Project: Blitter Bike

[junglesmacks]

Ah haaaaa.. it's because you're using such a low source voltage! Sure enough.. according to the LED wizard (http://led.linear1.org/led.wiz) you are indeed pulling 1920mA assuming a forward voltage of 3.2v per LED.

But.. check out if you were able to run 18v source voltage with 5 LEDs in series.. you're power consumption would only be 400mA.

I never thought about this dynamic of less power/more efficient power usage by using a higher source voltage and having longer strings together.

I think you're confusing power consumption with current. The power consumption remains the same if you're targeting the same amount of per-LED current with sufficient forward voltage. Only in power lines does it really make sense to use high voltage instead of large amounts of current for transmission purposes. The strips use minimal voltage because each channel of each LED is independently controlled. You may choose to use LEDs in series rather than in parallel if you're worried about the current carrying ability of the wires you're using, as well as only needing one current-limiting device; however, it may mean that your LED string as a whole stops functioning if one goes out.

I meant current, yes. Typing too fast.

The main reason that one would take current into account in an application like this would be the current carrying capacity of the components/controllers involved. For instance.. the MOSFET board that I'm using has a max of 4A per channel due to the sizes of the board traces and other factors. That's why personally I'm interested in my own current load.

Another good reason when using a large array of LEDs to use a higher source voltage with a greater amount of LEDs in series is the overall wiring work load and simplicty of design. I would much rather wire together 400 LEDS together via 80 strings of 5 LEDs with a resistor vs 400 strings with 400 resistors.. Sure, if one goes out then my whole micro string of 5 goes with it, but that's A) easily fixable B) unlikely to randomly happen as long as connections are secure C) almost a moot point if soldering to a PCB.. which.. is exactly what I'll be doing from now on.

[BBadger]

Sorry junglesmalcks, I didn't mean it to sound like you didn't know your shit.

Yeah... I know first hand just what a pain it is to have to wire up so many damn resistors to individual LEDs. Such a pain in the ass.

Are you going to use PCBs or protoboard? Fabbing PCBs can be pretty expensive, and rarely can you get super bulk rates of ten cents per square inch deals; protoboard might work better if you just want dense grids. Also, have you looked into wire-wrap instead of solder? It is more vibration resistant and quite fast for connecting things up.

[waxpraxis]

I'm working with junglesmacks to cut some boards on my CNC.



If you get the LED bug let me know - happy to help out fellow burners who have LED erm... issues.

[BBadger]

Oh nice! Is that the flexible kind of substrate, or stiff?

[junglesmacks]

Copper clad custom milled on a waxpraxis CNC, baby

Yeah.. PCBs only next year! Forget flexible backings.. I'm going bulletproof next time around!

[waxpraxis]

Yeah, I've only tested rigid material, but at some point I'd like to get some pyralux to try out. I've had a lot of success milling other semi-soft plastics (HDPE, etc).

[BBadger]

Yeah, I've only tested rigid material, but at some point I'd like to get some pyralux to try out. I've had a lot of success milling other semi-soft plastics (HDPE, etc).

If you can polarize that HDPE you could make it into a piezo-electric sensor too!

[junglesmacks]

at some point I'd like to get some pyralux to try out.

I had to google this.. I WANT!!!

[waxpraxis]

Yeah, it looks like fun!

Watched some videos today about SMT soldering... Given that with my CNC machine I have to do a tool change to handle drilling my through holes SMT is starting to look like a better and better idea. Time to start practicing!

I finally ordered all the components I need for the voltage regulators and have started sketching up the circuits I'm going to need to handle talking to all of the LEDs. The voltage drop across all 32 strands will be such that I think I'm going to have to segment it just like I'm doing with the power - 8 segments of 4 strands each. Folks on the adafruit forum have recommended that I just demux the clock signal and that'll do the trick. I've got another meter of lights and an Arduino Mega on the way to test that theory this weekend.

[waxpraxis]

Just sat down and did my first surface mount desoldering (and soldering) project. The desoldering, as expected, was a royal pain in the ass, but the actual soldering wasn't bad. I found a decent video on YouTube that shows a few different techniques (most of which fly right in the face of what I was taught to do with doing through-hole soldering!) and it was pretty easy to pick up.

Up until now I've been avoiding surface mount stuff, but I may just get into doing it since it'll be easier to make surface mount boards with my mill (no tool change required).

