Spectral Clock - Propeller Clock

I had it at hand, which does not change the fact that they are very nice drivers

Great design, all I have to add is have you thought about sucking the air out of the inside? It is known that then you will not be able to remove the back cover to get to the system, but thanks to this there would be practically no air resistance and it could become even quieter.

Hi, a beautiful project, I would also be interested in building something like this.
Throw some diagram, program, etc.

I am interested in one thing, whether you need to somehow control the brightness of the LEDs depending on the distance from the center, because the "pixel" of the same size, for example in the middle, will shine constantly, and the "pixel" will appear at a fraction of the time or not will there be a difference in brightness?

Can I have photos of the brushes? How did you solve them? Weight is provided by the motor housing?

big_foot wrote:

Can I have some pictures of the brushcakes? How did you solve them? Weight is provided by the motor housing?

a better solution would be to use two coils placed on the axis of rotation, one on the board and control the alternating voltage that would be contactless

Here there is something about contactless current transfer.

Quote:

... have you thought about sucking air from inside?

This thought flashed to me somewhere, but it would require a very tight casing. Besides, I wouldn't even know how to do it and how to suck the air out of it, because not with my mouth. .

Quote:

I wonder how after soldering all these SMD components?

An ordinary soldering iron. I do not like through-hole assembly, who would like to drill so many holes.

Quote:

I am interested in one thing, whether you need to somehow control the brightness of the LEDs depending on the distance from the center

Of course you can, but in this project, the resources of the microcontroller do not allow it. The only thing I could get was a 3-level brightness scale, so I gave it up altogether. The effect you described is noticeable.

Quote:

Can I have photos of the brushes? How did you solve them? Weight is provided by the motor housing?

The brushes can be seen in the video (they are made of wire with black insulation, so take a closer look). At the end of the wires, the proper brushes ( ) removed from the Walkman motor. The mass, as you well noticed, goes through the motor body and through the mounting bolts. The plus, on the other hand, is transferred to the copper circle on the bottom of the plate.

As for air suction, it's not a stupid idea I do not know if I am thinking correctly, but it seems to me that if the bicycle valve was installed there, of course, inverted, maybe the air could be sucked out. Could you help yourself with a vacuum cleaner with an adapter, and seal the housing with some colorless silicone? hmm ... there is no need to wonder for sure it would silence the whole system. Regards

Quote:

I am interested in one thing, whether you need to somehow control the brightness of the LEDs depending on the distance from the center

Of course you can, but in this project, the resources of the microcontroller do not allow it. The only thing I could get was a 3-step brightness scale, so I gave it up altogether. The effect you described is noticeable.

you don't need to control it from a prock, just select the resistor values in series with the leds :-)

oh, I noticed a bug in reasoning, you have to change the way of counting the time between successive jumps,
as you know, the engine speed is not constant, which causes rounding errors in this 8-bit timer and the displayed image jumps slightly,
to solve this problem you have to use COMPARE MATCH interrupt something like that and add integer with fractional parts to some register which will hold the next timer1 value at which the next step will take place, of course you will round this value to integer already :-)

that is:
It turned out that between jumps you have 230.33333 timer cycles
first you have some REG = 0 in the register; then you add it 230.3333 and you have REG = 230, then you add again 230.3333 and you have REG = 461, of course you compare this REG with the current value of the timer, which probably AVR can do hardware and throw an interrupt, you have to come up with an algorithm for it
then there will be no these jumps
I wrote too much
good luck


I like the project in general

Quote:

you don't need to control it from a prock, just select the resistor values in series with the leds

Yes, when, for example, I light all the LEDs at 360 degrees, your solutions work well, but when I light one line (i.e. a pointer), it does not shine evenly. The effect turned out to be the opposite . The only advice is a different driver, and as I mentioned, I miss megahertz for this (or I can't use them properly ).

