New Keithley DMM6500 and now DAQ6510

[elbot]

Couple of bug's today..

Blue Screen Swiping in Diode Function  see DMM6500_Diode Swipe

After working in Diode function, loading a script and this message show up    see DMM6500_remote
DMM are not controlled remotely, hereafter this..     see Error_5738_after_remote
DVM behaves strange so power cycle to get i running again.

I will later report a bug where the DMM stops measuring but still responds to GUI, but out of time for now.

[exe]

'old' K2700

If you plan to trash them I can help reducing e-waste

[Brad O]

There is one thing which did not migrated properly: external triggering using the Digital I/O connector in K2700 emulation mode. We had to (re)connect the trigger line to the BNC connector on te back of unit to make it work.

I see what you're talking about, I couldn't get it to work either.  I know 2700 emulation was tested pretty extensively so I'm not sure if there's a problem with the manual or the emulation mode.  The person who tested it is out this week, so I'll get back to you next week with an update.  I'm glad you were able to get your triggering to work over BNC though!

Running as Digi I as I often use (actually the reason to buy the DMM) I was fiddling around a bit with buffer-size and sampling rate
to restore things to "normal" by running a script I have used many times before, a few movments later error shows up...

NO fingers on the screen and no communications physically attached to the DMM so don't understand the message..

This "Communication error" message happens when the main processor talks to something else in the DMM and doesn't get a response back withing its timeout period.  It can sometimes happen when the DMM is busy processing a complex setup command and you send it another one too quickly.  Mostly, the processor handles these holdups itself, but with your lightning fingers, I could see this error happening.  Its rarely a fatal error though and the DMM6500 almost always recovers from it, sometimes without even dropping any commands.

There's another cause that's more common though: Most commands execute sequentially, so they'll always wait for the previous one to finish, but a few, like the trigger model, are overlapped so you can continue executing other commands while they do their own thing.  If you send a command that's not expected or conflicts with the current overlapped command while this is happening, you can get the communication error because the DMM won't know how to respond.  This usually only happens with scripts, which is why the waitcomplete() command exists, front panel operation should automatically wait for the overlapped commands to finish before sending a conflicting command.

The next firmware release will have more advice and help where that "No additional information available" message is now. 

Blue Screen Swiping in Diode Function  see DMM6500_Diode Swipe

After working in Diode function, loading a script and this message show up    see DMM6500_remote
DMM are not controlled remotely, hereafter this..     see Error_5738_after_remote
DVM behaves strange so power cycle to get i running again.

Was this blue screen also caused by selecting the settings shortcut while the swipe screen is moving?  That shortcut works the same in every measurement function so I would expect that bluescreen to be there for each function.

For the other errors, the Communication error message almost certainly comes from the remote control message.  But you say you got this message while you didn't have *any* interface cables attached?  That'll require some more investigation... The DMM should not be able to put itself in remote mode as far as I know.

A few other things:

• You can enable a full log of commands that would be really helpful for us in figuring out what's going on with these bugs you're finding.  Go to MENU > Event Log and in the Log group, turn the Command setting On.  That will record most commands the box executes to the event log which you can then save to a USB drive.  That event log would help us in figuring out the string of commands leading up to an error since some of them (like this remote operation message) is hard to replicate.
• You can take a screenshot of the box's screen by pressing ENTER while holding down the HOME button.  That'll save a screenshot to an inserted USB drive, you may find it easier that taking pictures with a camera.

[elbot]

Brad O,

The complexity of most bugs found from now on increases and therefore gets
more and more difficult to trace down... Quite sure the easy ones mostly found now
so I will try to use the log function, great with the screenshot function, I properly need an email
to be able to upload files directly to you.

Regards
Claus

[Brad O]

At the company i work at we are currently migrating from the 'old' K2700 to the newer DAQ6510, for now in the K2700 emulation mode. IMHO these meters are really great to work with and really a step forward from the K2700. We use the RS232 option card and it was -almost- plug 'n play in our automated measurement setups.

There is one thing which did not migrated properly: external triggering using the Digital I/O connector in K2700 emulation mode. We had to (re)connect the trigger line to the BNC connector on te back of unit to make it work. The migration manual mentions the DIO DB-9 as a valid external trigger source, so @Keithley, are we missing something like an additional configuration? The setup worked flawless with a K2700 unit. Trigger is at 5V TTL level.

Yep!  That's a bug!  I'm very sorry you had to change your setup.  It's AR61773, it will be fixed in the next firmware release.  DI/O triggering on Pin 6 was initially not supported, so it never made it onto the testing plan.  Later, engineering found out it was possible so the manual was updated to include it, but it still didn't get onto the test plan.  So, an error made it through to the first release that means the DMM never starts listening for the trigger signal on that line. 

[MikeP]

 Hello Brad O.
 
 Not about bugs, nevertheless hope for your answer.

 In the ref manual there is no mention of the hybrid filter. Please explain differences of various filters in practical use.

 This is important question for me. Appendix C: Performance verification, C-25 page. For 1 Ohm and 10 Ohm resistors typical values 1,000157 and 9,999450 are shown. For the best result it is necessary to accept THESE values, but not 1,000000 and 10,00000?
 By the way, on the C-31 page there is a small disorder for 1A range.

 How I can get calibration results? Does it exist?

 Thanks for good work.
 
 Why this multimeter is called 6 1/2 digits? It can show +/-1.200.000, but not +/-5.999.999 or +/-2.999.999. This is 6 1/12 or 6 1/6 digits DMM.

[Brad O]

Sure MikeP, in fact, I would say that I prefer talking about things that aren't bugs!  Hopefully that doesn't bother any entomologists 

The Hybrid filter is the same as the moving average filter, but it doesn't return any measurements until the whole filter block fills up for the first time.  This delay's the first measurement, but ensures you never get an unfiltered (or partially filtered) reading.  The reason it's called hybrid is because the first reading acts like the repeat filter, but every reading after that behaves like the moving average filter described on 5-54 in the manual.  This was added for the most recent firmware release and so was included in the 1.0.01f Release Notes, rather than the Reference Manual, it'll be added to the Manual when we put out a new update to it.

The measurements on C-25 are example measurements preformed by our cal lab for the manual.  So the "Typical reference DMM reading" values of 1.000157 and 9.999450 are the actual measured values of the specific resistors we used for the manual example.  You should use your own resistor or calibration source with nominal values of 1 Ohm and 10 Ohms, the actual value should just be close to that, but I don't think it's spec'd how close.  Page C-4 lists the how to find the upper and lower limits, you simply use the accuracy listed in the specification: Actual Value +- (% of actual value + % of range)

For the full calibration report, the best thing to do is be sure to ask for it when you order your unit.  If you got your unit recently, try emailing RMA at tek.com with your model number and serial number (they might need your calibration certificate number too).  They keep the full data for some amount of time so they may still have it.  Otherwise, you will need to ask for the full calibration report the next time you send your unit in.  I think the full data costs a little bit extra for the DMM6500/10, maybe $10?

The 1/2 digit refers to the first digit in the reading, which can be either a 0 or a 1.  So the DMM can read 6 full scale digits, where each can be any number 0-9, and one 1/2 digit that's either a 0 or a 1.  A 1 would be the full scale of the range (i.e. 1.000000V), but most meters let you measure a little above the maximum of the range, often 20% in Keithley equipment.  As far as I know, calling that a half digit is industry standard.  I believe the unofficial formula is fractional digit = (max value)/(number of possible values), I say unofficial because it isn't always true.  Keithley usually uses 1/2 digit to mean a digit that doesn't have a full 0-9 scale, other manufactures do similar.  Even in the DMM6500, some measurements can read more than 1 in the first digit.  Current for example has a 3A range, but it's still considered 6 1/2 digits, using the formula I gave would mean that particular range is 6 3/4 digits.  It would be confusing to list the precise fractional digits for each range and function, so the whole meter is classed as a 6 1/2 digit.

Hopefully that answers your questions?
EDIT: Fixed when the Hybrid filter was released and where it's documented.

[MikeP]

 Brad, thanks for your answer. This is great explanation. 
 Can you show some examples of dual measure functions (secondary swipe screen)? I think this is very interesting function. But I don't know how to use that - ref manual without enought info.

[Brad O]

Like how people use the secondary measure function?  Well a big one is to measure voltage and current of a circuit at the "same time" (measurement functions are usually separated by physical relays that take time to change).  But you could also diagnose an AC line by measuring the frequency and voltage at the same time, or measure an AC Voltage signal on top of a DC Voltage signal.  You could also measure the voltage to a thermocouple and the temperature of the thermocouple at the same time.  Maybe make a DC current measurement but also measure frequency to detect ground loops or AC noise.  I think there's lots of uses out there, maybe some other people have examples of how they've used a DMM's secondary measure feature?

[MikeP]

 Brad thanks.

 I was very confused with opportunity to set two voltages at the same time. Yes. It does not work. 
 For the rest, dual meas. is really very good function.
 Screenshots - burden voltage (front connectors). Some instability in the fuse contact is present. Valid range of 0,3-0,4 Ohms.

[MikeP]

  So it turned out that I received this calibration document. I believe that this is useful and interesting for everyone. I do not have much experience using this device. But something seems to be true, and repeated very well.

 Brad, maybe this is not a very stupid recommendation - to make cursors with different colors.

 Thanks.

