Advice for a Keithley 2001M

[jfet]

I would love to purchase a HP3458A, but its beyond my means.  I have been considering a Keithley 2001M, its suppose to be a 7.5 digit multi-meter.

I am looking for any advice, good or bad on the instrument.  My needs is to test voltage references and the likes, and to log data.

I own several HP3456As and I am happy with them.  I also own a HP3457A and its not a good voltmeter compared to the 3456A, its noisy.  But its compact compared to the 3456As.

Thanks.

[Kleinstein]

The Keithley 2001 is sold as a 7.5 digit meter, but it is still relatively noisy. It also still has the same LM399 type reference as the 3456 (most of them) or 3457.  The good thing that it has a 20 V range with high impedance which sometimes can be handy. For logging a 7 V or 10 V reference this is however more of a problem.

[DavidKo]

Or you can look for Keithley 2010 (same concept with selected LM399, but without fan). I do not know how it is with multiplexer card in other FW, but mine is able to manage only 10 switches (compared to 20 in FW19 and higher on Keithley 2000).

[alm]

Depending on which voltage standards, they are likely more stable than the meter. So what you're going to use the meter for is measuring differences between the voltage standards (if they are close in value) or measuring ratios (for example to compare a 7V standard to a 10V standard.

For the former you need low noise, and for the latter you also need good linearity and short term stability (10 minutes). You can find posts on the noise testing and linearity on this forum. The Keithley 2001M should perform the same as the regular 2001. It wouldn't surprise me if the Keithley 2000, HPAK 34401A and HP 3456A all outperform the Keithley 2001 in noise. Not sure about linearity.

[MuShan]

I don't really recommend you to buy 2001. I have owned it before, but it's really bad. I think it's 6 and a half digits

[J-R]

I don't see how you could say the K2001 is "bad", it's still a reasonably decent bench DMM for the price IMHO.

The K2001 is a 7.5 digit DMM with an 8.5 digit mode, which could be of some use combined with additional sampling/filtering.  The dot matrix VFD and UI makes it quite a bit nicer to use than the K2010 in my opinion.  The fan noise in the K2001 can be annoying, but you can swap it out.

I had a couple K2001 units for a bit, had to send them back because they were sold as fully operational but had errors.  I also have had a K2010 for a couple years now.  When I have money again, I could see trying to snag a K2001 in fully working condition.

If you're just using these with a computer in a metrology context, then the display and UI probably don't matter in practice but otherwise can be somewhat old and clunky feeling for daily bench use.

So maybe consider a used 34470A or DMM7510 instead?


Some other notes:
The number of counts on a DMM can make a difference for certain measurements, since you might have to lose a digit when going up a range.  Maybe that will matter.

I have a few good 10V references, so I can use them to verify the K2010 regularly.  If I had the K2001, it might be slightly more tricky.  I'd probably just ship it out for calibration.

[maxwell3e10]

Take a look at this post
https://www.eevblog.com/forum/testgear/restoration-glory-of-keithley-2001-dmm/msg4290661/#msg4290661

Keithley 2001 is not really much better than 3457A in terms of noise or temperature stability. Only possible advantage is 20V high impedance range and some more specialized functions.

[J-R]

Just realized we are mixing the K2001 and K2001M?  Both units I had were the 2001M versions, which appear to have some improvements over the 2001, as mentioned here: https://xdevs.com/review/kei2001m/

[Kleinstein]

The difference from the 2001 to 2001M should not be that large. According to the XDEVs article the main point is with the AC part that also support average responding AC mode.

From the reports here in the forum it is a relatively noisy meter. Form the ADC design it is a bit similar to the HP3456 and may have comparable performance. The 20 V range can be nice if one needs it, but for directly meassuring a reference a 10(12) V range would likely be more suitable.
A slightly odd point in the K2001 is that the ADC itself uses a separate (low noise, but not very stable) reference and there are extra 7 V reference measurements from time to time. I don't know the details - it could be a old style cycle like in the older K192/3 with a reference measurment in every cycle, but clould also be more complex. The extra reference reading adds noise (at least from spending some time there), but can in real time correct ADC gain drift. The other point is that the 20 V range has a divider (amplifier with gain -0.5) in the signal path that is not corrected with the reference reading. The HW side seems to support a 10 V range (likely used for the Ohms part), but AFAIK it is not normally available for DCV  .

