USB Cable Hack to Power Coin Door Lights

[grafixmonkey]

I had to laugh a bit when I stumbled across this topic    As for how nuclear bombs work...who knows, as for how to rotory engines work...who knows,...as for electrons and circuits?

I FREAKIN KNOW!

There is absolutely 100% no reason why running 2, 4, 6, 8 or however many leds around your cabinet off of a 5V or 12V USB port or any other power source on your motherboard/usb ports will do any harm PROVIDED that:

1.  You KNOW what voltage is there.
2.  You KNOW what the current (Amps) is available.
3.  You KNOW what voltage your LED runs at.
4.  You KNOW what current your LED consumes.
5.  You make SURE that you have your LEDs (and each associated resistor) wired in parallel with your power supply - NOT IN SERIES!

Bottom line is each LED will consume its required current (say 20mA or .020A).  When you add up all of your LEDs, it better NOT be more than what's available current wise from that power supply(or USB port).  If it is...you will be F%$&d!

No, I'm not a smartass that read it in a book at radio shack (though you could).  You could say this is my area of expertise... I did get a degree in Electrical Engineering.

you know what, dude, you just might not be the only one here in electrical engineering.  I took all the way through processor design and transmission lines at the U of I before I opted to use it as a course sequence for a math degree instead, and go into computer graphics.  So maybe check and see if you really read something right before ya break out the cannons, huh?

Bottom line is each LED will consume its required current (say 20mA or .020A).  When you add up all of your LEDs, it better NOT be more than what's available current wise from that power supply(or USB port).  If it is...you will be F%$&d!

No.  That's wrong.  You drive an LED with a current.  If you look at a datasheet for one, you'll see that in its operation range dV is nearly constant for any dI that doesn't burn the thing out.  You must provide the correct limitation of current to the LED to keep it from sucking up too much.  It doesn't just "consume its 20mA" and neither is it likely to if you drive it directly with a voltage source.  I calculated the margin of error for not burning out a voltage driven LED from one bin in Fairchild to be in the hundredths of volts, assuming the device would draw current exactly according to its datasheet.
EDIT:  It's even more wrong than I thought, because if the LEDs are in series like they should be if you connect them to 12v, you don't add up their currents.  You calculate the total power consumption, sigma[V*I].

Also, I've done current measurements with IPB and coin door lamps from Happ, and they consume 160mA at 12V.  If a USB port could power this, you could run three 160mA devices before being in the red line.  Granted you can't plug a 12V lamp into a USB power source because there isn't a 12v pin, but if the 5 or 6v lamps are anything remotely similar, they should work too.  I also know of flashlight bulbs that sure as heck aren't drawing 160mA, because they're running on batteries, and they get a bright bulb out of as little as 3v.

And dude with the 8's in his name:
I know how hard a EE degree is to get.  So congrats on having one.  However, I also know that other people here don't have one, so you laughing at people for asking questions about ground loops and power consumption is kind of like me lording it over on you for not knowing how to draw a straight line between points in the Poincare model of hyperbolic space, or what the heck I mean when I say that having under-weighted points in a subdivision skinned skeletal system will result in improper results when the root joint translates.

Besides, I knew a lot of EE people in the U of I.  I lived for three years in a dorm where 70 percent of residents were EE.  They're one of the best schools for that, and I was frequently helping grad students to actually build their projects, because while they knew a heck of a lot about theory, they knew nothing when it came down to the actual devices.  So don't tell me you know everything just because you have the degree.

[grafixmonkey]

I know randy, but making such a fatal screwup isn't allowed when you are flaming people for not knowing what they are talking about.

Incandescent lights will NOT work properly under usb power, they pull 12v or more and a buttload more amperage, that was my point.  So you are absolutely correct even though you just elaborated on what I just said.  

They don't "pull 12v."  The light will not force the port to 12v, neither will it draw more current than it would if it were connected to 12v.  A 12v lamp connected to a 5v source would operate dimmer, with less current and less power consumption.  And, as I stated in my other post as well as on another thread (possibly this one but I don't feel like looking) the incandescent lamps I've tested draw well under the power supply limit for a USB port.  There are also lamps that do operate at the voltages a USB port can provide.  Examples are here:
http://www.happcontrols.com/lighting/91003700.htm
and here:
http://www.happcontrols.com/lighting/91204400.htm
ideal lamp numbers being #'s 44, 47, 86, 259, 447, and 555, all of which should operate quite nicely on 5V.  Happ does not list the power requirements for those lamps, so I can't guess how many would be safe to plug in without testing, but it's certain that at least one would work and be perfectly safe for the port.

[Carsten Carlos]

Don't wanna jump to much into this discussion, but regarding the USB-specifications, you can only drive 100mA from each port until the driver of your USB-device (that you won't have at this hack) asks for ok to demand 500mA.