In case anyone's interested:

[youtube][/youtube]

[junglesmacks]

Very, very timely post. I was just having a back and forth with Jaycerochester about this. I have a good assortment of SMTs on order from Mouser to play with and I'm unclear on how to solder/mount/use them as of yet. Can't wait to get my hands on them next week. I'm thinking SMTs + Pyralux could make some pretty wicked custom LED strips..

[waxpraxis]

Hell yeah!

BTW, I finally said fuck it and ordered a relatively small (12" x8") and cheap ($75) vacuum table for my CNC (it can be powered by my shop vac, score!). I think that's the only way I'm going to get really accurate depth for milling these PCBs.

[BBadger]

Watched some videos today about SMT soldering... Given that with my CNC machine I have to do a tool change to handle drilling my through holes SMT is starting to look like a better and better idea. Time to start practicing!

Soldering SMT chips is relatively easy with solder masks and a programmed toaster oven (or you can do it manually too). You could even make your own masks rather than pay for them. You still need to manually pick-and-place the components, and double-sided boards can be a pain. I'm also not sure if it will work with pyralux with current DIY setups.

[waxpraxis]

I've seen the toaster oven "trick" before, and I had always assumed that was just the way you had to do it. I was honestly surprised at how easy the SMT soldering was when I used a generous amount of flux paste and used the solder-on-the-tip technique in that video. I held down the component with tweezers, touched the corner pins and I was off to the races!

On a side note I'm not sure if I'm happy or not at how thorough I'm being with the accounting for this project - just on all of my new tools, sample strips, arduinos, etc, I'm already over $450!

Ah well... a lot of it included tools and other resources that will certainly get used in other projects and are really long-term investments. That, and usually when I needed a particular chip or other component I bought 5-10x as many as I needed to cover learning curve, breakage, etc. I mean, why get one $0.68 multiplexer IC when it's such a commonly needed part?

[junglesmacks]

Dude I feel you. I've got bags of LEDs and un-done mini DIY boards laying around from the light suit last year. Everytime you get something online to use, by the time it gets there you've had a better idea and go in a different direction. So much trial and error and refining with all of this stuff! I mean hey.. I just spent $45 on a sheet of Pyralux that I don't know if I'll ever be able to use.. lol.

[waxpraxis]

So Mr. Junglesmacks is acting as my PCB milling guinea pig...

So far, so good!

[junglesmacks]

Mmmmmmmm.. 2"x4.5" panels for 50 LEDs. I wonder what say.. 48 or so of those suckers strategically placed on clothing would look like while computer controlled. Hmmmm..

[waxpraxis]

Hell yeah!

Does your controller have a PWM per channel? Cause yeah, I could see all sorts of crazy shit if you could fade them in and out in addition to just on and off!

[junglesmacks]

It has not necessarily PWM but % of voltage control within the sequencing.. that's one thing that makes the patterns look so cool and sweeping instead of blocky. As in.. for a simple chase the lead channel is 100%, following is 50%, last is 20%. It's all open source hex programmable but I haven't wrapped my brain around it enough to do custom patterns yet. A bit much on the plate right now.. whew..

[waxpraxis]

Ah, ok. Does the output of LEDs related to the voltage in a linear fashion? I mean as long as the forward voltage is over 0.6v you should get some kind of glow, but I've always heard if you want to dim LEDs that a PWM is the way to go.

[junglesmacks]

Ah, ok. Does the output of LEDs related to the voltage in a linear fashion? I mean as long as the forward voltage is over 0.6v you should get some kind of glow, but I've always heard if you want to dim LEDs that a PWM is the way to go.

Yeah.. PWM isn't the only way to dim them.. you can just lower the input voltage. In fact, that's how I'm going to be adding an overall brightness control on my suit for next year: a simple dial potentiometer attached to the + side of the v input to the board.

Now.. if you wanted to selectively dim certain addressable LEDs within your array.. then yeah, PWM. If you want to a rudimentary dimming device for your entire array.. then just increase the resistance on the + side to lower the input voltage.

[waxpraxis]

Well, where you put the pot is unimportant as long as it's still in series, but yeah, I get what you're saying.

I'd be interested to find out the relationship between current and voltage of an LED. Something tells me it's not linear, but that's somewhat because everytime I think something is simple with electronics I fall down some new rabbit hole!

[BBadger]

Well, where you put the pot is unimportant as long as it's still in series, but yeah, I get what you're saying.

I'd be interested to find out the relationship between current and voltage of an LED. Something tells me it's not linear, but that's somewhat because everytime I think something is simple with electronics I fall down some new rabbit hole!

