Go Down

Topic: rgb ledstring lightsaber (Read 13648 times) previous topic - next topic

racemaniac

#15
Apr 01, 2016, 08:34 pm Last Edit: Apr 01, 2016, 08:35 pm by racemaniac
Hi racemaniac,

Do you mind if I move this discussion over to the LED string saber thread? Now there is LSOS support for neopixel (not yet released, but I included a video, quite recent), I think it will give a new boost to the dicussion about this!
no problem, great that you're also supporting it XD
i'm also hard at work making the next version of my own saber, still using the stm32 microcontroller, and trying to push every ounce of power it has out of it (currently extensively experimenting with its hardware peripherals & dma, so i can do some serious multitasking on that bad little mofo :p).

My power supply is 4 nimh batteries cells (4000 mA each). Theese are special Radio Ccontroled car batteries (should deliver up to 20A)

At the begining I'm at 4.9V and at the end I was at 2.8V
i'm about to receive some of my own 144 leds/m ledstrips, so i'll experiment with them, indeed wouldn't surprise me if you got such a voltage drop (probably also the reason they are specced for 5V).
when having 144 leds, i'm also thinking about not putting them to full brightness, with just 60 leds/m i already had good brightness, so lowering their consumption a bit may also help a bit. But i've also yet to experiment with those strips. and if you're going for powering them on both ends, maybe also make an extra connection in the middle (or every 36 leds or so)? the wire is passing there anyway :).

bodtx

#16
Apr 01, 2016, 10:57 pm Last Edit: Apr 02, 2016, 08:16 am by bodtx
Ok so back with a new test, I've powered up both line, so I run a test for 8 minutes, no batteries problem, I run two 144 led with no voltage drop problem
so I run 288 led at full green end full blue.
it draws 4A

I've checked the voltage and I have a strange voltage drop that I don't understand, why near my batteriy I have 4.1V and why on the strip both end I have 2.7?
If I plug the arduino power on the breadboard near the strip it reset when voltage on strip is nears 3.3 (wich is the minimum for a Atmega328P@20Mhz

I thought connection in // does not suffer from voltage problem?
But that said it works correctly, but with a good voltage my strip would be even lighter

 see the schema (it is one led in it, but it is to compact the view, the 2 strips are with 144 led)


thx

racemaniac

Ok so back with a new test, I've powered up both line, so I run a test for 8 minutes, no batteries problem, I run two 144 led with no voltage drop problem
so I run 288 led at full green end full blue.
it draws 4A

I've checked the voltage and I have a strange voltage drop that I don't understand, why near my batteriy I have 4.1V and why on the strip both end I have 2.7?
If I plug the arduino power on the breadboard near the strip it reset when voltage on strip is nears 3.3 (wich is the minimum for a Atmega328P@20Mhz

I thought connection in // does not suffer from voltage problem?
But that said it works correctly, but with a good voltage my strip would be even lighter

 see the schema (it is one led in it, but it is to compact the view, the 2 strips are with 144 led)


thx
Could it be that your voltmeter is confused by all the noise these strips generate?
When building my saber i put several capacitors in between them and the rest of my project since i was getting a lot of noise on the powerline.

racemaniac

A small update of what i'm trying to achieve :).
As i mentioned before, i'm trying to get a single stm32 to do everything :). After lots of programming & learning, i've now got an stm32 reading the mpu6050 via dma transfers, and getting the orientation of the sensor, and the acceleration corrected for gravity :).
at the same time the program is driving an I2C DAC at 16khz and sending out a sinewave (that will later on become the sound), and with still plenty of cpu cycles to spend on other things.
All transfers are done using very lightweight systems, so only the math for the sensor is taking some(5%) of the cpu time, for 100 updates per second.
Next to do: read an sd card for getting sound samples, and drive the ledstring, probably also via DMA :).

racemaniac

Hi di you have a problem with the voltage drop in the led strip?
I'm doing the same thing with  144led/M strip and with powering only one side it drops at least 2V, I'll test powering both sides
my first 144leds/m strips arrived :). i'm wondering, do you guys also have strips that are composed of 2 strips of half a meter? (so in the middle you've got 2 leds that are a bit further apart).

bodtx

Yes there is some soldering i do not remember where. So the two leds have some more space between. But as i will put some cooking paper in the saber, specific led position won't be a problem

Protonerd

my first 144leds/m strips arrived :). i'm wondering, do you guys also have strips that are composed of 2 strips of half a meter? (so in the middle you've got 2 leds that are a bit further apart).
Yes, this is normal, my stripes also have some real solderings in-between.

