Theoretical vs measured electrical resistance: Percent error relation

[Eucliwood]

Mentor Note -- Thread moved from the technical forums, so no Homework Template is shown.

So, we've conducted an experiment on resistors in a circuit. The theoretical calculation was based of on E-24 series color band table and the measured resistance of the resistor was measured by a meter. We've identified each percent error, and I happen to noticed that as the theoretical resistance value increases so does the percent error. Why is that? Along with the theoretical vs measured in series the percent error is higher than of the parallel. I was just trying to get a hint so that I can relate other laws regarding it and make my interpretation on the observation. Thank you in advance! I only have three here, i have 3 other with my friend and i also did my own experimentation.

 

[sophiecentaur]

I happen to noticed that as the theoretical resistance value increases so does the percent error.

It would help if you included a table of results. How large a sample of results do you base your statement on? Also, what actual measurement circuit were you using (and details of the meter you were using etc.)?

 

[Nugatory]

You used the same meter for all the measurements, I hope Do you get similar results if you redo all the measurements using a different meter?

 

[Eucliwood]

You used the same meter for all the measurements, I hope Do you get similar results if you redo all the measurements using a different meter?

I did and it was all on a breadboard(large). The resistors were enclosed on a glass casing with its terminal popping out. I redid the experiment at home and it seems like the culprit with mine is the length of the terminals, the percent error I've got is inconsistent, although it still seems suspicious to me since the other group from our class have the same proportionality although they have different resistor values they did get an ascending percentage of error as the theoretical resistance increases.

 

[Eucliwood]

It would help if you included a table of results. How large a sample of results do you base your statement on? Also, what actual measurement circuit were you using (and details of the meter you were using etc.)?

I edited it now with 3 of the resistors we've used. The other group from our class did have the same result that the percent error increases along with the theoretical resistance. I do not know the details regarding the meter used. I redid it at home with different percent errors (different type of resistors) (2 meters) and the results were inconsistent. I think the culprit was the probing.

 

[256bits]

Leads me to suggest that the resistors at the school came from the same manufacturing batch, and yours at home came from another. Or the resistors supplied to the school had a selection bias for laboratory material.

And 3 resistors do not a trend make.
You would have to test a whole lot more.

 

[Eucliwood]

I agree. I'd leave it as a hypothesis for now. Thanks for the reply anyways.

 

[sophiecentaur]

Or the resistors supplied to the school had a selection bias for laboratory material.

The cheapest resistors (used to) come out of the production process in a wide range. They were measured (samples) and grouped into values (with none thrown away). If you bought 10% tolerance, a brown black red could be 950Ω to 1050Ω. But can buy 1% tolerance resistors. The metal film resistors can be made much closer to their quoted value. Did you look at the tolerance bands?

 

[lewando]

If I may offer a correction:

If you bought 10% tolerance, a brown black red could be 950Ω to 1050Ω.

The phrase "10% tolerance" is an informality that can lead to mayhem. The range of values for a "10% tolerance" resistor would be between +10% and -10% of the nominal value.

In the case of the E24 series (informally, the "5% tolerance" series) the resistor range is +/-5% of nominal, which the OP correctly tabulated in post #1.

 

[256bits]

If you bought 10% tolerance, a brown black red could be 950Ω to 1050

Theoretically, the mean should be 1000 ohms, with some value of standard deviation for the population.
Taking a sample from the population of resistors, one should be hopeful that the sample represents the population.
If one decides to re-check the values, which a supplier may want to do for certain customers/distributers , so as to ensure his reputation can be withheld or perhaps the customer chooses to do so himself, and the testing equipment is biased in one direction or another, perfectly "good" items may be rejected. The mean value of the new sample will be shifted. An unintended bias or skew within the items sold to the final user crops up, due to double checked quality control.

Maybe that is all a bogus surmise and non of the handlers do that, and you can call me out to lunch, since,
As @lewando states an E24 should be an E24 all the way through the selling and purchasing line.