posted
I'm doing a statistical analysis on resistances, and I noticed something interesting. Nearly all of the values I got with a digital ohmmeter were below that of the stated resistance. I'm planning to chalk this up to bias in the equipment in my report, but I'm wondering if anyone is aware of the makers of resistors erring on the lower side of the stated value.
I swear, this isn't necessary for the report, and I'm not asking the forum to do my homework. I'm just curious.
Also, if anyone would like to recommend a good book on statistics, I'd appreciate it. I can use the formulas, but I don't think I have a good conceptual intuitive grasp of the topic. I might try to do some reading over the summer so I actually can understand the material I learned this semester.
posted
What kind of resistor (it's not a decade box, right)? How are you measuring it? What the tolerance on the resistor (is it within tolerance)? There are many crappy ohmmeters around, so that’s a choice too.
We used Devore's, Probability and Statistics for Engineering and the Sciences. It was okay.
edit: I found my two stat classes rather unfulfilling. I took stat specifically because of the plug-and-chug or "use Excel" mentality taught in my analytical chemistry classes. Specifically as it relates to ANOVA and regression analysis. Actually understanding how PLS works would have been nice, too. We spent the last two days of the second quarter on linear least-squares. It wasn't even on the test. It sounds like your enjoying your class a little more, so I wish you luck.
posted
Hmmm...the results might be skewed if the ohmmeter isn't large enough compared to the resistor.
I have no idea whether the resistors were from the same batch, nor was I able to calibrate the ohmmeter (the equipment belongs to the school and I didn't want to mess around with it too much). The ohmmeter was supposed to be one of the best ones in the lab, though.
Anyhow, I was using the small cylindrical resistors with wires coming out either end, and I was using a digital ohmmeter to measure the resistance. I was getting four significant digits, so it's at least precise, if not accurate. The resistors turned out to be well within tolerance ranges (5%).
I don't find the stats class that fascinating, but actually performing an experiment is somewhat useful. I'd prefer to learn statistics by applying it to actual problems that I care about rather than learning formulas and occasionally applying them to made-up problems. Oh, well. Thanks for the ideas, everyone. I'm guessing the internal resistance of the ohmmeter is largely responsible for the measured resistance values being lower rather than higher than the stated value.
Posts: 3546 | Registered: Jul 2002
| IP: Logged |
posted
Actually, the lab supervisor shorted the leads when he turned the ohmmeter on. The bands were gold. I was expecting to see some deviation from the value the resistors were supposed to be, and in fact that was the subject of my analysis. What I wasn't expecting was that the values would almost all be lower than the stated resistance.
Posts: 3546 | Registered: Jul 2002
| IP: Logged |
posted
I can't help but wonder if an ohmmeter originally designed to use a carbon-zinc battery would read lower resistance when using a higher voltage alkaline battery.
Printer-friendly view of this topic
![[Smile]](smile.gif)
