# TDS2004B Oscilloscope Accuracy Determination

1. Oct 15, 2016

### Jamison Lahman

Hello,
I recently did the Franck-Hertz experiment (a short summary of the experiment can be found here http://inpp.ohiou.edu/~roche/3701_web/man-fh-a3b.pdf).
I am currently doing data analysis and am having a hard time determining the accuracy of the oscilloscope, the Tektronix TDS2004B. I am fairly certain it was a DC current and according to page 131 of the manual (found here: http://inpp.ohiou.edu/~roche/3701_web/tektronix-tds.pdf) the accuracy of a DC reading is ±(3% × reading + 0.1 div + 1 mV).
My voltages range from 12V-65V so the precision of the smaller measurements is about 5x that of higher measurements? Additionally, I'm plotting current as a function of voltage and I can't find any mention of accuracy of current measurements. Am I supposed to assume current measurements are also within 3% according to Ohm's Law? I don't know exactly how oscilloscopes work, but I was under the impression they have many capacitors and transistors to get very high impedances so I am skeptical the assumption is applicable. Any guidance is appreciated. Thanks. (If any over information is important, let me know.)

2. Oct 15, 2016

### Staff: Mentor

Welcome to the PF.

Sorry, I clicked into your links briefly, but my attention span is short at the moment, unfortunately. (I'm at work)

Could you please show us the math that you are asking about? Please show your calcs using the datasheet accuracy and precision specs. What are you using for the current measurements? A current probe on the 'scope? What are the current probe's specs?

Sorry again if I'm skimming too much.

3. Oct 16, 2016

### Jamison Lahman

Thank you!
Basically, I have recorded data, where the stars are the position of local maximum used in determining the period of atom excitation. Plotting the local maximum as a function of integers yields this graph, to which the residual plot for linear regression looks like . As you can see on the residual plot, the precision of the later data points is much less pronounced which makes me somewhat uneasy. Basically, I am trying to figure out what the error bars associated with the residual plot should be. Right now, they're just a constant 3% of the voltage reading.
As for the current measurements, a cord, I am amusing a probe, was connected directly from the 3B Scientific™ Power Supply Unit for Franck-Hertz to the oscilloscope. I think the probe specifications can be found on page 137 (page 115 of the actual manual) http://inpp.ohiou.edu/~roche/3701_web/tektronix-tds.pdf. With this being said, I am not sure what this information means or how to use it to find accuracy of current readings. Additionally, the specs for the PSU can be found here: https://www.fishersci.com/shop/prod...franck-hertz-experiment-115v-3/p-4085788#tab2 and "Output A for collector current I: I = UA ™ 7nA/V (0 to 12V)." maybe be of some importance though I am unsure.

The raw CSV file from the oscilloscope had the following included:
Record Length 2500
Sample Interval 0.00004
Trigger Point 1250

Source: CH1 CH2
Vertical Units: V A
Vertical Scale: 10 20
Vertical Offset: -44.4 -48.8
Horizontal Units: s s
Horizontal Scale: 0.01 0.01
Pt Fmt: Y Y
Yzero: 0 0
Probe Atten: 10 10
Model Number TDS2004B TDS2004B
Serial Number C031499 C031499
Firmware Version FV:v22.16 FV:v22.16

Also, not sure if its important, but I am using Libre Calc 32-bit