DMM DCV Comparison Keithley 2000 vs 2002 vs 2750

[VintageNut]

Everybody likes pictures, below is a contribution to the collection here on EEVB. The intent of this configuration of gear is to several fold. I recently bought a model 2000 DMM and want to verify its state of calibration. My 2750 is approaching the two year mark since calibration and I want to check it as well. I calibrated my 237 last year and wanted to check its sourcing cal. I had an opportunity to borrow a 2002 that is withing its cal date. So  this exercise is a DCV check only. The 237 source ranges of 1.1V, 11V, 110 V and 1100V. For 100mV I used a borrowed 2450. All three DMMs have DCV ranges from 100mV to 1000V. The 2002 can measure 1100V.

Here is the rig after completing the 100mV check.



The measurements and calculation of uncertainty of the 2002 are here. The 100mV results are not completed so I omit those.

Interestingly, the model 2000 is closer to the model 2002.

[TiN]

Nice stack of SMUs you have there I wanted to get one of those, but they go pretty high on prices.
Also would be interesting to compare ohms function, as that's one big difference between 2001/2 and 2000 series.

As a side note, you have meters warming up each other, that could affect your results a little.

P.S. any chance to crack the hood open on 2750 and for few photos? I never saw what's inside of those. Perhaps in another thread?

[VintageNut]

The 237 was as-is out-of-calibration on eBay for $350. The case was bent and cracked. The fan was destroyed and locked in place. It had been obviously dropped on its corner from some considerable height. I bought a new fan and installed it. It runs perfectly. The 1000V range force V,  measured 1000.66 using the 2750 which is what the eBay seller claimed. It was only over spec by 0.02V but definitely out of spec.

I spent a couple of weekends learning how to calibrate the SMU and did so. The 1000V force range now measures 1000.06 using the 2750 which is about 10X better than required. Almost all of the 23X SMUs uncertainties are 2x to 10X better than spec.

The 236 I got for $450. It was also as-is out-of-calibration but in good physical condition.

Both of these SMUs were pretty bad forcing and measuring current below 1uA. I spent considerable time verifying low current force and measure and calibrating them. Both SMUs can now easily  force and measure 1pA.

The 238 I paid real money for but was able to negotiate below $1000.00.

I wrote a Python program to run the 236 and 237 to extract Vds-Ids FET transistor curves. A future endeavor is to create a complete python transistor matching application.

I will do an ohms comparison and post it. I will also do a separate 2750 open-the-case thread with pictures. Good idea!

[TiN]

Might also add threads about 23x SMUs
I'd love to play with these beasts, but shipping alone to here would cost few hundreds easily. So will stick to 2400 I have

[VintageNut]

Here are the DMMs measuring 10V from the 237. The model 2000 is about 1 ppm higher than the model 2002 on +10V and 2 ppm on -10V. I am pleased that this is the case.

For +/- 5V, the same holds true. I have a 5V and 10V reference on the way to me. Looking forward to characterizing the references.

I have also purchased a Fluke 332A and a K-V divider. This will allow me to perform a confidence check on the 237's ability to measure 1000V.

I also want to try to extend the K-V divider's voltage range with a series resistor and a trimmer. I will perform a compare of the divided high voltage with the 10V reference. 

The 2750 is repeatably 20ppm higher than the 2002 with just about every DC range and reading.