I subscribe to this 100% It's just that the next copy is much better :)
Added after 39 [minutes]:
Well, I will not write that 05.kg of filament went into the trash, because the filament jammed on the spool, or there was a mistake in the dimensions of the solid, that the PCB was ok the third time, because the concept of the scheme changed, that the electronics housing had to be redesigned, because it came out too low and the plugs did not fit .. and others .. But that's the passion of creating...
And it's very good that you're writing about it now - let "Młody Talent" know that even when someone spends half their life "sitting" in electronics, such corrections and mistakes happen - that it's not as easy as it seems to some people. And you really need to have a passion in addition to passion to keep learning/training - mostly on your own mistakes.
I immediately associate topics such as "I will commission a design and construction (...), I have a budget of PLN 300"
Exactly. They approach equipment repairs in the same way - it doesn't matter that there is no documentation; After all, electronics is there to know. And if he says he can't, it means that he maliciously doesn't want to help.
I'll join from the other side. Well, because there is not much to do - done neatly and aesthetically. A lot of work. And taking this work into account, I'm thinking: instead of making a 40x16 display by hand, why not use an LCD TV of this size + any SBC computer? It's about the same price, and it's much cheaper if you factor in labor. The possibilities are many times greater, of course.
Well, you know - hobbies, fun, etc. But looking rationally, there is no sense We have lived to such times that there are few things that are really worth doing from scratch...
As for me, the image from a cheap LCD at dusk is simply poor. We all know the disadvantages of such displays. But it all depends on the preferences of the recipient. One does not fit, the other is enough. I don't even want to think about an OLED display in this size (price). I have "loved" LEDs since the dawn of my history Besides, with LED constructions, whether they are segment displays, or like this device, you can actually freely compose the proportions of the entire display. And so it is here too. With LCD screens, we are limited by their factory sizes and proportions. Here's another example. I have a weather station built of LED displays (dimensions 56 x 19 cm) hanging on my wall. Everyone, everyone who sees it for the first time admires it and asks where I bought it and how much I gave for it. There's something about these LEDs And I think it's worth doing uncommon and somewhat unique things.
Chair backs for proportions
Etching machine controller.
I may have written this before. I am glad that I have lived to such a time that I can build such constructions without looking at rationality. Because where rationality begins, DIY ends ... (of course there are some limits). Well, I can't imagine this spectrum on a small TV... But that's my personal preference
I recently saw something like this on the net and bought myself a roll of ws2812b, esp32 devkit v1 and a microphone for max4466. A few days ago I found half an hour to connect it and try to fire it, but it doesn't work unfortunately. The belt is functional, esp probably too. The LED flashes and the program is uploaded. The microphone also, the output is a sine wave consistent with the frequency from the generator. I used the program you provided at the beginning. By the way, you gave the wrong address there. The correct one is: https://github.com/s-marley/ESP32_FFT_VU A few LEDs at the beginning of the strip are turned on and later sometimes a few more are turned on somewhere further. What can I check? In the program at the beginning I changed only the number of bands and LEDs in the bars. I would try to upload your program, but I wanted to apply the stripes vertically, not horizontally. Combine them more classically. You got a great effect. I ordered a few sheets of white milky thin diffusing foil on alliexpress, but I haven't looked at what it will look like yet.
>>20461107 Is the mask really necessary? Assuming that the walls adhere to the fast and the diffuser will probably stick to the fast itself anyway?
I'm sure you did the tests "without the mask" and how did it look like?
If I was looking for such a diffuser, let's say 1 meter square, where would you recommend me to look?
The grille is 0.4mm thick and fulfills two tasks. The pixel is clearer. It also has slightly rounded corners. The walls are black which improves the pixel contrast. Then look at my scribbles #22. It is bent around towards the walls. The baffles and the focusing screen do not move sideways, because this fold holds everything together. The partitions are printed with two posts and it is 10 sets. I did not glue them together, because I expected a weak pressure of such a large grid. Each grille with two posts is pressed individually. As for the screen, I have not found anything larger than A4. I bought a white binder with a rigid binding as a test and it fit perfectly. It diffused nicely and did not suppress the light too much. That's why the screen is in parts. I can send you a piece and you can follow the trail of Polish stationery stores in search of something similar I don't have a picture of my spectrum without the mask, but it looks something like this: https://www.youtube.com/watch?v=Mgh2WblO5_c
A few LEDs at the beginning of the strip turn on and then sometimes a few more turn on somewhere further ...
I'm assuming you didn't make a matrix out of these strips, you're just looking at a loosely unwound strip. Then it looks like this. Make a matrix out of these strips, connect as in the diagram and everything will be ok. Besides, it is not enough to change the height and width parameters of the matrix in the config. You need to reassign the range of FFT fringes to each bar even further in the code. On the author's website (thanks for correcting the address) there is an xml file for calculating these values logarithmically for a different number of bars and audio band. This is the part of the code: the code for 20 bars.
if (i 2 && i 3 && i 5 && i 8 && i 10 && i 12 && i 16 && i 20 && i 25 && i 30 && i 36 && i 45 && i 55 && i 66 && i 82 && i 106 && i 137 && i 176 && i 227) bandValues += (int)vReal[i];
but I wanted to apply the stripes vertically, not horizontally. Combine them more classically.
Classic, or how? Say how you want to connect these bars, from which diodes the next bar should start and end. I don't know how it is supposed to look like for you, but I warn you that with such a connection (vertical bar). the minimum pixel height is 16mm (with a strip of 300 LEDs/5m).
I ordered a few sheets of white milky thin diffusing foil on alliexpress, but I haven't looked at what it will look like yet.
