New Puck.js production run

A vote for a sensor which determines whether the Puck is moving or not.

@Jean-Philippe_Rey sorry I missed your post - that looks like a neat device (and yes, tiny - and very shallow!). I'll have to see about using it on other devices in the future (it might have been possible to get RGB on the Pico I guess!).

The luminous intensity of red at 2mA seems to be around 14 mcd, and the one I'm using is 100mcd! I wonder if people wouldn't complain more about the lack of brightness or battery life than they would about the colour mixing :)

@user73560 http://forum.espruino.com/conversations/301185/#comment13502846 ?

They're all Kingbright - APT1608LSECKJ3RV, APT1608LZGCK and APT1608LVBCD. The idea is they're better at low currents like ~2mA - since you shouldn't be drawing loads of power out of the CR2032

Wow, those are very efficient! I'll see if I can find something comparable in RGB that's thin and diffused.

I've noticed the silicone picks up dust too - I'm not sure there's much that can be done about that. At least if I were to sell tinted black covers then it would be less obvious :)

I'd just try washing the silicone cover with some hand soap. Tinted covers might reduce LED brightness further, of course.

They're all Kingbright - APT1608LSECKJ3RV, APT1608LZGCK and APT1608LVBCD. The idea is they're better at low currents like ~2mA

Your LEDs are 100mcd, 50mcd, 24mcd for RGB, respectively.

This RGB LED, also Kingbright, might be interesting.

It's cheaper, brighter, and blended.

It's 60mcd, 180mcd, 30mcd (typical) at 2mA. Overall, that's 100mcd brighter than your current ones. (chart attached)

It's about USD 0.31 ea, qty 2000, vs about USD 0.46 for your current 3 LEDs. It's 2mm high (about 1mm taller than your current ones) but should have a more blended color output. It might end up with a smaller beam, but that would also be perceptively brighter. I think this would increase perceived quality.

One could still do the half dome mod (ping pong ball or 3D printed form).

I'm wondering now whether a thin disc of perspex with one side roughed up would solve both problems?

Edge-lit, laser-etched Perspex (acrylic) is a great way to do diffusion in small quantities; I do this frequently when prototyping products at the design firm where I work. But there isn't enough room in the Puck to do it properly. And that's fine.

1 Attachment

@gordon, which model of IR LED are you using on the current Puck hardware? I'd love to help track down one that extends the range.

:) thanks! It's the APT1608F3C - so also Kingbright, although not one of the low power variants.

If there was one I could stick in, that'd be amazing! That really would be a nice easy improvement for this next production run.

edit: given the recent firmware change that's pushed the range up to almost a meter, this might not be needed now :)

The SunLED XZTNI53W-8 may be a drop-in replacement.

• 60° angle, vs 120° (Kingbright)
  

• 2.8 mW/sr, vs 2mW/sr (Kingbright) (CIE127-2007) (both at 20mA)

• USD 0.23/ea in qty 100.

Something like the SunLED XZTNI55W-3 might be interesting down the road. The footprint is 3.2 x 1.6mm (vs 1.6 x 0.8, for the Kingbright APT1608F3C).

• 40° angle, vs 120° (Kingbright)
  

• 4.8 mW/sr, vs 2mW/sr (Kingbright) (CIE127-2007) (both at 20mA)

• USD 0.22/ea in qty 100.

Supposedly, these are “Ideal for indication light on hand held products.” 🤔

Thanks! What are your thoughts about the emission angle though? It feels to me like the Puck is usually going to be face up with the thing receiving at right angles, so you want something pretty wide. I guess the silicone case would do a very good job of diffusing it though.

Also, I could be wrong here as it's new to me, but mW/sr is power/angle - so to get a radiated power value you need to multiply by the radiated angle (ish). It sounds like the Kingbright one might actually kick out more power total, just over a wider angle? I guess the only way to be sure would be to get some, solder them on, and do some range tests.

Side subject, on the subject of pogo pins, can someone find what is likely to work?
ie a pogo that is correct length, fit into the pins on the bottom of the puck with plastic on bottom and female that allows for adhesion (or a 3d print of black bottom) for fixation.
I think this could be nice for the new (and old) puck.js.
Once found (with link to EU supplier), the pogo pins could be tested in the community.

The pogo pins could be bought from Gordon if time/value allows with a buy of a puck.js (with price on pogo pins ((add on sales, be it ever so little)) ).
It would show ability of puck to be puck-ered on to something else and retain the cool design

3D print of the bottom could come from community (as there are some cool guys out there who are good at 3d design ((I am not)) ). The 3D should contain openings for the pins that go into the bottom, and whatever is needed to hold the pogos in place

The Ruuvi guys gave me the part number for their pogo pins: Mill-Max 0926-1-15-20-75-14-11-0: http://www.digikey.co.uk/product-detail/en/mill-max-manufacturing-corp/0926-1-15-20-75-14-11-0/ED1652-ND/2784012

They look really nice - they're really compact, however as with all the Pogo pins you'd have to come up with a separate way to clip them on.

Having said that, they're not that cheap - having a bunch of them starts to get expensive quickly.

I am wondering if there finds pogos with holes in the bottom to accept male pins: the pogos look solid in the bottom.

Edit:
not cheap and a quick link: seems like male-male is mostly used which gives something that sticks out of puck.js bottom, but one can live with that

http://www.digikey.co.uk/products/en/connectors-interconnects/contacts-spring-loaded-and-pressure/311?FV=1140050%2C1f140000%2Cffe00137&mnonly=0&newproducts=0&ColumnSort=0&page=1&stock=1&pbfree=0&rohs=1&quantity=&ptm=0&fid=0&pageSize=500

and on bottom silk, the pin names: I am sitting turning my puck.js round all the time to make sure I have taken the correct hole...

Now that's a good idea - unfortunately they've already gone to production with the PCB as-is :(

I'll make a note about it for next time though.