It's definitely non-linear, especially the transition around the forward voltage, which goes from passing virtually no current (I = 0 when V < Vf) to passing exponential current (I = e^(kV) when V > Vf). Most people just assume that you provide at least the forward voltage and then apply (regulate) whatever current is needed for the brightness (without burning the things out). When powering things like laser diodes beyond their spec (like hundreds of milliamps), you can definitely see a change in forward voltage as more current is applied. For specific LEDs, pull up its datasheet and look at the forward current vs. forward voltage graphs.

[waxpraxis]

Thanks, that actually clarified a couple of things for me! It's funny cause I just realized I wrote the relationship between voltage and current and what I MEANT to write was current and perceived (or actual) brightness. Derp.

Anyhoo, I finally got in all of my capacitors, coils, etc I need to start building my power supplies for Blitter Bike this week. I just got done throwing together a quick test of my circuit on a breadboard and I'll be damned - it works! I got it hooked up my my 12v sump pump backup battery and out came perfectly smooth 5v that the RGB LED strips just loved - woo hoo!

[BBadger]

Thanks, that actually clarified a couple of things for me! It's funny cause I just realized I wrote the relationship between voltage and current and what I MEANT to write was current and perceived (or actual) brightness. Derp.

For that, pull up the lumens graph in the datasheet for the LED. Lumens capture the perceived brightness of light. So for example, a green LED will have a higher lumen rating than a red LED, for the same radiant power, because of the eye's greater sensitivity to green wavelengths.

Another thing to consider when sequencing your LEDs is to use opponent colors to enhance the perceived brightness of certain colors.

Anyhoo, I finally got in all of my capacitors, coils, etc I need to start building my power supplies for Blitter Bike this week. I just got done throwing together a quick test of my circuit on a breadboard and I'll be damned - it works! I got it hooked up my my 12v sump pump backup battery and out came perfectly smooth 5v that the RGB LED strips just loved - woo hoo!

Did you build a bucking circuit?

[waxpraxis]

Good to know about the lumens. I'm really just screwing around with RGB LED strips so I don't have to directly worry about laying out the individual LEDs, but that's really good to know. A few years I did a bunch of work with sound reactive code and had to do a lot of fun math what with FFTs and taking into account psycho-acoustics, etc - which I at least had an idea that perceived brightness wouldn't be directly tied to current. Then again, I don't think you generally do too poorly assuming that things are more complicated than they look!

And yes, I built a step-down switching regulator. I'm barely hanging on in terms of my understanding, but as I did my research it looked like switching regulators, while more complicated, draw quite a bit less current than other options. Since I'm trying very hard to keep battery life up and battery size/weight down, it seemed to be the best choice.

[waxpraxis]

Finally figured out Eagle enough to get a decent board built. I'm not entirely sure the feedback line coming off of the IC is separated enough from the inductor (the data sheet on the IC says to keep it away from the inductor due to potential issues with flux), but I figure it's at least worth a shot. Eagle is a bit odd at times, and I'm sure my design is far from optimal, but I'm pretty happy with this as first pass!

[BBadger]

Finally figured out Eagle enough to get a decent board built. I'm not entirely sure the feedback line coming off of the IC is separated enough from the inductor (the data sheet on the IC says to keep it away from the inductor due to potential issues with flux), but I figure it's at least worth a shot. Eagle is a bit odd at times, and I'm sure my design is far from optimal, but I'm pretty happy with this as first pass!

(image)

Before you dive into laying out the board for a DC-DC converter you should read some of the design guides and application notes for DC-DC converters. Those traces you have in the photo there look too long, and there is a lot of thought that must go into making a decent DC-DC converter on a PCB because they operate at such high-frequencies where trace length, ground planes, etc. are important to consider and plan for.

Some refs:

Reducing Ringing Through PCB Layout Techniques

Building a DC-DC Power Supply that Works

Application Note 1229 SIMPLE SWITCHER PCB Layout Guidelines

[waxpraxis]

Thanks - I've been reading a lot of the same docs. I'm using one of TIs SimpleSwitcher modules since one of it's major "selling points" is that you don't have to be quite as careful with the PCB layout. As best I can tell I'm sticking to the layout guidelines they've provided.

I'm also well aware that this is my first board layout, it's not a simple one, and there's a good chance it ain't gonna work.

[waxpraxis]

Well I'll be damned, it did work!

I need to track down a scope to make for certain, but my analog and digital multimeters are reading a clean 5v and it powers up both my Arduino and the RGB LED strips! That's not to say it it won't catastrophically fail later, but for now... IT WORKS!