When it comes to the smoothness of the display, it can certainly be better written. I am a novice programmer, my programs are exemplary, but the most important thing is that something is shining there, something is winking and the joy is there. Thank you for all valuable comments

mb1988 wrote:

Yes, when, for example, I light all the LEDs at 360 degrees, your solutions work well, but when I light one line (i.e. a pointer), it does not shine evenly. The effect turned out to be the opposite . The only advice is a different driver, and as I mentioned, I miss megahertz (or I can't use them properly) ).

maybe you chose them wrong, too much difference between the largest and the smallest

mb1988 wrote:

When it comes to the smoothness of the display, it can certainly be better written. I am a novice programmer, my programs are exemplary, but the most important thing is that something is shining there, something is blinking and the joy is there. Thank you very much for all valuable comments

when I find moments of time, the algo will describe you. how to do it without using fractions or multiplication / division so it works fast ;-)

mb1988 wrote:

I am a novice programmer, my programs are exemplary, but the most important thing is that something is shining there, something is blinking and the joy is there. Thank you very much for all valuable comments

and for such a beginner, you don't name the variables well, e.g. gCostam means a global variable

Hello, I like your Clock very much, I would like to make it, just no schematic. And that's why I ask when it would be ready.
Greetings

Please be patient, I will share the materials soon.

By the way, after listening to your opinions, I am already a little tempted to remake this clock. I will leave the enclosure because I like it ( ) but inside I would change:
1) coreless transformer instead of brushes
2) 63 (sic!) Blue diodes on the diameter instead of 32 on the radius (this means reducing the engine speed twice, which means almost inaudible operation). I just wonder if it will drag it.
3) added simple IR communication, which would not require a computer and a programmer to change the time.

Eh, only when I have time for all this ...

regarding 63 diodes on the diameter - I do not know if the moment of lighting of these LEDs on the other side of the propeller after half a turn, exactly in the same place as the previous ones, will match. The image would double a bit shifted.

wow! something beautiful ... I love the spectral watches, and this one is especially true!
PS: The theme from the end of the video is brilliant! I didn't think anyone would be able to project an animation on it .

and where did you get the TDA5140

It would look pretty good, but it would be nice to do a brightness adjustment right away so you can get more than 8 colors. But taking into account the number of LEDs, it would be very difficult to make a tile at home for such a layout. Besides, if you want to get the aforementioned colors at such a speed, I think that you would need to put something more powerful than the atmega inside.

I got the TDA5140A at tme (no advertising).

Great respect
I just admire and envy my skills and knowledge because you can and say that it's simple, but not for me.

For me, I would need a step-by-step description (so peasant) plus a applet.

Hello. I have to say that the clock really impressed me !! I'm already thinking about my own construction, and here's my question: what do you think will it be possible to write software in bascom? Isn't he
too "slow" for such applications ?? How much space does yours take up
code in processor ??

Come on, buddy, respect .
Projekcik great, I am waiting impatiently for the documentation and in the meantime I will think about another way to power the chip e.g. a rotating transformer
Greetings

Quote:

Zubiik , in my opinion and from what I heard, the bascom is terribly muscular and eats up memory. and after that, the test version gives too little possibility to generate the code, the size is that the code cannot exceed 4096kb.

If you can write the correct programs and get to know Bascom, you will be able to do the acoustic spectrum analyzer in Bascom.

It can be done in Bascom and the display support should fit within 4kB of the result code (of course, it depends what you want to have in addition to the display / watch).

Congratulations, your clock is even on the monster joe.
http://www.joemonster.org/filmy/7287/Zegar-widmowy-na-smigle


This copper ring (commutator) on which the brushes slide looks as if it is the same thickness as the tracks, are you not afraid that it will wear off quickly? After all, the clock runs 24 hours a day.

Download diagrams. As for these brushes, I think I have already said it, but I will repeat it again: yes, I am well aware that the clock will not work soon, but I will not despair about it. At least I will have an excuse to make another version, more perfect .

Buddy, could you provide a link or source where you put these schematics ?? Today I am a bit yesterday so if I missed something, I am sorry

Where documentation should usually be, which is in the first post .