Though it would work on many motherboards anyway, this is not the right way to do it.
Leave the USB where it is, and get +5V/+12V from your powersupply. You never know if you need more current anyway in future.

[matt888]

Bottom line is each LED will consume its required current (say 20mA or .020A).  When you add up all of your LEDs, it better NOT be more than what's available current wise from that power supply(or USB port).  If it is...you will be F%$&d!

No.  That's wrong.  You drive an LED with a current.  If you look at a datasheet for one, you'll see that in its operation range dV is nearly constant for any dI that doesn't burn the thing out.  You must provide the correct limitation of current to the LED to keep it from sucking up too much.  It doesn't just "consume its 20mA" and neither is it likely to if you drive it directly with a voltage source.  I calculated the margin of error for not burning out a voltage driven LED from one bin in Fairchild to be in the hundredths of volts, assuming the device would draw current exactly according to its datasheet.

Hmmm... I am NOT wrong!  It will consume it's current rating.  End of story.   You are also misleading others!  Higher than it's tolerance VOLTAGE blows an LED... NOT CURRENT.  You must provide the VOLTAGE limitation to the LED!  ie - if you want an LED that operates at 20mA/2V to run off a 5V USB port, you must DROP the voltage that you don't need across a resistor.  So, since V=IR, you have 3V(the additional voltage you DON'T want across the LED)=.020A x R.  Find R, so 150 ohm resistor is what you need to buy in this case.  Now there are some fluctuations in tolerances, but you will not find them to be an issue in this "light up my coin door case."
 

EDIT:  It's even more wrong than I thought, because if the LEDs are in series like they should be if you connect them to 12v, you don't add up their currents.  You calculate the total power consumption, sigma[V*I.]

AGREED  

Also, I've done current measurements with IPB and coin door lamps from Happ, and they consume 160mA at 12V.  If a USB port could power this, you could run three 160mA devices before being in the red line.  Granted you can't plug a 12V lamp into a USB power source because there isn't a 12v pin, but if the 5 or 6v lamps are anything remotely similar, they should work too.  I also know of flashlight bulbs that sure as heck aren't drawing 160mA, because they're running on batteries, and they get a bright bulb out of as little as 3v.

AGREED  

And dude with the 8's in his name:
I know how hard a EE degree is to get.  So congrats on having one.  However, I also know that other people here don't have one, so you laughing at people for asking questions about ground loops and power consumption is kind of like me lording it over on you for not knowing how to draw a straight line between points in the Poincare model of hyperbolic space, or what the heck I mean when I say that having under-weighted points in a subdivision skinned skeletal system will result in improper results when the root joint translates.

Seems as though you have read the original post like HC.  It was never meant to disrespect anyone, and was more of an entertaining way of posting on the dry material of electricity.  I apologize to those of you that have helped EE's get their degrees       - That's good stuff too!  

Besides, I knew a lot of EE people in the U of I.  I lived for three years in a dorm where 70 percent of residents were EE.  They're one of the best schools for that, and I was frequently helping grad students to actually build their projects, because while they knew a heck of a lot about theory, they knew nothing when it came down to the actual devices.  So don't tell me you know everything just because you have the degree.

As for the Poincare of hyperbolic space, sounds like a very exciting and stimulating topic also.  Perhaps we can get together sometime and discuss this...along with my current task that I get paid for: Determining the optimal OBSA (Off Bore Sight Angle) that the Aim-9X Sidewinder needs to achieve to beat the 4th generation MIGs after it and our F-15C hit the "merge" and begin the single or double circle fight.  See, I don't exactly fall into the "EE's have theory down, but no practicality" pot.  As the lead test engineer of the Aim-9x program at Nellis AFB, I have quite a bit of hands on action and real world theory tested also.  Sometimes good, sometimes bad, but the F-15C Eagle is a can of whoopass thanks to more than theory.

Uh oh, here comes a post from someone that designed the space shuttle  

Seems like alot of these posts get out of control.  This community/forum is an awesome place and I would not have 2 amazing MAME machines right now if it wasn't for the large collection of experience and knowledge around here.  My genuine thanks goes out to everyone ...  including Grafix and HC!  Can't we all just get along?

[grafixmonkey]

Bottom line is each LED will consume its required current (say 20mA or .020A).  When you add up all of your LEDs, it better NOT be more than what's available current wise from that power supply(or USB port).  If it is...you will be F%$&d!

No.  That's wrong.  You drive an LED with a current.  If you look at a datasheet for one, you'll see that in its operation range dV is nearly constant for any dI that doesn't burn the thing out.  You must provide the correct limitation of current to the LED to keep it from sucking up too much.  It doesn't just "consume its 20mA" and neither is it likely to if you drive it directly with a voltage source.  I calculated the margin of error for not burning out a voltage driven LED from one bin in Fairchild to be in the hundredths of volts, assuming the device would draw current exactly according to its datasheet.

Hmmm... I am NOT wrong!  It will consume it's current rating.  End of story.   You are also misleading others!  Higher than it's tolerance VOLTAGE blows an LED... NOT CURRENT.  You must provide the VOLTAGE limitation to the LED!  ie - if you want an LED that operates at 20mA/2V to run off a 5V USB port, you must DROP the voltage that you don't need across a resistor.  So, since V=IR, you have 3V(the additional voltage you DON'T want across the LED)=.020A x R.  Find R, so 150 ohm resistor is what you need to buy in this case.  Now there are some fluctuations in tolerances, but you will not find them to be an issue in this "light up my coin door case."

In which case, what you are doing is inserting an ohmic device that gives you control over the current (which the LED does not), and adjusting the ohmic device until the current is correct.  It's a way of converting a supply of a specific voltage supply into a supply of a specific current.  I've attached a VI curve from an LED datasheet at Fairchild.  http://www.fairchildsemi.com   I highlighted the operating range of the LED yellow, and the V and I operating ranges in red on the axes.  If you want to "give this thing 1.75v" and use its VI curve to drive it, you'd better be sure your power supply doesn't vary by more than 0.02 volts.  And if you measure your PC power supply, I guarantee you will not see exactly 5.00v on your meter.  

Another approach:  if what you say is correct, then I would be able to use a Zener diode of exactly 3.25v, hook it up in series with this LED, and it would work.  But I'm sure you'll agree that it won't, because now that you've removed the ohmic device that lets you solve for current, if your power supply does vary by a +0.1 volt, your LED will burn out, and if it varies by -0.1v it will be dim.

Seems as though you have read the original post like HC.  It was never meant to disrespect anyone, and was more of an entertaining way of posting on the dry material of electricity.  I apologize to those of you that have helped EE's get their degrees       - That's good stuff too!  

I guess it's an effect of seeing the replies on my way up to the post.  LEDs have been a frequent topic of argument on the board, since they are difficult to properly drive unless you know the theory.  There was a guy insisting that it was just fine to put a bunch of resistorless LEDs in parallel, with one limiting resistor for all of them, and was selling devices build this way.  While he may have found a part number with a shallower VI curve that allows this, it's asking for trouble if you try it with any old LED, or a mixed batch.  There were some other people too which were taking LEDs out of little LED flashlights, attaching them to a power source of precisely the same voltage as their batteries, and wondering why the things just instantly burnt out.  The reason they burnt out is exactly what I'm talking about with this new LED discussion.

As for the Poincare of hyperbolic space, sounds like a very exciting and stimulating topic also.  Perhaps we can get together sometime and discuss this...

Picture two people starting out within arm's reach of each other, facing perfectly parallel to each other, and walking straight forward (perfectly straight!) at exactly the same speed for exactly the same time.  Now one of them turns exactly 90 degrees to face the other person.  He can see the other person in front of him, 100 feet away.  That's hyperbolic space.  Weird huh?

My genuine thanks goes out to everyone ...  including Grafix and HC!  Can't we all just get along?

ohhhh, all right...  group hug!!!  

[JustMichael]

Personally the easiest way for most people to do this is to wire a resistor inline with each led to limit its current.  
First, find out what the maximum current is that the led can handle.  (example 20ma [same as .02A])
Second, how much voltage is available.  (example 5V)
Third, figure out how much resistance is needed to keep the led from getting too much current (example   Resistance =  Voltage / Current)
Resistance = 5 / .02
Resistance = 250 ohms

This simply gives is the smallest size resistor (or series of resistors) that you want to use inline with the led.  Larger resistors than the value found will result in a dimmer led.

[grafixmonkey]

yes...  the way you use the LED ends up being the same in almost all cases, we're just debating a rather technical point about it - that being that you must arrange a circuit such that it provides the proper current to an LED, not one that attempts to provide the proper voltage to one.  (although if you can control the voltage supply to the tenth or hundredth of volts, accurately, you can do the latter.)

[matt888]

Just Michael - yes!

Grafix- Do I get to post an LED spec graph with a MUCH steeper current curve?!  J/K  Also, there's no need to wonder if the voltage source is substantially off of say...a 5V USB port.  That's why man created a digital multimeter.  Now you've got THE VOLTAGE available and your resistance pinned down and the guessing only lies with how well the LED meets its tolerances/specs.      

I'm off to play MK2...  anyone else find it amusing in revision 2.1 to do the Babalities multiple times and cause the game to go in cardiac arrest?  

[grafixmonkey]

oh well...   I don't completely agree but you know what's going on and I'm sure everybody else stopped reading a loooong time ago.