Statistics project

Alright, I think I have already posted this once and gotten a response, but I then forgot my password and the location of this site, so please don't get angry with me anyone if you happened to be someone who answered my post the first time.

ok, so, for my statistics class we have to do this giant year long statistics project in which we gather numerical data and then use various statistics methods to interpret it, etc. Well, we were encouraged to do some type of experiment, and since I like electronics, I decided to build different types of DC-DC power conversion circuits and compare them. I haven't really worked with these before though, so I don't really know much about them. I'd like to convert 9v to 5v using different types of circuits, so does anyone know of different types I could build and where I could get the schematics. I'm a highschooler, so my knowledge is limited. Thanks for anyhelp you can give me.

Alienminds said:

Alright, I think I have already posted this once and gotten a response, but I then forgot my password and the location of this site, so please don't get angry with me anyone if you happened to be someone who answered my post the first time.

ok, so, for my statistics class we have to do this giant year long statistics project in which we gather numerical data and then use various statistics methods to interpret it, etc. Well, we were encouraged to do some type of experiment, and since I like electronics, I decided to build different types of DC-DC power conversion circuits and compare them. I haven't really worked with these before though, so I don't really know much about them. I'd like to convert 9v to 5v using different types of circuits, so does anyone know of different types I could build and where I could get the schematics. I'm a highschooler, so my knowledge is limited. Thanks for anyhelp you can give me.

Click to expand...

Well.

Get ready to google!

Look for "7805 regulator" (or LM7805 regulator) follow links, eventually you will get to a circuit example. This is the "linear" way of regulating from 9V to 5V. It is very inefficient but low cost and simple in design.

Here is another method:

Look for "5V buck converter" or "DC/DC buck converter" you will find many links and most will lead you to the general topology of this method.
This method is a "switching" type regulator. Much more efficient than the linear way but as you'll see, it is more complicated in design. Chances are you will not find an example for 9V input to 5V output (although you might!) but that is ok. If you can get to a company that sells one of these "Switching regulator" chips, their datasheet will have an example and enough explanation so that you can come up with your own circuit design that will be 9V to 5V and if you have trouble just post your questions..

Another statistically "significant" method would be to stack solar cells in series (as close to 5V as you can get)

First, give it a good go at google and see what you come up with. Everyone on this forum is very helpful. Try and see what you come up with!

I'd recommend doing plenty of planning before you start building. It sounds like you could loose sight of the 'statistics' focus if you aren't careful. I think I responded to your earlier post with similar worries.

Part of your planning might be to think yourself and work with the forum. The first step might be to come up with three different methods of converting DC to DC. Next might be to ask the experts what the differences might be between them. No reason for you to 'discover' everything if it's already known. Then you might make a plan to measure those differences. A very important step here is to determine what you will do with the data. Again - all of this is planning. The planning part may seem tedious but it's imaginary, in a way. You pretend you've done things - rather than consume the time and resources. If in your planning you can't see your way thru to the end you can make adjustments - so you can get to your goal. My worry is that you'd do a lot of construction that would not get you to your goal.

As I wrote this I had thought of some things you could do, that would require some electronics but would be less dependent on complicated electronics.

A. Some people like to run LEDs so they are brighter than normal - the problem is that the life is reduced significantly. You might construct a regulated power supply then test banks of LEDs at various levels of overvoltage. You'd have to shop (we'll help guide you) to get a bunch of the same LEDs and dropping resistors. What you'll do is record the point of failure - at a given voltage or current. All of them won't fail at the same time and your statistics will help you to describe that. You'll probably see that as you push the voltage or current up that failure occurs sooner. To add some spice to this you might create a means of pulsing the LEDs. A low cost version of this might be to push 1/8 watt resistors to failure at various power levels. These resistors can be had by the reel at low cost - if you shop carefully.

B. A variation of 'A' could be to gather data from observers on the attention getting qualities of various pulse rates or brightnesses.

C. You might ask friends to give you their discarded batteries. You could build a constant current supply (or several) then drain the batteries to nothing. The data might help reveal patterns about people - and how much life is actually left in batteries that they discard.

D. You might purchase some packs of batteries (on sale) and with your electronics skills, build power supplies to discharge the batteries at predictable rates. You'll see some variation - possibly between brands. Maybe the best bang for the buck will be apparent.

E. Another spin on batteries - compare the life of rechargeable batteries at various rates of charge or discharge. You might take 10 batteries from the same mfr and see how they behave with all conditions being the same. That might be your baseline data. Then vary things, one at a time, to see what happens.

Note that in all cases, if you are making measurements, that there will always be some variation - if nothing else, from the measurement process itself. If you aren't seeing variation it's a clue that the resolution of your measurement process is is flawed. A project by itself could be to take the measurements of the same thing with several instruments over time. You'll see differences and maybe discover that one instrument varies more with temperature than another.

Anyway, hope this helps. Main advice is to make sure that you don't loose sight of the reason for doing the work. I'd keep the electronics simple so you can learn about statistics. Best of luck.

i like suggestion E
that would have been my choice..
yuo could measure the effects of 'NI-CAD memory' and see wether compleatly discharging them really helps them or not..