need some help with a little hardware hack

[aloishis89]

So you may have seen this on thinkgeek: http://www.thinkgeek.com/geektoys/japanfan/a620/

It is a USB owl that just runs off the 5v USB provides. I was going to get one for my friend, but thought that only moving around randomly was pretty lame, it needed some modification. So I thought it would be cool to make it start to move whenever you get a new email. At least then it has some functionality. My friend would get a little more use of it and I would get to try a neat project. I need some help decided which route to take to make this happen. The first option is to buy one of these: http://usb.brando.com.hk/prod_detail.php?prod_id=00681

It already has software that works with most major email clients and turns on lights. I figured that I could buy one, take it apart and desolder the LED(s) and install a DC to DC converter to bump the voltage up to 5v, then connect the 5v pin from the owl to it. I noticed that the light pulses when you get a new email, but I think a DC to DC converter would stabilize the voltage automatically right? Then my second option was to use an arduino micro and solder the 5v USB pin to an output on it and interface with it via a python script that I found online here: http://successlessness.blogspot.com/2008/0...-some-code.html

My final option is to just remake the original project that that code was made for here: http://successlessness.blogspot.com/2007/0...l-notifier.html

But since this is just to drive an LED, I would need a DC to DC converter here also.

So that is what I am trying to decide, obviously the cheaper the better but I also want to make sure that it is not too hard (I am only going to order 1, so I can't make any mistakes). Which do you all think would be the best to try, and would these even work? thanks, and sorry this post is so long.

[dallaskorben]

Maybe there is a way in software to enable/disable a specific USB port? In that case you could plug the owl into a USB port, and have a program enable that USB port when an email is received. But I don't know if it's possible to enable/disable individual USB ports, so maybe it won't work - just throwing that idea out there.

edit: doing some googling and found this post saying what I am suggesting is not possible - http://www.mail-archive.com/python-win32@p...g/msg02616.html

[aloishis89]

I was wondering whether or not it was possible to do that, but I guess it won't be that simple.

[dallaskorben]

The 'USB Webmail Notifier' from Usb.brando.com.hk (link in your first post) looks like it would probably be the easiest/cheapest way to do what you want. At the price ($17), it would be hard to build hardware from scratch to do what you want. (and most homebrew projects use parallel port or serial port for control, which many PCs don't even have anymore)

If you're looking for a hardware 'do it yourself' project, there are many ways to accomplish what you want. If you want to control the device by parallel or serial port then you can build very simple hardware to do so (basically just a relay). If you want to control the device by USB then you need to build slightly more advanced hardware (with a microcontroller that can talk USB).

The USB Webmail Notifier is a nice option because then you only need to plug one thing into USB (the webmail notifier), and the other device you want to power (the owl) can be connected to the webmail notifier for power. No need for parallel or serial ports, no need for external power. I don't think a DC to DC converter is what I would use, but it would be pretty straightforward to use a relay to control power to the owl.

[aloishis89]

The USB notifier was what I was leaning towards, but I have no idea what the circuit in there looks like. Would I be able to desolder an LED and use that point to power a relay which would connect the continuous 5v from USB to the owl?

[dallaskorben]

The USB notifier was what I was leaning towards, but I have no idea what the circuit in there looks like. Would I be able to desolder an LED and use that point to power a relay which would connect the continuous 5v from USB to the owl?

You should be able to do something like that, yes. (you may need a transistor or something between the LED connection and the relay, to provide the necessary current to switch the relay - same basic idea, just a detail of the circuit)

I'm going to order a couple of the USB notifiers to tinker with myself. I have a little electronics lab to tinker in, I can try to help you figure out how to wire it up. It shouldn't take more than a few dollars of parts (in addition to the USB owl and the USB notifier), but will probably require a small amount of soldering.

[aloishis89]

That's awesome, thanks a lot. I have experience soldering, but not as much experience with knowing which components to use. I will be ordering the owl and notifier very soon (like tomorrow). I guess we will not know more until we can see the inside of the notifier. Let me know when you get yours.

[dallaskorben]

I will be ordering the owl and notifier very soon (like tomorrow). I guess we will not know more until we can see the inside of the notifier. Let me know when you get yours.

I placed my orders for the owl and notifier tonight. (and a short browse at ThinkGeek ended in a $223 order. way too much neat stuff! lol)

[aloishis89]

$223?? Wow lol. While we are waiting for the stuff to come in, I came across this page on SparkFun: http://www.sparkfun.com/commerce/tutorial_...utorials_id=119

It is a really cool tutorial on how to use a relay to control large (120v) devices. With those schematics, you could control something much larger than a USB owl (email activated blender? lol). We just need this, only smaller. They do use a transistor to help control the relay and even provide schematics for a control board for the relay, which is nice. Can we do the same thing as in the tutorial only with components rated at smaller voltages?

[dallaskorben]

$223?? Wow lol.

They sell Astronaut Ice Cream. ASTRONAUT ICE CREAM! Soon, I'll be dining like a spaceman!

While we are waiting for the stuff to come in, I came across this page on SparkFun:

That is a nice tutorial, and it probably could be easily controlled by the USB email notifier. (but it's overkill for controlling the USB owl)

Can we do the same thing as in the tutorial only with components rated at smaller voltages?

I have used a MIC2981 chip in another project, and it worked well. It's only about $3 for the chip. I think it might work for this project also.

There are lots of simple circuits that could work for this. I think once we get the devices and open the USB email notifier up to see what its circuit looks like, then we can figure out the simplest way to control the owl.

[dallaskorben]

I got my ThinkGeek order today! Attached is a picture of the USB owl. (but I still haven't received the USB email notifier)

[aloishis89]

Awesome, I got my ThinkGeek order as well. However, it's at my house in Florida, and I have finals this week so I won't be down there until Friday to see it. I haven't received my email notifier yet either.

[adlerweb]

Uhm... DC-DC, Relay, ICs? Why not a simple optocoupler or transistor?

[aloishis89]

Ok, I got everything. I took apart the notifier and noticed that it was much smaller than I thought and it was much lighter than I thought. My camera is a crappy Sony CyberShot, so I couldn't get a good picture of the circuit board, but I noticed that it only has 1 LED that is a 4 pin SMT RGB. It looks very similar to this: http://www.ferret.com.au/c/Manuco-Electron...SMT-LED-n690032

It doesn't have a traditional IC controller, it has the same thing calculators do. I don't know what they are called, but it looks like a black dot of silicon. Other than that, it is just resistors and a capacitor or two. Its all surface mount, which means that soldering and desoldering is going to be a pain. I plugged it in and tried it out, and it seemed to work fine. I'm going to do more research on the use of transistors tomorrow (finished exams finally, so I have tons of free time now!) and see if I can find the pinout of the exact LED they use and take some voltage levels.

[aloishis89]

I think the pinout of the RGB LED is

1 - cathode (red)

2 - common anode

3 - cathode (blue)

4 - cathode (green)

Please correct me if I am wrong, but could I take an NPN transistor and connect the collector to the 5v pin on USB and the emitter to the 5v pin on the owl and the base to pins 1, 3, and 4 of the LED? The base may need a resistor to allow saturation, but since the same amount of voltage will always be put out by those 3 pins combined, would that work?

[dallaskorben]

Ok, I got everything.

Me too! But I left it at work and I'm at home right now. I'll get the photos later.

I took apart the notifier and noticed that it was much smaller than I thought and it was much lighter than I thought.

I thought the same thing! I was expecting something much bigger.

It doesn't have a traditional IC controller, it has the same thing calculators do. I don't know what they are called, but it looks like a black dot of silicon.

It's an IC under a glob-top encapsulant.

From Wikipedia IC packaging:

""blob on PCB" that attaches the raw die directly to a PCB, bonds the die wires directly to the PCB traces, then covers the die and the die wires with a blob of insulator;"

Instead of putting the silicon die into a plastic package, they put it directly on the PCB and cover it for protection. (it's cheaper)

Its all surface mount, which means that soldering and desoldering is going to be a pain.

You can make the connection to the resistor that is in series with the LED, as long as you don't overheat it (and pull the LED off the board) it should be pretty easy - there is plenty of space between parts at least.

Please correct me if I am wrong, but could I take an NPN transistor and connect the collector to the 5v pin on USB and the emitter to the 5v pin on the owl and the base to pins 1, 3, and 4 of the LED? The base may need a resistor to allow saturation, but since the same amount of voltage will always be put out by those 3 pins combined, would that work?

The R, G, B pins of the LED are not all at the same voltage level. Normally G is at 0v (to have green led on), when a mail is received it turns off G (sets pin to +5v) and pulls the B pin low. (it flashes the blue led, so actually the B pin is pulsed low, not just held low) After the new email alert completes, the B pin goes back to idle (+5v) and the G pin goes to 0v to turn the green led back on.

Using the B pin for control will mean that the output is flashed on and off.

Using the G pin works better - when the G pin is at 0v, we want output to be disabled, when the G pin is at 5v output should be enabled. (I tested it this way using a MIC2981 driver ic, and it worked well)

A USB-powered snowman (that lights up) worked fine when I had it controlled from the USB email notifier, but I had problems getting the USB owl to work. It seems like the owl is starved for power - I tested the control circuit using a 5v power supply and it worked well, but when I piggybacked 5v from the USB email notifier it had problems.

