Author Topic: How accurate does it really need to be?  (Read 342 times)

0 Members and 1 Guest are viewing this topic.

Offline aswallace88

  • Newbie
  • Posts: 1
  • Country: us
  • Country: us
How accurate does it really need to be?
« on: June 27, 2019, 07:43:16 pm »
I've recently started designing and building circuits, and to this end, I've gotten some tools that will help me understand what's going on. I got a digital oscilloscope (Siglent SDS1052DL+) and I happened upon an old Tektronix 463A for a reasonable price. They both power on and both seem to measure pretty well. The only issue is that the last proof of calibration I have for the 463A is from 1984. I want to calibrate it, and Tektronix will calibrate it for me but it'll cost $190, assuming there aren't any broken or worn out parts that need replacement. Can I calibrate it myself and get close enough for my hobby work seeing as it seems to be working correctly, or would I be better off sending it off to Tektronix and shelling out the money to make sure it's done to the nth degree?
 

Offline bob91343

  • Frequent Contributor
  • **
  • Posts: 475
  • Country: us
  • Country: us
Re: How accurate does it really need to be?
« Reply #1 on: June 27, 2019, 07:51:55 pm »
Why do you need a calibrated unit?  Once you answer that you can answer your main question.

To calibrate an oscilloscope you only need an accurate voltage and an accurate time or frequency.

The oscilloscope has a calibrator so you can just verify that.

Nearly any cheap DMM is more accurate than an oscilloscope for voltage so you can compare with that.  Frequency is easy too; even the power line is close enough.  The 60 Hz line is 16-2/3 milliseconds per cycle.

Chances are the units are just fine as is and don't need calibration.  You don't measure stuff precisely with those instruments anyway; other gear is for that.
 

Offline David Hess

  • Super Contributor
  • ***
  • Posts: 9902
  • Country: us
  • Country: us
  • DavidH
Re: How accurate does it really need to be?
« Reply #2 on: June 28, 2019, 08:15:13 am »
First check the 463A against voltage, time, and transient response references to see if calibration is even required.

The chances are only the gain adjustments for vertical and horizontal deflection are required.
« Last Edit: June 28, 2019, 08:19:10 am by David Hess »
 

Offline Kleinstein

  • Super Contributor
  • ***
  • Posts: 6086
  • Country: de
  • Country: de
Re: How accurate does it really need to be?
« Reply #3 on: June 29, 2019, 06:59:46 am »
In most cases one does not need a calibrated scope. The main purpose of calibration is a check that the instrument is still working oK and some commercial uses can need thus for legal reasons. Especially for a scope such a test is relatively easy and for hobby use it is a good point to really learn how to use the scope and know it's limits. One may not get all the details a good cal lab could check (e.g. divider compensation, gain flatness).

Even if one really want's an official cal paper,  one would do a first test anyway before sending it in - this is especially true for hobby use where time is not that expensive.

A scope is usually not that accurate anyway, so that adjustments (e.g. gain, time scale) may not be needed, even after many years. SOme adjustments (e.g. divider compensation) are tricky and it may be better to just leave it as is unless one really knows how to do it and has the right instruments.
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf