Author Topic: Keithley 2000 - Adjustment with 20kHz rather than 50kHz? at 100mV  (Read 983 times)

0 Members and 1 Guest are viewing this topic.

Offline mendip_discoveryTopic starter

  • Frequent Contributor
  • **
  • Posts: 824
  • Country: gb
I have had the odd Keithley 2000 come in and I have felt it could do with running through the calibration procedure to adjust it. Well, I can do the 50kHz at the higher voltages but my Transmille won't do it at 100mV, the max being 20kHz.

How much of an issue could you see this causing? I know I should just get the boss to spend lots of money on a Fluke or better Transmille but let us just say he is tight.
Motorcyclist, Nerd, and I work in a Calibration Lab :-)
--
So everyone is clear, Calibration = Taking Measurement against a known source, Verification = Checking Calibration against Specification, Adjustment = Adjusting the unit to be within specifications.
 

Offline bdunham7

  • Super Contributor
  • ***
  • Posts: 7729
  • Country: us
Re: Keithley 2000 - Adjustment with 20kHz rather than 50kHz? at 100mV
« Reply #1 on: July 15, 2021, 05:53:37 pm »
Since you appear to be using a 50-ppm basic DC accuracy calibrator to calibrate a 30-ppm basic DC accuracy meter, I'll assume you are characterizing your calibrator with a more accurate DMM either in parallel or in close succession.  You can just use whatever you have to divide down a higher voltage at 50kHz and then characterize and adjust that in the same way.  Setting the HF flatness correction with 20kHz instead of 50kHz will likely give a total crap HF response--you'd be better off using an AWG at 50kHz with an accurate parallel meter. 

Now what you list on the calibration certificate is another story. 
A 3.5 digit 4.5 digit 5 digit 5.5 digit 6.5 digit 7.5 digit DMM is good enough for most people.
 

Offline mendip_discoveryTopic starter

  • Frequent Contributor
  • **
  • Posts: 824
  • Country: gb
Re: Keithley 2000 - Adjustment with 20kHz rather than 50kHz? at 100mV
« Reply #2 on: July 15, 2021, 07:45:30 pm »
UKAS (17025) accredited with an Unc of ~60ppm for DC, but the annual drift data I have I can say DC I am within 10pmm drift over the past 5 years. I also have an EN ratio of less than 0.5 against an interlab with Keysight UK. So please don't scoff at the lowly Transmille. My Unc is just high as I am gradually building up my knowledge of the budget and still using the manufacturer's specification as its erring on the side of caution until I know that none of the other engineers will break something and I have to go down the route of chasing customers due to a failure of my equipment.

I haven't made any adjustments just curious as to if I needed to get something that is better than nothing.
Motorcyclist, Nerd, and I work in a Calibration Lab :-)
--
So everyone is clear, Calibration = Taking Measurement against a known source, Verification = Checking Calibration against Specification, Adjustment = Adjusting the unit to be within specifications.
 

Offline bdunham7

  • Super Contributor
  • ***
  • Posts: 7729
  • Country: us
Re: Keithley 2000 - Adjustment with 20kHz rather than 50kHz? at 100mV
« Reply #3 on: July 16, 2021, 12:30:07 am »
So please don't scoff at the lowly Transmille.

I haven't made any adjustments just curious as to if I needed to get something that is better than nothing.

Not scoffing, just trying to point out a way to accomplish what you are trying to do.  I do have opinions on the matter, but since I'm not a recognized expert on either metrology or calibration lab procedure, I'll keep them to myself.  Oh wait, this is EEVBlog.  Well, I'll tell ya......

You may know all this, but for anyone following along, the HF correction constants are done differently than the gain constants.  I haven't worked on a K2000, but I have worked on a number of same-class meters and they all do the HF correction in a similar way.  The normal AC measurement simply has the signal pass through the ranging, voltage dividing and gain stages, then to the TRMS converter and then the ADC.  After the ADC spits out a number, the meter's MCU multiplies it by the calibration gain constant for that range.  The HF correction constant is a frequency response adjustment in one of the gain stages--usually voltage controlled--and as each range is selected, the MCU commands a DAC to output a voltage to the circuit to obtain the desired response.  The idea is to get things as close as possible at one point (50kHz) and hope for the best as the frequency goes up--and you can see by the specs that things rapidly get worse out to the 300kHz advertised response.  Typically the meters are fairly accurate to 20kHz already, so setting the HF gain constant there would not be helpful.  The expectation is that the meter, once calibrated with good HF correction constants, will extend that accuracy out to 50kHz plus, after which it gets worse and worse. 

An AWG or similar in parallel with a more accurate DMM would be how I would do it, but I don't have to certify anything.  I don't know what procedures would be acceptable under ISO 17025.  I have a calibrator, much older but similar in capability to yours--but its HF AC accuracy is not much better than a Siglent AWG in practice.  The AWG allows me to verify the performance out to 300kHz as well.
« Last Edit: July 16, 2021, 12:37:22 am by bdunham7 »
A 3.5 digit 4.5 digit 5 digit 5.5 digit 6.5 digit 7.5 digit DMM is good enough for most people.
 
The following users thanked this post: jeffjmr

Offline mendip_discoveryTopic starter

  • Frequent Contributor
  • **
  • Posts: 824
  • Country: gb
Re: Keithley 2000 - Adjustment with 20kHz rather than 50kHz? at 100mV
« Reply #4 on: July 17, 2021, 06:20:41 pm »
Sorry, bdunham7 for my defensiveness. I see a fair bit of scoffing about the lowly UK Transmille so I get riled up when people start to suggest they are no good.

But thank you for the comments about Frequency. I was just hoping that it made little but not much difference. Not that I know many people who do HF voltage measurements in general, often I see 6.5digit meters just sat there in continuity mode for buzzing out wiring looms etc.
Motorcyclist, Nerd, and I work in a Calibration Lab :-)
--
So everyone is clear, Calibration = Taking Measurement against a known source, Verification = Checking Calibration against Specification, Adjustment = Adjusting the unit to be within specifications.
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf