The first batch of DrinkShield boards showed up on my doorstep last night. We will be assembling a few kits and will start working on tutorials and preparing for classes. Once the online store is fully tested we will start selling there as well.
DrinkShield Boards
6 Comments
It looks like those are 220 ohm resistors. Did you know that each pin of the 74HC595 can only source 6mA? If the LEDs are typical with a voltage drop of about 1.7V, then you are drawing 15mA from each pin of the shift register. This will burn out the chip after a little while. And the total current through the 74HC595 cannot exceed 70mA, but with all the LEDs on you would be pulling 120mA. You really should be using larger resistors.
And the MQ-3 heater pulls 150mA (I have one so I measured). You really don’t want to power this device using the Arduino voltage regulator. The regulator can source 800mA which is a lot, but think about the power dissipation. WIth 150mA for the heater, and up to 165mA for the LEDs, that’s over 300mA. You should never be pulling more than 200mA through an Arduino. Most people say 150mA is a better limit to use. Isn’t your Arduino’s voltage regulator getting REALLY hot?
Just trying to help.
I’ve been running the sheilds for a while and haven’t noticed the voltage regulator getting hot. The longest I’ve had it on is about a day and a half straight at the Next HOPE Conference. The MQ-3 Gets a little hot but that’s all I noticed. They are 220 Ohm resistors but I don’t believe there is an issue with them in this configuration…they actually may not even be needed… I should be able to just run the LEDs directly from the micro w/o a resistor, similar to other kits. I’ll keep my eye on it and see if any users have any issues with the current configuration.
Also I just posted the eagle files on http://drinkshield.googlecode.com/. Feel free to check them out and comment on them. I appreciate the feedback!
Hmm. Why aren’t you concerned about the values of your current limiting resistors? The maximum current you can source from the shift register pins is 6mA, and the maximum for all pins combined is 70mA. Like I said, you are allowing 15mA to flow through each pin which is way over the maximums stated in the datasheet. The shift register chips will fail if you run them like this. It will take some time, but they will fail.
You ALWAYS need a current-limiting resistor with an LED. If you don’t use a resistor, what will prevent a HUGE amount of current from flowing from your shift register? You will burn out the LED, but even worse, you’ll burn out the pin sourcing the current. What other kits don’t use current limiting resistors? The most common way for someone to burn out an arduino pin is to connect an LED without a resistor. An arduino pin can only source 40mA (but you are limited to 6mA because you are using shift registers).
I still would never source that much current through the Arduino MC33269 voltage regulator. A couple hundred mA will make it pretty hot and it’s life will be shortened. Remember that everyone paid about $30 for their Arduinos.
Just remember, that damage to semiconductors happens over time. When you exceed the limits stated by the manufacturer, it does damage and causes the component to fail. I’m just trying to help because I’ve had the same concerns with my own projects/products.
Here’s some stuff to read:
http://www.thebox.myzen.co.uk/Tutorial/LEDs.html
http://www.hebeiltd.com.cn/?p=zz.led.resistor.calculator
I received my information about the LEDs from Jimmie Rodgers when we was in town and I assume he is basing it on such projects of his like the LOLshield (http://jimmieprodgers.com/kits/lolshield/) and his open heart kit (http://jimmieprodgers.com/kits/openheart/). I’m not as concerned about the LEDs as I am about what you are saying about the voltage regulator. Since this shield was primarily built to work with the DrinkDuino game it is usually powered via the USB cable. Would this make a difference? Would this be why I don’t notice anything getting hot?
I originally made this project for me and my friends, then others wanted one so I made a kit. I am totally open to suggestions and if I can learn how to make it better I will.
By causing damage over time? How much time are we talking here? Like, if you ran a breathalyzer game every day the lifespan would be cut in half? Better than that? Worse than that?
I think I have the device on longer than most because I spend time developing the game and I have not seen an issue but then again I created the game just a few months ago. If this is a realistic concern for such a limited use shield then I would like to post a warning on the page to let people know not to run it for extended periods of time or something.
Don’t worry about criticisms the kits are a evolving open source idea and if there are improvements we will definitely use them!
You should ask Jimmie why he does not have any current limiting resistors on the LOL shield. It may have something to do with charlieplexed LEDs, because the current draw at any given instant is lower. But you are not charlieplexing, so you DO need current limiting resistors of the right size. If you just ship 470 ohm resistors instead of 220, then you will reduce the probability of damage to the shift register chips. Do some research on this topic — this is a fundamental principle. You need the resistors, and there are many calculators to help you chose the right value.
Voltage regulator: no, powering from USB makes no difference. Your shield draws a lot of current, and you need to measure it. Use a multimeter to measure how much your running shield is drawing when all the LEDs are on. It could be over 300mA, which is a lot. Even though the voltage regulator can handle 800mA, the real limiting factor is power dissipation. Read this: http://www.thebox.myzen.co.uk/Tutorial/Power.html. The vreg needs to be able to dissipate the heat. I would think that your vreg is pretty warm to the touch. It’s the little chip next to the power jack. Have you touched it when your shield is on?
I’m sure you saw this: http://www.danielandrade.net/2010/03/07/building-an-breathalyzer-with-mq-3-and-arduino/. Notice that they said to use an EXTERNAL power supply for the heater. You don’t want to run a heater off of the arduino. If you included a dedicated voltage regulator on the shield, then you would not stress the Arduino board. Connect VIN to a 7805 regulator in a T-220 package and power the heater from that.
It’s hard to say how long it will take to damage semiconductor components. The point is, the damage happens over time. The junctions in the semiconductor get weaker over time. The damage is cumulative. It doesn’t matter how long someone uses the shield in one particular session. The damage will accumulate.
Also, it’s a good practice to use decoupling capacitors on all ICs. Your shift registers don’t have them. There should be a .1uF capacitor right next to the VCC pin of each chip. One lead connected to VCC, the other to GND. You can do some research on this concept, too.
This is a lot of fantastic information! Thank you very much! I will definitely look into an external regulator and while I’m redoing the board design I’ll add the capacitors. Well, that’s if I keep the shift registers. Might try to experiment with the charlieplexing…not because of this discussion but I’m still fascinated by that and it would be a good way for me to learn it
Thanks again for all the info!