Quote
I've checked the voltage and I have a strange voltage drop that I don't understand, why near my batteriy I have 4.1V and why on the strip both end I have 2.7?
If I plug the arduino power on the breadboard near the strip it reset when voltage on strip is nears 3.3 (wich is the minimum for a Atmega328P@20Mhz
Could it be the famous IR drop? If the LED stripe eats 4As, even 350mOhms of resistance over the connections will cause the 1.4V drop you measured. Try soldering instead of breadboard plugs and I hope that this problem will go away.

racemaniac

#22
Apr 22, 2016, 11:13 am Last Edit: Apr 23, 2016, 12:17 pm by racemaniac
Little update on my lightsaber progress: I've got functional frameworks for my maple mini for every part i need :)
- I2C dac i can drive with varying sampling rates (so it can reactly very subtly to even the slightest motion).
- 4MB Serial flash memory chip where i will cache my sound samples i fetch from an sd card (an sd card with fat32 is just annoying & slow to work with, so i'll cache it on an in between storage) -> i can use a standard sd card library that's too slow & blocking for my needs to fill the flash memory
- driving the neopixel chain via the spi port (using its dataline to generate the signal) -> got a demo running of updating a 144 led chain at 215 fps with hardly any cpu usage at all :). It's mostly just the DMA transmitting the data.
- reading the motion sensor via i2c

All of the above either work asynchronously via DMA, or are non blocking in nature, so i can do all of this IO simultaneously without having blocking calls anywhere, or the IO interfering eachother. (okay, i still have to test them all together, but it's starting to look really good :) ).

Now slowly moving on to the part of actually making all those parts work together as a lightsaber :).

At first i wasn't sure if i'd be able to squeeze all of that IO & interaction out of a single maple mini, but as i'm getting closer, it's really starting to work out :).

bodtx

Ok I've droped the breadboard and soldered strip directly on battery for test, it's working :)
here is an example with only one strip, it is already a strong green, I will do a test with two strips:

Lightsaber Test

I've added a little effect, there is a kind of light ring traveling on the tube, it is not a malfunction

racemaniac

Ok I've droped the breadboard and soldered strip directly on battery for test, it's working :)
here is an example with only one strip, it is already a strong green, I will do a test with two strips:

Lightsaber Test

I've added a little effect, there is a kind of light ring traveling on the tube, it is not a malfunction

nice :)
i'm also thinking of doing some effects :).

what are you using as tube around the leds? and how well does it look irl? can you see the individual leds or is it well blended?

bodtx

In the tube it is some baking paper:

racemaniac

In the tube it is some baking paper:

and does it work well? i know from experience that a camera makes any chain of light look perfect, but irl it's not always that perfect when seeing it with your own eyes :)

bodtx

Sorry i was not able to ├ędit my answer

IRL it is leSS bright
But it is quite bright with one strip
You cannot see the led if it does not touch the paper.
I have to build a centering ring to put every 20 centimeter in the tube

racemaniac

Sorry i was not able to ├ędit my answer

IRL it is leSS bright
But it is quite bright with one strip
You cannot see the led if it does not touch the paper.
I have to build a centering ring to put every 20 centimeter in the tube
i also thought about that path for a while, but i don't quite see it working. You'd also have to put something stiff in between the strips, or it'd just be chaos with your centering strips, and something small & rigid enough isn't so easy to find.
I'll probably stick to a layer of bubblewrap, it also immediately protects the leds against impacts, and also diffracts their light without losing too much brightness. The only concerns are thermal issues, but 2 60leds/m strips withstood a test of 15 minutes at full white (and then i was starting to feel some heat build up, but they probably could have lasted for quite some longer). but i'll stick to that option for now :).

lakron

Hello.
I have arduino nano, led strip 60 led/meter with 100 leds, and 18650 battery.
How to power the arduino and strip?
For test can i power arduino from usb, and connect strip to 5v arduino?
For production power strip directly from 18650 battery and arduino from 18650? Or for arduino i should use 2 18650 batteries, because input voltage for arduino 5-12v?

Go Up