Keithley 2000 TRMS converter noise

[sync]

Hi

I repaired a dead Keithley 2000. The power supply caps leaked and eaten PCB traces. It's working now again.

I would like to know the typical TRMS converter noise. It would be nice if someone with a Keithley 2000/2015 could measure it. Set the meter to ACV, 5.5 digits and short the input. The reading is the TRMS converter noise. I get 121 counts. The manual states 100 counts typically.

[ratsept]

Hi,

I just tried my 2015 THD and I'm getting 127 counts. Hope it helps.

[sync]

Yes, it helped. The TRMS noise of my unit didn't degraded.
Thanks!

[TiN]

5.5digit , 1 NPLC, REL + Auto filter

[quarks]

I would like to know the typical TRMS converter noise. It would be nice if someone with a Keithley 2000/2015 could measure it. Set the meter to ACV, 5.5 digits and short the input. The reading is the TRMS converter noise. I get 121 counts.

I have never really checked this on my Keithley.
Therefore out of interest I just ckecked my 2015 and it is around 0.13mV and the Tek DMM 4050 is 0.04mV.
Is this an indication that the 4050 is 3x better in this respect?

The manual states 100 counts typically.

@sync, where in the Manual can I find this "100 Count" information?

5.5digit , 1 NPLC, REL + Auto filter

@TiN, I would be interested to see your values without "REL", if possible.

thanks
quarks

[sync]

Keithley 2000 user manual, Page 2-19:

AC voltage offset
The Model 2000, at 51?2 digits resolution, will typically display 100 counts of offset on AC volts with the input shorted. This offset is caused by the offset of the TRMS converter. This offset will not affect reading accuracy and should not be zeroed out using the REL feature. The following equation expresses how this offset (V_OFFSET) is added to the signal input (V_IN):

Displayed reading = sqrt(V_IN² + V_OFFSET²)

Regarding if the 4050 is 3x better. I don't know. Maybe they use a less noisy analog TRMS converter or a different TRMS conversion method.

@TiN: Many thanks for hosting the K2000 schematic. It helped a lot during repair.

[quarks]

Keithley 2000 user manual, Page 2-19:

thanks sync
it is the same in the 2015 Manual (see att.)

[quarks]

because I made my first test in 100mV range, I now repeated it in 1V range
Keithley 2015 is now 1.27mV
TEK 4050 is 0,33mV

[TiN]

Time to open KI fanclub? Btw, this is after AC CAL (internal AC calibration procedure)

KI2002:

200mVAC range	Inputs shorted	0.052xxx mVAC RMS	10 NPLC, 8.5d, no REL
200mVAC range	Inputs open	6.559xxx mVAC RMS	10 NPLC, 8.5d, no REL
2 VAC range	Inputs shorted	0.000142xx mVAC RMS	10 NPLC, 8.5d, no REL
2 VAC range	Inputs open	0.00657xxx mVAC RMS	10 NPLC, 8.5d, no REL
20 VAC range	Inputs shorted	00.0032xxx mVAC RMS	10 NPLC, 8.5d, no REL
20 VAC range	Inputs open	00.0084xxx mVAC RMS	10 NPLC, 8.5d, no REL
200 VAC range	Inputs shorted	000.010xxx mVAC RMS	10 NPLC, 8.5d, no REL
200 VAC range	Inputs open	000.0146xx mVAC RMS	10 NPLC, 8.5d, no REL
750 VAC range	Inputs shorted	0000.048xx mVAC RMS	10 NPLC, 8.5d, no REL
750 VAC range	Inputs open	0000.051xx mVAC RMS	10 NPLC, 8.5d, no REL

Sorry, 2001 off now, messing with NVRAMs in it ...

[robrenz]

What I found in this old thread implies the 8846A AC performance may be better than average

[quarks]

@TiN, thanks for the numbers.
but can you please recheck beginnen from 2V range your values are probably VAC and not mVAC

@robrenz, yes once again the Fluke8846A/TEK DMM4050 show amazing good results. 

Just to see what my 8.5 Digits meters show with this quick test, I fired them up.
Whatever this really means, the little 4050 is second best!!!

