Electronics and Embedded Programming V2

[ROBO_DONUT]

[QUOTE=Protocol7;31672946]I have an old broken harddrive stripped apart. Time to fuck with the motor :v:.[/QUOTE]
Good luck. I got nowhere with my back-emf sensing BLDC motor driver.
Funny thing is, each of the individual parts seemed to work fine. Zero-crossing detector worked fine, the constant acceleration spin-up routine seemed OK (after a little debugging), etc. But when I put it all together it just sat there like a lump of spaghetti.

Probably doesn't help that I don't have any of the proper tools...

[Chryseus]

[img]http://i.imgur.com/5dzeF.jpg[/img]

I was not expecting this when I lifted the mystery SOIC part, has to be one of the more
unusual things I've found.

[ROBO_DONUT]

wat.

[bobthe2lol]

I think I may have fried my arduino. Yesterday, I had it working nicely, powering my 10 meters of a total of 600 RGB Leds. Then, today I plug it in and the program that's in there is acting VERY strangely (really slow responce time, not doing correct colors, etc.) and I haven't touched it since yesterday. The power was unplugged (when I wasn't using it). So I think that maybe the program got corrupted.
So I go to re-upload it and here's exactly what happens (Arduino Uno): Click upload. Wait 1/2/3 seconds. L flashes. RX flashes 3 times in succession (100-300 ms apart).

Then the IDE says,
"avrdude: stk500_getsync(): not in sync: resp=0x00
avrdude: stk500_disable(): protocol error, expect=0x14, resp=0x51"

Any ideas?

[Chryseus]

[QUOTE=bobthe2lol;31676288]I think I may have fried my arduino. Yesterday, I had it working nicely, powering my 10 meters of a total of 600 RGB Leds. Then, today I plug it in and the program that's in there is acting VERY strangely (really slow responce time, not doing correct colors, etc.) and I haven't touched it since yesterday. The power was unplugged (when I wasn't using it). So I think that maybe the program got corrupted.
So I go to re-upload it and here's exactly what happens (Arduino Uno): Click upload. Wait 1/2/3 seconds. L flashes. RX flashes 3 times in succession (100-300 ms apart).

Then the IDE says,
"avrdude: stk500_getsync(): not in sync: resp=0x00
avrdude: stk500_disable(): protocol error, expect=0x14, resp=0x51"

Any ideas?[/QUOTE]

Make sure you are programming it correctly.
But it sounds to me like you've damaged it, powering 600 LEDs directly will almost certainly exceed the maximum pin current.

[bobthe2lol]

[QUOTE=Chryseus;31676685]Make sure you are programming it correctly.
But it sounds to me like you've damaged it, powering 600 LEDs directly will almost certainly exceed the maximum pin current.[/QUOTE]

So... How should I power them? I got the 12v 5a power supply from adafruit, and it has a connector that I don't think I should just wrap a wire around and stick in the led strip. And doesn't the board have overcurrent protection?

[ROBO_DONUT]

Maximum pin current is 40mA. Small LEDs can draw 20mA each.
600*20 = 12000mA.
12000 > 40

Hope you had a nice big FET driving those LEDs.
[QUOTE=bobthe2lol;31677224]And doesn't the board have overcurrent protection?[/QUOTE]
No, because you shouldn't be powering anything through an I/O pin. They're strictly for signalling.
You need a power transistor to switch larger loads.

[Chryseus]

[QUOTE=bobthe2lol;31677224]So... How should I power them? I got the 12v 5a power supply from adafruit, and it has a connector that I don't think I should just wrap a wire around and stick in the led strip. And doesn't the board have overcurrent protection?[/QUOTE]

You make use of a transistor like this:
[img]http://www.ladyada.net/images/tvbgone/transistordrive.jpg[/img]

[QUOTE=ROBO_DONUT;31677258]Maximum pin current is 40mA. Small LEDs can draw 20mA each.
600*20 = 12000mA.
12000 > 40
[/QUOTE]

They are probably in mixed series / parallel configuration with reduced current such as 4mA per branch.

