PMEL Forum

General => General Discussion => Topic started by: Broken_Wings on 08-20-2009 -- 16:25:56

Title: Conundrum of math. K3 related but ...
Post by: Broken_Wings on 08-20-2009 -- 16:25:56
I have a unit making a measurement in a power mode. I don't want to be specific but lets say its not in volts. Its in dBm more or less but there's a relative value.

Here's the question best I can summarise it:
I have a unit with an accuracy of +/-0.75 db. Which linear unit do I convert to? What is the percent error at any point? Why do I use this linear unit?

So far I think Watts or milliWatts to be most appropriate. Units that I know of that I believe to be linear and give appropriate % error:
Watts, Volts , Amps


V will change based on R however the power dissipated will remain the same. Also the varying R from 1 to 1000 the % error in volts remains the same seeing how it figures in to the formula.

More additional but  information: matching pad NCR , input resistance of TI doesn't appear to be spec'd, TI also doesn't appear to be converting internally from Volts as the 3458A does using SMATH RES.

Sorry to serve you a wall of text but I don't like it when I don't understand what some people are doing. I'm striving for knowledge.

edited for mass rambling.
Title: Re: Conundrum of math. K3 related but ...
Post by: pmelgeek on 08-21-2009 -- 01:57:43
Ok, I apologize in advance if I'm missing what you're getting at here, but it sounds like there's a little confusion between logarithmic and linear measurements.  For your measurement concerning dB values is logarithmic and you'll have to convert all your linear measurements, (voltage, milliwatts, etc) into dBm or dBV, then add your +/-. 75dB tolerance, then convert them back to the original units. 

Since I can't really put math formulas in here with standard text, I can't get in to this too deeply.  But if you go to www. sengpielaudio. com/calculator-volt there is a good explanation of all this along with a on-line calculator that can help you with the math.  It also gets into different impedances as well, so maybe this will help.

I hope this helps, and sorry again if I missed it.