 

[casinada]

On my 2015 it shows:
100mV range: 000.0084 mVAC
1V       range: 0.00074 VAC
10V     range: 00.0073 VAC
100V   range: 000.073 VAC
750V   range: 0000.71 VAC
 

[sync]

On my 2015 it shows:
100mV range: 000.0084 mVAC
1V       range: 0.00074 VAC
10V     range: 00.0073 VAC
100V   range: 000.073 VAC
750V   range: 0000.71 VAC

Interesting. That is considerably less than the 120-130 counts we had so far. Is this a newer unit?

[quarks]

On my 2015 it shows:
100mV range: 000.0084 mVAC
1V       range: 0.00074 VAC
10V     range: 00.0073 VAC
100V   range: 000.073 VAC
750V   range: 0000.71 VAC

That is interesting and is well below the 100 number from the manual
(only the 000.0084 mVAC should probably be 000.084 mVAC).

I wonder if the test results can be improved somehow.
Have you observed if the "Offset" changed over time?

I hope some more members do this test, and Report their values.

[dr.diesel]

I hope some more members do this test, and Report their values.

K2010, eta: Monday afternoon.

[mimmus78]

Mine numbers:

100mv 000,032
1V       0,00028
10V     00,0028
100V   000,027
750V   0000,26

so I'm the winner of K2000 contest?

[quarks]

Mine numbers:

100mv 000,032
1V       0,00028
10V     00,0028
100V   000,027
750V   0000,26

so I'm the winner of K2000 contest?

That looks really good.
Can you give us more details of your meter and setup?

Here are links to some more info about this topic:

https://www.eevblog.com/forum/testgear/dmm-measuring-ac-volts/msg233469/#msg233469

http://www.ni.com/white-paper/2966/en/#toc6

[robrenz]

Numbers from my trusty 8846A   (yes, I kiss my meter)

Shorted inputs, not zeroed out, high resolution, stat mode of 10 readings.  No typo's possible with screen shots

100mV


1V


10V


100V


1000V

[dr.diesel]

robrenz's weaker brother, the DMM4020, not zeroed, slow rate:

Code: [Select]

200mV 0.000mV
2V 0.00024V
20V 0.0018V
200V 0.011V
750V 0.04V

[quarks]

Numbers from my trusty 8846A   (yes, I kiss my meter)

Shorted inputs, not zeroed out, high resolution, stat mode of 10 readings.  No typo's possible with screen shots

wow, that is 3x lower than my own measurements

[KJDS]

I just measured three 2015THD on the 1V range

measured counts were 171, 80, 71

[sync]

Robrenz, these numbers are very low. Do you know how good it measures at the low end of the scale? My 34401A shows also very low noise with shorted input. Less that 6ppm of the range. But the measurements drop off dramatically under 0.4% of the range, 0.5% on the 100mV range. IMHO these low noise readings are faked on the 34401A. Geller Labs have observed this too (http://www.gellerlabs.com/34401A%20AC%20zero.htm).

[dr.diesel]

Code: [Select]

                          HMC8012, ADC=Slow, AC Filter=Med, Not zeroed
Range Mean         Min         Max         Std Dev         Pk to Pk # Measures (1000=~3min20sec)
400mV 0.062mV 0.058mV 0.065mV 0.001mV 0.007mV 1000
4V 0.00072V 0.00070V 0.00073V 0.00001V 0.00004V 1000
40V 0.0080V 0.0078V 0.0081V 0.0000V 0.0003V 1000
400V 0.080V  0.077V  0.082V  0.001V  0.005V  1000
750V 0.06V    0.05V    0.06V    0.00V    0.01V    1000



[robrenz]

Robrenz, these numbers are very low. Do you know how good it measures at the low end of the scale? My 34401A shows also very low noise with shorted input. Less that 6ppm of the range. But the measurements drop off dramatically under 0.4% of the range, 0.5% on the 100mV range. IMHO these low noise readings are faked on the 34401A. Geller Labs have observed this too (http://www.gellerlabs.com/34401A%20AC%20zero.htm).

I addressed this in the link I posted on reply #9 of this thread.  Its here again for convienience

[sync]

I addressed this in the link I posted on reply #9 of this thread.  Its here again for convienience

Sorry, I missed that. Wow, the 8846A is impressive! I subtracted the 12.5uV noise from your measurements.
V_MEAS = sqrt(V_READING² - V_NOISE²).

  V_IN     V_MEAS
 20uV    19.1uV
 30uV    29.7uV
 40uV    40.7uV
 50uV    50.7uV
 60uV    61.3uV
 70uV    74.5uV
 80uV    82.3uV
 90uV    92.6uV
100uV   102.9uV