[ROBO_DONUT]

There's no load resistor on that LED...
[QUOTE=Chryseus;31677302]You make use of a transistor like this:
[img]They are probably in mixed series / parallel configuration with reduced current such as 4mA per branch.[/QUOTE]
Using a 12V supply, you can have no more than 6 in series, which means 100 strands in parallel.
That's still well over 10 times the rated current.

[Chryseus]

[QUOTE=ROBO_DONUT;31677336]There's no load resistor on that LED...[/QUOTE]

Good job spotting that I just quickly grabbed an image from Google.
This is more like it:
[img]http://hades.mech.northwestern.edu/images/d/d6/Transistor_common_emitter.gif[/img]

Actual value for the base resistor depends on load, you can work it out as follows:
1. Find the maximum load current
2. Find the required base current (Iload / 10) {10 is a good low current gain as might be expected in a average transistor}
3. Using the base current work out the resistor size ((Vpin - Vbe) / Ibase)

[bobthe2lol]

[QUOTE=ROBO_DONUT;31677258]Maximum pin current is 40mA. Small LEDs can draw 20mA each.
600*20 = 12000mA.
12000 > 40

Hope you had a nice big FET driving those LEDs.

No, because you shouldn't be powering anything through an I/O pin. They're strictly for signalling.
You need a power transistor to switch larger loads.[/QUOTE]

Well, I'm not very good at electronics and math, but I did follow the tutorial, and it involved using 3 MOSFETS (whatever those are) and connecting the 12v line of the led strips to Vin on the board, so I wasn't powering them via the I/O pins. I belive I had the left line of the mosfet connected to the pwm pins for brightness modulation.

I wired it up like in the picture here:

[t]http://www.ladyada.net/wiki/_media/products/rgbledstrip/ledstripfet.gif[/t]

[ROBO_DONUT]

[QUOTE=bobthe2lol;31677501]Well, I'm not very good at electronics and math, but I did follow the tutorial, and it involved using 3 MOSFETS (whatever those are) and connecting the 12v line of the led strips to Vin on the board, so I wasn't powering them via the I/O pins. I belive I had the left line of the mosfet connected to the pwm pins for brightness modulation.[/QUOTE]
It sounds like you had everything set up right. Your board shouldn't be damaged. Perhaps it's a software issue?

[bobthe2lol]

[QUOTE=ROBO_DONUT;31677592]It sounds like you had everything set up right. Your board shouldn't be damaged. Perhaps it's a software issue?[/QUOTE]

What sort of software issue? The only thing I can think of is that the bootloader got corrupted, and I can't really fix that without some piece of equipment right (or another arduino, I kind of want a Mega... more pins)?

[Xera]

You're better off posting on the Arduino forums.

[ROBO_DONUT]

[QUOTE=bobthe2lol;31677655]What sort of software issue? The only thing I can think of is that the bootloader got corrupted, and I can't really fix that without some piece of equipment right (or another arduino, I kind of want a Mega... more pins)?[/QUOTE]
You can reprogram AVRs using the parallel port and three resistors.

[Lapsus]

[QUOTE=ROBO_DONUT;31678365]You can reprogram AVRs using the parallel port and three resistors.[/QUOTE]

Well, here's hoping that he actually has a parallel port on his motherboard, if that's the case.

@bob2thelol: do you get any different results if you unhook everything from the arduino and just use it from USB power with something simple like the blink sketch? It's a long shot, but it can't hurt.

[bobthe2lol]

[QUOTE=Lapsus;31679681]Well, here's hoping that he actually has a parallel port on his motherboard, if that's the case.

@bob2thelol: do you get any different results if you unhook everything from the arduino and just use it from USB power with something simple like the blink sketch? It's a long shot, but it can't hurt.[/QUOTE]
That was one of the first thing I tried. I thought, is something shorting out, or are some of the pins shorting, or doing something else wierd? So I unplugged everything, let it sit for a bit, then plugged in USB only and tried the blink sketch. Same result. :(

And nope, I don't have a parallel port. Heh... damn.