For testing a reference the ideal case is measuring the difference to a 2nd, stable reference.  A direct reading would have the DMMs reference as a limiting factor.

The way the K2001 input is made I would expect extra noise, as there is only a somewhat limited AZ mode for the 20 V range. It may look different from the noise hump seen in many other Keithley meters.

[guenthert]

Well, I have some sympathy for the desire to buy a better DMM than one already owns, but given the cost it's perhaps more prudent to re-evalute one's needs.  The 3457A isn't actually all that bad in the 3V range (with low bias current suitable for standard cells still common when it was developed).  Direct measurement of 10V references is convenient and popular around here, but requires one to put quite some trust in the linearity of the DMM.  With multiple references one can also measure the differences or chose the approach NIST tried in the eighties [1].


[1] https://www.nist.gov/system/files/documents/calibrations/im-34-2a.pdf

[jfet]

Thanks for the comments. 

Sounds like the best thing I can do, is spend the money on a proven voltage reference and do differential measurements between the proven voltage reference and the references I want to check. 

Considering a Fluke 732.

[J-R]

With regard to the 2001M improvements, this caught my eye: "Resistor divider R607 (9.9 MOhm / 100 KOhm) is now built using pair of alumina-based precision stable CADDOCK resistors, TF050R TF-255 and TF020R, glued together to equalize temperature of divider. This helps to improve ratio stability and tracking of such divider, and reduce errors from temperature variations."

A Fluke 732A/B/C seems awkward.  You have limited output voltages so I'd think you'd also want a precise voltage divider and some additional references.  At this point are you not at used 3458A money?  Don't forget yearly calibrations too.

Also, with only two posts it is somewhat difficult to determine exactly what you are trying to accomplish...

[Kleinstein]

When looking at voltage references one would normally not use the 10 M input divider. The K2001 even has a 20 V high Z range. With the 20 V range there is even additional uncertainty from the gain -0.5 stage. From the circuit I would expect rather good stability for the 2 V range and not that great at 20 V.

The Fluke 732 would help with 10 V or maybe 1 V references, but not so much with a 7 V or 2.5 V reference.

Meters with a better reference could be the Keithley DMM7510, KS 34470 (not really cheap) or an used ADTV R6581(T) that may sometimes come up for a not so high price (but often weak display).

[jfet]

I have found a couple of DC multimeters that may fit the bill, but the manual for the Advantest R6581T is in Japanese, and I can not find a manual for the ADCMT 7480T.  But there was info in the forum that the ADCMT was made under contract for Advantest.

Kinda tough to weed out which unit is better.

Comments or experiences with these units ?  Thanks.

[Kleinstein]

The Advantest R6581T is a limited (no AC, limited max voltage and it seems to be no good factory calibration - linearity cal) version of the full 8 digit meter without the T.  There is a quite lenghty teardown and repair / improvement thread in the forum. This includes quite some extra information and AFAIK also some translation / rewrite of a manual.  The display tends to get dim and may be an issue. With the LTZ1000 reference it is quite stable, but also with some quirks, causing quite some current spikes from AZ switching. The high ohm ranges also seem to have an issue with ACAL, which may be for this reason.

[branadic]

If you want to go Advantest then better get the R6581(D). But be aware that, as Kleinstein already mentioned, the INL is initially poor due to the fact that Advantest went with standard (batch?) INL correction coefficients. However, thanks to MickleT., who investigated the firmware and spotted how to change the coefficients, it is possible to adjust them and to get highly improved INL out of the unit similar to what is shown in several papers.

https://www.eevblog.com/forum/metrology/advantest-r6581-service-manual/

I rather like the meter, it is quiet and after some maintenance a valuable unit with its quirks, but in my opinion every 8.5-digit meter has one or the other drawback.

-branadic-

[IanJ]

FYI.......here's my K2001 I repaired some time ago and was undergoing testing, 156mins from power up.
Noise looks to average 2.3uV pk to pk......but can't remember what cables etc I was using.

Ian.