New Keithley DMM6500 and now DAQ6510

[AG7CK]

...
It was similar for me, I have a bunch of other, older meters, some more stable, some with more digits, but I wanted a modern one with all the bells & whistles.

Thank you very much for your very informative post. I missed it yesterday when I glanced through the last posts. Sorry.

I will get a DMM6500. Cool box.


[Edit: Deleted some rambling]

[kj7e]

I gave all the pics to TiN, he did a splendid job with retouching and postprocessing them as my DSLR skills are very mediocre I believe he will publish them on xDevs soon

For now I'll just say: it is LM399, but as usual pre-aged and pre-selected with custom markings

Found them;

https://doc.xdevs.com/doc/Keithley/DMM6500/lukier/

Thanks!

[hwj-d]

As i suspected.
(lukier's pictures from xdevs.com processed)





This 70k isn't a VHP100/200 best quality ...?!

Edit: Big orig pictures to smaller ones

[kj7e]

Amazing the level of effort needed to go from 6.5 to 7.5 and then 8.5 digit.

[hwj-d]

Amazing the level of effort needed to go from 6.5 to 7.5 and then 8.5 digit.

Yes, really. But I'm thinking, if I would really buy something like that again next to my Keysight

[IAmBack]

Hi.

Any news about DMM6500?
I can buy Keysight 34461A about 15% cheaper than DMM6500. Considering specifications and planned exploatation period (10 years maybe) I'm more convinced to buy Keithley rather than Keysight. Anyone would choose other way?

Regards.

[hwj-d]

Is there any newer Firmware as V1.0.0j in the meantime?

I've got mine dmm6500 today, and had not less than 4 bluescreens in the first two hours at the trial to measure my 10V references graphically. Also the y-scaling is relatively meaningless as well as the vertical indication of the cursoposition as voltage.

Btw, the calibration certificate is from 15-May-2018, says nothing else as "In Tolerance",and doesn't contain any information about actually measured values! Is that normal? Big difference to Keysight.

[maxwell3e10]

Yea, funny vertical scale and those 10 uV jumps look pretty bad if you used a good 10V reference.

Of course this meter is meant to be more of an eye candy than a metrology instrument.

[HighVoltage]

Btw, the calibration certificate is from 15-May-2018, says nothing else as "In Tolerance",and doesn't contain any information about actually measured values! Is that normal? Big difference to Keysight.

I heard that Keithley changed their policies regarding calibration certificates.
You can order a full calibration report with all detailed values, but you have to pay extra for that!

[hwj-d]

Of course this meter is meant to be more of an eye candy than a metrology instrument.

Would you rather suggest a 34465a?

[maxwell3e10]

For absolute stability one would need to go to a 7-1/2 digit meter, like a 34470A. At the 6-1/2 digit level, measurements suggest that 34461A is actually better in terms of noise than 34465A, https://www.eevblog.com/forum/metrology/dmm-adc-noise-comparison-testing-project/msg1658819/#msg1658819

Certainly Keysight 344x series never has a blue screen as far as I know. It doesn't have all of the latest user interface features of the Keithley meters, but it is still a big step up from the old style meters that just give a number.

[HighVoltage]

It took Keithley a while to fix the blue screen problems on the DMM7510, 2450 and 2460 SMUs. It is a big surprise to me to see a blue screen of the DMM6500. I will wait until this problem is fixed, before I will buy one.
 

[hwj-d]

I own a 344651a for >2 years, and that's the same reference:
https://www.eevblog.com/forum/metrology/build-a-10v-buffer-inside-the-teko371-ltz1000-box/msg1611664/#msg1611664

Confessedly a bit rough construction, that should be a prototype, a feasibility study, but it works actually very good in my opinion. Don't see those 10 uV jumps on the Keysight.The Keysight is (edit: 10uV) 60uV off to the 6500 and need calibration for shure. But now, i really don't know, what's going on with this brand new dmm6500.

Frankly said, there to add a note that says no more than "all ok my dear", and to label that with "calalibration certificate" is a joke! They would hardly sell a dmm with "NOT in Tolerance", instead they can put the advertising as a certificate immediately. 

Ok, no more bluescreens in the moment after >24h, the connectivity is better with the dmm6500, and i have some expectations in TSP and APPS. So i will give the meter another chance.
 

[Kleinstein]

The low frequency noise shown for the DMM6500 does look nasty at first view. However it is not that much different from what you are expected to see for an LM399 reference. It may be the difference between a simple of the shelf LM399 and one selected for low popcorn noise.

It is a 6 digit meter after all and 10 µV in the 10 V range is just 1 ppm. They show some extra resolution, but one can not expect this to be really stable.

If in 10 V range, a 60 µV and thus 6 ppm difference to another meter is also well within the specs.

For me the blue screens would be more something to worry about - a software problem should not get better over time.

[hwj-d]

The low frequency noise shown for the DMM6500 does look nasty at first view. However it is not that much different from what you are expected to see for an LM399 reference. It may be the difference between a simple of the shelf LM399 and one selected for low popcorn noise.

