave positive feedback so it burns the LED up.
Matt, thanks but the pulsed inductor driving circuit isn’t able to provide the rated current as far as I can tell from the LT Spice simulation, so how can it burn up an LED, even one with 0 resistance?
Why not put some tape on a long plastic clamp, like say one for meant for hair (to name a random example)? That way you don’t need to desperately search for magnetic LED.
Or make your own clamp, from plastic or wood, with a nice easy lever to operate it.
That sounds fine, although it was the fact of noticing that these chips were magnetic that made me think of this, rather than making the holder and trying to find suitable LEDs.
Some aspects of your suggestion would be worth looking at in more detail – the first is that the magnetic method makes connection easily with the jumper wire terminals, which are a different thickness to the LEDs and so might not be grabbed by the clamp. On a similar note, the magnetic method appears to be quite easy to scale, whereas adding and removing an LED from the clamp might affect the support and connection with the remaining ones. Finally, the magnetic method applies a fairly even contact pressure to make the electrical connections (I think the slight elastic “give” in the insulation tape helps), whereas a clip is like a beam, especially if it’s wide. Once you have more than two LEDs keeping the jaws of the clip apart, it becomes more difficult to ensure that all components are making electrical contact. Perhaps a layer of foam between the jaws and the copper tape would assist with this.