So i'm working on a project and I am trying to figure out a chart on error budget analysis but it's getting me somewhat confused. I notice that the chart lists +- errors for Gain, Vos, Ib, etc. It lists plus and minus errors but when it lists the total algebraic worst-case error, it only lists a positive error. Shouldn't the total error be +- like the individual errors? If my output voltage is 10 V, and my error budget is 50 mV, should my allowed output range from 9.95 V to 10.05 V or is my range just 10 V to 10.05 V?

I'm sorry if i'm just thinking about it the wrong way. I've attached the chart for reference.

 

The total error budget is +,- because you can't predict the polarity of the error.

 

Your final output range is +/- x number of millivolts. The only problem is in how you are expressing the error in numbers. You have to add up the worst possible errors to find the worst possible result and that is kind of like adding up the absolute values of the worst errors. If you added together both the positive and negative limits of the errors, you would come up with zero. Right? Kind of like a sine wave. If you add up all of it's excursions away from the zero line, the results are zero. If you believe the math without thinking properly, all sine waves add up to zero, so they don't exist in the real world!

You have to go for magnitude only when you want the answer for the worst magnitude. After you get the amount, you can usually put the +/- back on the result because it is the truth. You only took the signs off the numbers long enough to find their worst possible result.

 

Ok, I get it. Thanks for the help. I was going to use the +/- range but I just couldn't get over the fact that the table only listed a + value. Thanks again.