How to they get a line of them to all blink in a "cascade"?
I'm guessing some sort of Bluetooth link and some software to control triggering but I'm not sure what initiates the cascade and what stops it running backwards.
Tim
Why so complex? simply one triggers the next which fires after a delay would be my guess. if one does not get a trigger after a longer delay, it flashes any way.
I've always assumed that one lamp looks at the one in front and triggers its own lamp after a prescribed amount of time.
Nothing more than a lens, a photocell and a timer.
If it doesn't see anything for a specific amount of time then it would launch into autonomous mode.
BICBW
The cascade predates bluetooth - I think it's optical innit? Whatever it is it's going to be very dumb. Set one at the front to flash. Then point sensor of the next one to the one in front, fixed delay after it sees the light come on, turn your own light on, repeat. So you need one flasher and the rest are identical slaves. The electronics for that is trivial.
Despite looking as closely as possible, I've never seen a visible sensor and many lamps are unidirectional with no light visible from the back. How does the second lamp know when to flash if it can't "see" the first?
Tim
I think the master one flashes by itself and all the rest have a photodiode and delay circuit to "fire" shortly after they see the one up stream has flashed ...
I think I have a couple of the burners in the garage. No housings though.
According to the document I posted a link to there is no master. You just plonk the lamp onto a cone, which switches it on and they automagically sort themselves out into the blocks of six.
Red or Yellow case and Lens ? I gathered a Red one from somewhere and used it to take the chill off my A35 Van overnight in Winter It had no heater so stayed cold even with the engine going. Yellow ones came in shortly after around 1974 but it wasn't long before the battery ones took over though a match was still the easiest way to turn one on , the switch was operated by sticking the match through a small hole and moving it. To bring a bit of DIY into this I drilled two similar holes in the case of a Dimplex wall mounted fan heater so a matchstick could move the hidden switch which changed it from 1kW to 2kW which was normally set on installation. I could then change it as required without taking the cover off.
G.Harman
Really, are you sure its not just that someone started them all off by walking down the row and switching them on?
Brian
These lamps produce a visual illusion of a light moving along the line of the lights. Actually it's several lights all "moving" in the same direction and six positions apart.
If the lamps all free ran they wouldn't stay synchronised enough to produce the constant visual effect. Well not without spending a relatively large amount on an accurate clock in each one.
"Accurate" meaning that they all tick at the same rate *and* all drift by the same amount due to temperature or WHY. IIRC the step time between lamps is 200 ms if that changed by more than 20 ms I think it would be very noticeable. 20 ms over 24 hours is 1 part in
4.32 million. That's not easy to consistently mass produce.
It's not that difficult with the divide by 2^22 counters that take a crystal of 32,768 Hz down to 1 Hz that get you some pretty good accuracies for watches that cost less than a birthday card.
OOI, if you want to go to the other extreme, geta watch with an atomic clock in it! I recently visted these people:
I put a pic of the guts here:-
An ordinary white LED is on the other side of the pcb fed via the 150 ohm resistor visible. When placed on a cone the long contact is pressed against the pcb . It works on 6 volts and free-runs. The rectangular chip under the (epoxy?) lens is about 0.15" x 0.1" and has fine bonding wires to the pcb. Presumable would sync - but I have not got time just now to set up a test.
But we don't want a 1 second tick but a 200 ms one *and* happening in sync, all the time, with all the other lamps (up to 250 IIRC) over a temperature range of at least -10 to +30 C, though in full sun the lamps may well get considerablly hotter inside. And with free running clocks you still need some means of setting, to a pretty fine tolerance, when a given lamp is to flash.
Flashing an IR light with an embeded code of some sort is far simpler and very much more reliable compared to free running clocks.
A "birthday card" watch with an accuracy of 1 s in 24 hours (30 ish seconds a month) is only 1 part in 86,400 not 1 in 4,320,000.
Some corrections here:-
It has a white body but is actually yellow when lit. It is pulsed 1mS on and 9nS off.
Hmm - maybe mine is a bit dumb. It does one flash when power is connected (in lighted room). In the dark, it is switched on permanently. Flashing another LED at the sensor turns the yellow LED off after about 200mS. Not sure if that would cascade.
Then pay a bit more or maybe less for a faster clock frequency.
- - - - that should read 9mS off
Actually I think it should be 9ms off.
MilliSieverts? Are you sure? That's pretty radioactive.
HomeOwnersHub website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.