[analogRF]

I am seriously considering buying a DMM6500 after I sold my 34401A and have no bench meter now.
However, I noticed that the test currents in R and C measurements are quite high (much higher than 34411A and even slightly higher than 34401A)
How does it work for in-circuit measurements when troubleshooting? Sometimes I had wrong readings with my 34401A when I did in-circuit resistance measurement and I had to check with my Fluke 289 and it would give a much better result. This meter, however, I suspect will not be very useful for in-circuit troubleshooting.
Can anybody do some tests and give an idea as to how it performs?

one last questions, does it have the auto hold function the same way as 34461A? that meter holds the last 8 readings on screen and updates the last one every time you do a new measurement, that is very useful in troubleshooting

[Octane]

Hi analogRF

Serious question: how does the measurement current influence in circuit measurements? If there are multiple path for the current to go (in circuit), how does that change with measurement current (assuming linearity of all involved parts)? Differently asked, why is lower current better?

We always learned: never measure in circuit. You dont know where the current goes and can‘t trust the results.

Thanks,
Michael

[Kleinstein]

If everything is linear, it does not matter at which current one measures. However diodes / transistors are not linear and tend to not conduct if the voltage is well below 0.5 V. So at low voltage (low current) chances are a little better to get away with an in circuit resistance / capacitance measurement. One always has to be a little careful about the results and should not expect high accuracy anyway.

With manual range selection, one can just choose a high resistance range to get a low test current - high accuracy is usually not needed anyway.
However auto-ranging is very convenient if one does not know the range before. So ideally one would have a separate setting for high/low Ohms test current.  This might also apply to more normal measurements on resistors of small physical size. The choice of test current is a balance between self heating and low voltage related errors like thermal EMF. In principle this is to a large part a software question.

[analogRF]

Hi analogRF

Serious question: how does the measurement current influence in circuit measurements? If there are multiple path for the current to go (in circuit), how does that change with measurement current (assuming linearity of all involved parts)? Differently asked, why is lower current better?

We always learned: never measure in circuit. You dont know where the current goes and can‘t trust the results.

Thanks,
Michael

Yes, that's absolutely true and "generally" we never expect to get accurate results by in-circuit measurements, however, a lot of times during troubleshooting we do that for example as a sanity check or to find "open" or get an idea about a capacitor in circuit, etc...now when the test current is too high (and usually the open circuit voltage of the R/C range is also too high if you check their specs) it can strongly forward bias pn junctions in the circuit (around the area of measurement) and thus have too much influence on the result leading to wrong conclusions.  I have personally checked when I used my Fluke 289 instead of 34401A, a lot of the times I would read numbers much closer to what was expected.

I actually remember a specific case when I finally detected an open SMD resistor in circuit (reading a value much higher than its nominal value) when I used my Fluke 289 but my 34401A was giving me an answer which was not that far off and I could not suspect the resistor was open! It wasted my time for a while until I finally found the problem when I switched to my fluke 289

[analogRF]

....but still this thing has so many goodies packed in it that makes it very very tempting

34410A/11A/34465A are really the best if low test current and also accurate AC measurements are important.
Keithley ones don't come close in my opinion but now with this DMM6500 they have lot of other advantages over the Keysight ones
including but not limited to their better accuracy in DC and resistance

[HKJ]

34465A are really the best if low test current and also accurate AC measurements are important.

Is it? I would guess most handheld meters uses about the same test current and much lower voltage voltage.

[analogRF]

34465A are really the best if low test current and also accurate AC measurements are important.

Is it? I would guess most handheld meters uses about the same test current and much lower voltage voltage.

sorry I was only comparing among 6.5 digit bench meters. yes, hadhelds are better in that particular spec

[MikeP]

 For me, a large current was one of the reasons for buying. Since a large test current is an important part in the accurate measurement of small resistances.

[Kleinstein]

Small resistors (e.g. 10 Ohms) need a relative high test current - no question here. The problem here is more about larger resistor ranges like 100 K. Here there is a choice of using something like 1 µA to read in a 100 mV range, or use something like 10 µA and read up to 1 V. It is possibly use even more and go up to maybe 10 V.  Here is really depends on the application: sometimes 1 µA is better and sometimes 50 µA could give the best results.

Another point is about how residual voltage (like thermal EMF) are handled: Some meter use a DC current and some measure both with and without current (so kind of low frequency AC). This can also have an effect on how it reacts in circuit. This feature can usually be turned off and may not be used in all ranges. An important point here is that the instructions tell, what way is used, so one knows about possible pitfalls. With the large display it might even be nice to show the used  test current in the display.