[Zero-Point]

[QUOTE=Chryseus;31674919][img]http://i.imgur.com/5dzeF.jpg[/img]

I was not expecting this when I lifted the mystery SOIC part, has to be one of the more
unusual things I've found.[/QUOTE]

Ah, the joys of SMD chokes.

[Chryseus]

Added new supplier to the list, [url=http://www.rshelectronics.co.uk/]RSH Electronics[/url] they have some good bargain packs and ship international for £4.00

[q3k]

Just a quick question: what would be the best way of powering a 5V microcontroller off a 12V AC power supply? Just build a diode bridge, slap on some capacitors and use a linear regulator (7805)? Or is there a better way?

[Chryseus]

[QUOTE=q3k;31692137]Just a quick question: what would be the best way of powering a 5V microcontroller off a 12V AC power supply? Just build a diode bridge, slap on some capacitors and use a linear regulator (7805)? Or is there a better way?[/QUOTE]

That's pretty much as simple as you can get.
If you have an op-amp and transistors laying around you can build your own regulator which would save you the cost of a 7805 if you don't already have one.

[ROBO_DONUT]

[QUOTE=Chryseus;31692181]That's pretty much as simple as you can get.
If you have an op-amp and transistors laying around you can build your own regulator which would save you the cost of a 7805 if you don't already have one.[/QUOTE]
You still need a voltage reference, which would generally be some sort of diode (preferably zener). You can also drop the op-amp circuit and use a single transistor like a voltage follower with the zener diode at the input. I think the op-amp probably adds some stability, though.

[DrLuke]

[QUOTE=Chryseus;31674919][img]http://i.imgur.com/5dzeF.jpg[/img]

I was not expecting this when I lifted the mystery SOIC part, has to be one of the more
unusual things I've found.[/QUOTE]

Gotta wonder what the machine building these looks like.

[Vbitz]

Looks like was just inserted nearly complete into the package and then sealed.
I don't know anything about IC production but that is just a guess.

[Chryseus]

[QUOTE=DrLuke;31705178]Gotta wonder what the machine building these looks like.[/QUOTE]

This probably:
[img]http://moreintelligentlife.com/files/china%20factory.jpg[/img]

[b]Edit[/b]
My new article on resistors and potentiometers is done, [url]http://chryseus8086.co.uk/?p=176[/url] hopefully someone will find it useful.

[Nikita]

Curiosity: how would you go about recording a radio-frequency unfilitered signal, gathering data from it inside a windows program, then generating a new arbitrary signal and outputting it?

Unfiltered as in, all frequencies, both input and output.

[ROBO_DONUT]

[QUOTE=Nikita;31725043]Curiosity: how would you go about recording a radio-frequency unfilitered signal, gathering data from it inside a windows program, then generating a new arbitrary signal and outputting it?

Unfiltered as in, all frequencies, both input and output.[/QUOTE]
You can't record raw wideband RF in real-time without some really fancy hardware.
You can, however, bring a specific band down to an intermediate frequency (IF) by mixing (or sampling) it with a local oscillator. Your IF can be low enough that you can sample it with an ordinary sound card. You can find lots on this topic by searching for software-defined radio. Jeri Ellsworth recently did a [url=http://www.youtube.com/watch?v=YFVgq3ZB0Mo]bunch of videos[/url] where she builds and debugs a quadrature sampling detector.

[Mr.T]

Is Farnell a good supplier? Or is it expensive?

[Ddrl46]

[QUOTE=Mr.T;31725246]Is Farnell a good supplier? Or is it expensive?[/QUOTE]

Farnell is great, but ordering there as a consumer can be a pain in the butt.

[Nikita]

Another thought: How hard would it be to use real meat as an actuator?