[aloishis89]

A USB-powered snowman (that lights up) worked fine when I had it controlled from the USB email notifier, but I had problems getting the USB owl to work. It seems like the owl is starved for power - I tested the control circuit using a 5v power supply and it worked well, but when I piggybacked 5v from the USB email notifier it had problems.

Did you take the 5v from somewhere on the circuit board, or did you piggyback directly from the USB cable? Is it possible that the owl needs a higher amperage to run?

[dallaskorben]

I think the pinout of the RGB LED is

1 - cathode (red)

2 - common anode

3 - cathode (blue)

4 - cathode (green)

I'm not sure how the pins are numbered on this kind of package, I think they are numbered like a regular IC. (pin 1 at top left, counting down the left side and up the right side, with last pin at top right)

What I found is:

pin 1 = R6 (75 ohm) (common anode)
pin 2 = R4 (0 ohm) (blue cathode)
pin 3 = R5 (0 ohm) (green cathode)
pin 4 = R7 (56 ohm) (red cathode)

[dallaskorben]

Some photos of the usb webmail notifier, for anybody interested in what it looks like.

Also a photo of the PCB, I labeled pin numbers of the LED and what they control.

[dallaskorben]

I put together a circuit with a couple of transistors, connected to the green cathode, and it seems to work well. When an email is received, the owl blinks its eyes and moves its head, when the alert is dismissed the owl stops.

The USB connector I used sticks out of the case a little bit because of poor planning - I cut the hole for the socket in a place where there's not quite room for the socket to fit inside. But it turned out to be not too bad.

Photos are on this flickr set.

I'll draw and post a schematic in the next few days. I used a 2N2905A PNP transistor to switch +5v to the USB socket, and a 2N2219A NPN transistor to invert the signal for the green cathode, so the USB socket has +5v when the green led is off.

(edited to correct 2N2219A NPN transistor part number)

[aloishis89]

Good to hear that you got it working. I am ordering the transistors and I can't find the 2N2919A NPN transistor anywhere. Do you know of a place that has them?

edit: Nevermind, it was a typo. It should be a 2N2219A NPN. It was correct on your flickr page.

[dallaskorben]

edit: Nevermind, it was a typo. It should be a 2N2219A NPN. It was correct on your flickr page.

Sorry, I totally missed that typo! I swear I even proofread before posting.

I still haven't sat down to draw a schematic of the circuit, I don't know if you need one or if the photos are readable.

I basically followed the circuits on this page: http://www.kpsec.freeuk.com/trancirc.htm - it has well-written and to-the-point explanations of PNP and NPN transistor usage. The circuit I used is the "PNP transistor switch" diagram, with the input connected to the circuit from the "A transistor inverter (NOT gate)" diagram. I connected the output of the PNP transistor to a USB socket (from Digikey, 95 cents each - part #151-1114-ND).

[aloishis89]

Yea I think I can just use the picture, let me make sure that I have it right. Referring to the first picture in the set, the red wire is going from the 5v pin on the board to the transistor to the 5v pin on the female usb connector (assuming the female USB connector is not upside down), and the black wire (the one NOT soldered to the usb connection on the board) is soldered to the green pin on the LED. The only thing that I could not tell was which order the transistors are in. Is the one on the closest to the USB connector the PNP or the NPN? Let me know if I got anything wrong.

[dallaskorben]

Referring to the first picture in the set, the red wire is going from the 5v pin on the board to the transistor to the 5v pin on the female usb connector (assuming the female USB connector is not upside down), and the black wire (the one NOT soldered to the usb connection on the board) is soldered to the green pin on the LED. The only thing that I could not tell was which order the transistors are in. Is the one on the closest to the USB connector the PNP or the NPN? Let me know if I got anything wrong.

The one closest to the USB connector is the PNP. This transistor will supply +5v to the USB connector when the base is pulled low (to 0).

The other transistor is the NPN, which is used as an inverter. The base is connected to R5 (LED green cathode) so that when the green cathode is at 0v, the output of the NPN transistor will be +5v (so when LED is green (cathode=0), the output will be +5v).

When the green cathode is at +5v, the output of the NPN transistor will be 0v, which is connected to the base of the PNP transistor and so the PNP transistor will be switched on - providing +5v to the USB connector.

[aloishis89]

Well, I got all of the parts in today and put it all together. When I tested it, it had a problem though. When a new email is received, it blinks, but it does not have enough power to move its head. I checked all of my connections and everything seems good. When I put my hand on it, I can feel the motor trying to move, but it seems like it is starved for power. Any idea why this would be?