It is a 6 digit meter after all and 10 µV in the 10 V range is just 1 ppm. They show some extra resolution, but one can not expect this to be really stable.

If in 10 V range, a 60 µV and thus 6 ppm difference to another meter is also well within the specs.

For me the blue screens would be more something to worry about - a software problem should not get better over time.

Yes, i understand your judgment. I understand my somewhat subjective expectations of "such" a meter from a meta view, so thank you for that. But my answer is, with all due respect, NO, that's still unacceptable.

I don't want a >1000 Eur meter with all this metrological limitations at the lowest end. Starting, for example, with the seemingly only nickel-plated measuring sockets, over the compared to my 34461a unstable measurement results, right through to the absurdities, simply erroneous behavior of the gui, not to mention the withheld docs from the important first calibration and bluescreens. This meter is also at the metrological lower (-lowest) end of the "specs", a 34465a is, subjective diametral, at the higher end of what the lm399 technology is to get. And with this, it has to be measured. This meter is is finally a Keithley-Tek, the successor of the K2000, and as such, it really disappoints me.

A hobbyist, for whom this is the very first table dmm will say, what a beauty. An electronics enthusiast will be rather disappointed in the ranks of his older 6.5 meters, a metrologyist should not get such a dmm, also hardly in this price range. For me, that was a bad buy.

I am sorry to say so. Others may come to other conclusions.

[hwj-d]

Another example of what this dmm delivered during a long-term measurement (ltz1000 ref).

These 18uV spikes are not from my reference:



It ends up by simply touch the touch-screen to the left (only the screen!). After that, measurement again reasonably ok. 



It started here:



This dmm has BIG problems...     


Edit: taken with 5 NPLC, Filter 10 repeat, autozero, 10Mohm impendance, ap's teflon cable
My 34461a shows max 2 to 2.5uV 100 NPLC

[coromonadalix]

We had in our equipment a strange touch screen problem,   We had to put an 0.1 uf 1kv capacitor between the electronic ground and the earth ground to kill noise on the touch screen supply ?

The supply we use is CSA UL CE compliant / listed, but far from being prefect, some ac component is found on the dc outputs ?

Just touching the screen and the noise drop   wow

[hwj-d]

According to my observation, behavior has less to do with capacitive disturbances, but rather the algorithm triggered by shifting it to the left, cause a different processing of the signal to be measured. In other words, it was somewhat "out of takt" before.

[HighVoltage]

This dmm has BIG problems...     

Are you saying that without changing the anything on you setup the spikes just showed up
and went away again by swiping the screen?

[hwj-d]

This dmm has BIG problems...     

Are you saying that without changing the anything on you setup the spikes just showed up
and went away again by swiping the screen?

YES.

I did nothing but swiping the screen to the left and a gesture, 2 fingers, to adjust the y-scale a little.

[Zucca]

 

[Kleinstein]

Having the noise spikes disappear/reduced  when touching the screen is interesting. It somewhat points to a problem with common mode noise (e.g. from the DMM internal supply).  I don't think it is a problem of the touch screen itself, more like from a switching converter.

So this might really be a problem of the meter - maybe by design or a broken internal cap. The environment might also be a factor, e.g. with lots of EMI from the grid or maybe lamps.

We had in our equipment a strange touch screen problem,   We had to put an 0.1 uf 1kv capacitor between the electronic ground and the earth ground to kill noise on the touch screen supply ?

The supply we use is CSA UL CE compliant / listed, but far from being prefect, some ac component is found on the dc outputs ?

Just touching the screen and the noise drop   wow

0.1 µF sounds like a lot - normally the usual class Y caps should be sufficient (e.g. 5 nF with high voltage rating for safety). In a meter I would prefer even less.

[Synthtech]

Element 14 in Australia is of course out of stock of the Keithly for months otherwise I would have already purchased a DMM6500, perhaps I am lucky that they don’t have them. Of course they are also out of all the Keysight Truevolt meters but after reading this thread I am glad that I am forced to wait, it will be interesting to see how this all unfolds, I might be better with a Keysight.

[hwj-d]

Having the noise spikes disappear/reduced  when touching the screen is interesting. It somewhat points to a problem with common mode noise (e.g. from the DMM internal supply).  I don't think it is a problem of the touch screen itself, more like from a switching converter.

So this might really be a problem of the meter - maybe by design or a broken internal cap. The environment might also be a factor, e.g. with lots of EMI from the grid or maybe lamps.

Is this to me?
This is not "touching", this is swiping.
https://www.eevblog.com/forum/testgear/new-keithley-dmm6500/msg1675235/#msg1675235

And this can not be EMI from enviroment, if itself measured this 5h before without this (over night measurement); yes, this are wrong measurements from the meter itself.

[Kleinstein]

According to my observation, behavior has less to do with capacitive disturbances, but rather the algorithm triggered by shifting it to the left, cause a different processing of the signal to be measured. In other words, it was somewhat "out of takt" before.

I had overlooked this and my first thought was the effect of capacitive coupling to the screen. This description would point towards a software problem or maybe something like interference of the display with the meter itself, like massive interface traffic on the shift / redraw causing the spikes.