[analogRF]

yes, my concern was about measuring (roughly) high resistances, say less than 200K or 100K, in-circuit without activating any semiconductor junctions around that point.
I went through the datsheets of several 6.5 digit bench meters and did some simple calculations and here is what I found.
I have come to trust my Fluke 289 at 500K and 5M ranges to be the best for in-circuit testing. So I will compare the bench meters to my Fluke 289 at 500K range. This meter gives 10ohm resolution with 0.5V full scale voltage (1uA) at 500K range. The open circuit voltage of this meter is always 5V

I am not dwelling on the accuracy specs of these meters since it is not that important for this application.

In case of all Agilent 6.5 digits, old and new, it turns out that if I set the range to 10Mohm I get 10 ohm resolution with 5V full scale (0.5uA) which is even better than the fluke 289 at 500K range. At 1Mohm range the resolution is 1ohm at 5V full scale voltage (5uA), so for 100K resistor it will be 0.5V which is too high. So I should have used my 34401A more wisely 

In case of Fluke 8846A and Tek DMM4050 (identical), in 10M range we get 10ohm resolution with 10V full scale voltage (1uA) which is the same as Fluke 289 at 500K range (0.1V for 100k resistor). The open circuit voltage in this range is 13V. The 1M range is not good with 10uA test current

In case of DMM6500, in 10M range we get 10 ohm resolution with 3.5V full scale (0.7uA parallel with 10Mohm input) which is even better than the Fluke 289 at 500K. The open circuit voltage is about 7V.
BTW this meter is excellent for low resistance measurement with 1ohm range and 10mA test current.

So no concerns here 
now I need to find one with reasonable price 

[MikeP]

Another point is about how residual voltage (like thermal EMF) are handled: Some meter use a DC current and some measure both with and without current (so kind of low frequency AC). This can also have an effect on how it reacts in circuit. This feature can usually be turned off and may not be used in all ranges. An important point here is that the instructions tell, what way is used, so one knows about possible pitfalls. With the large display it might even be nice to show the used  test current in the display.

 Yes. Quite right. Here IT is called - Offset Compensation. You can enable (or disable) this feature in a range of up to 10 kOhm.

[MikeP]

 Today I used a digitizer. It works! The frequency is 100 Hz. Unfortunately, there are some defects. Probably I made a mistake in setting up. It would be interesting to know - how this should work and what recommendations for use.
 
 Thanks.

[Brad O]

MikeP:
Probably you set the digitize Count to 1 and went into continuous trigger mode?  When you do that, triggering is handled by the display processor, so when that processor gets busy it will stop triggering and catch up with whatever else it's being told to do (like update the graph image).  That's where those gaps in data are coming from.  A couple ways around this come to mind:

• Set a higher count to capture all the data you need.  Triggering within a Count set is handled by a separate processor that won't get caught up in display stuff, you also won't see the lines connecting separate groups of data that may be a little confusing.  The graph's smart scaling in x will by default show you the latest group of readings.
• Use a trigger model.  Triggering from a trigger model is also handled by a separate processor and a very simple model can allow continuous data capture.  I'm attaching a script (change the .txt to .tsp to use it) where the last lines set up a trigger model that starts a digitize voltage measurement (that continues infinitely) and then stops the measurement when the TRIGGER key is pressed.  The display might lag behind slightly depending on your other settings, but there won't be any gaps in the data.  If you digitize at a really high rate then you may see a pop-up like "Processing reading backlog...".  That message means the display processor needs to catch up with the data buffer and it will stop other activities until it catches up, usually no more than a second or two.

analogRF:
You seem to have figured out your test current question yourself, but I don't think anyone answered you on the hold functionality.  The DMM6500 doesn't have an exact equal to the 34461's probe hold feature, but it's easy enough to reproduce through scripting or a trigger model.  Tell you what, if you get a DMM6500, I'll write something for you myself
EDIT: Also in regards to AC measurement accuracy, the DMM6500 has a brand new AC measurement architecture from the model 2000 that performs much better, the specs are fairly conservative there too.

[analogRF]

analogRF:
You seem to have figured out your test current question yourself, but I don't think anyone answered you on the hold functionality.  The DMM6500 doesn't have an exact equal to the 34461's probe hold feature, but it's easy enough to reproduce through scripting or a trigger model.  Tell you what, if you get a DMM6500, I'll write something for you myself

Thanks! Finally that's the answer I was looking for  I have even asked the Tek support about the same thing and waiting for them to reply.
Reading a bit through the manual I thought that it might be possible with all those trigger models although I cannot figure out how.
Have no idea how the scripting works...

Have you tried to write something to emulate this? I am stunned they have not implemented this simple and very useful feature
in this sophisticated device. Even 2015 had this and all Keysight meters or even Rigol have it. There you can even adjust the threshold
for triggering to read a new value when you probe again. I have used it a lot and I am very used to it and I really want my new meter to have this.