I'm assuming you didn't make a matrix out of these strips, you're just looking at a loosely unwound strip. Then it looks like this. Make a matrix out of these strips, connect as in the diagram and everything will be ok.
Correct. I just plugged it in to see if it would work. Of course, it doesn't look like a display, but the LEDs in the episodes should be blinking, and there was no reaction.
Besides, it is not enough to change the height and width parameters of the matrix in the config. You need to reassign the range of FFT fringes to each bar even further in the code. On the author's website (thanks for correcting the address) there is an xml file for calculating these values logarithmically for a different number of bars and audio band.
And I haven't read that. Today I changed back to 16 bars and there is already a response to the sound, so you are probably right.
Say how you want to connect these bars, from which diodes the next bar should start and end. I don't know how it is supposed to look like for you, but I warn you that with such a connection (vertical bar). the minimum pixel height is 16mm (with a strip of 300 LEDs/5m).
I wanted to cut the 5m tape into 20 equal sections of 15 LEDs and set the strips vertically. One whole bar is responsible for one frequency. The higher the intensity at this frequency, the more LEDs on the strip light up. Because you probably have stripes horizontally and each stripe contributes to each stripe? Yes, the distance between the LEDs on the strip is over 16 mm, and I wanted to place the strips every few dozen mm from each other.
I think that about 500 - 550 PLN should be enough for everything. I don't think I missed anything.
You have omitted the most important thing - business registrations Alternatively, gofoundme application and fundraising.
Almost everything you have buttoned up, patent rights design is open source. The audio department can be cut. Also make a cheaper version of pliss in the budget up to PLN 200 https://allegro.pl/listing?string=analizator%20widma%20audio Remember to include a typical 50% margin - PLN 1,000 per device. Include customs costs from China. And ensure continuity of supply Oh, and social media promotion. I recommend sending a sample to Techmoan for testing. It is best to immediately export it in Poland, it is a land of poverty - who will buy it?
You can set these strips vertically, only if you want to control them from one port, then you have to decide on the following connections: AND Whether b ? Generally, it doesn't matter how the stripes are arranged. Vertically or horizontally. Each LED can be individually addressed. And it doesn't matter where she is on the bar. It is the program that decides from which LEDs to create a bar in the case of a spectrum. You can even slant ... The horizontal arrangement of the stripes "zig_zag" is quite often used in factory LED matrices with the WS2812 driver, hence probably many encoders write for this control system.
Added after 2 [hours] 3 [minutes]:
If you still insist on vertical stripes, this requires a code change.
Generally, it doesn't matter how the stripes are arranged. Vertically or horizontally. Each LED can be individually addressed. And it doesn't matter where she is on the bar. It is the program that decides from which LEDs to create a bar in the case of a spectrum. You can even slant ...
Yes, I know that you can only have limited time to modify the program. But from what you write, it seems that the horizontal arrangement is more popular. And the connection A or B itself can be changed in the program under the SERPENTINE variable if I understand correctly. I'm looking at your schematic now. Will uploaded to esp32 work without this part with atmega8? You don't need to connect anything to EN, I will give audio from the MAX4466 microphone board to input 36, the output to WS2812b works without CD4050. I will connect the button, but RX and TX? There must be info there for the analyzer to start?
Atmega is not required for this to work. Everything should work without it. SERPENTINE is not used anywhere in the code. Yesterday I did a test on a small 8x8 matrix. Just swap x and y patterns. Fuction (matrix->drawPixel(x, y, .....) Then the bars are along the bar. Connection A
You have omitted the most important thing - business registrations
I don't plan on doing business...
... Also make a cheaper version of pliss in the budget up to PLN 200 ...
For this amount at today's prices, it's hard to tinker something. You can order a small ready-made spectrum from the Chinese. In general, there may be a mono/stereo version with a typical width of 440, 430 mm for audio "blocks".
For you. He reads the whole world and such typos are often mistranslated by translators and the text may lose its sense. Therefore, it is worth taking care of the spelling. Have you thought about using, for example, tracing paper as a screen? A few years ago I used road glass beads as a "distraction", I still have a few kg left and I need to make a half-meter 7-seg display. in the near future. I'm waiting for the materials.
The topic of the screen is an endless topic. I also did tests with tracing paper. I mentioned this in #22. For this device, with a "pixel" size of 30 x 11 mm, a distance of 18 mm from the LED and a 0.8 mm gap between "pixels", three layers are needed and it is a bit too thick. I often deal with illuminated pictograms in my constructions. Generally, the smaller the distance between the backlights, the thinner the screen must be so that the light does not escape to the neighboring backlights. With larger spacing, I etch the pictogram on a 0.4 to 0.8 mm laminate and it's ok. I read about glass beads on Elektroda. Maybe it was even your description. I even intended to buy them for testing, but somewhere this intention got stuck in time ... A white PVC screen of about 0.16mm gives very good results. The diffusion is very nice and the light loss is also very low. The problem is only with the purchase of a larger sheet and in the case of this spectrum I had to do it in pieces. Friend @ noel200 ordered "milk" foil from China. Maybe it will come out after trials A half-meter display with balls, it will probably weigh a bit ... and finding a screen for it is quite a challenge. Although here it is enough to reliably mask only segments.
The LED WS2812 itself is and probably the usual Atmega will handle it with a decent refresh rate, but if you add 1024 audio samples and 512 FFT bands, it's not so colorful anymore ... Current consumption depends on the displayed pattern. On the default color pattern, with maximum brightness, the power consumption is about 30 - 50 W (measured). With white pixels, it can go up to 100 W.