New Keithley DMM6500 and now DAQ6510

[Kleinstein]

....
It's alive!!!!

I still need to buy more solid-state relays to populate the rest of the board but what's there seems to work flawlessly (so far), silent and fast. Since I was trying to keep the cost as low as possible (you can tell from the all DIY PCB board), the relays are not the best, the on state resistance of each channel is around 2 Ohm. But it's not a big deal as I have plenty of channels to use kelvin connections if needed. A $12 Atmega32u4 based board is handling all the logic.

(top board: Keithley 2000-Scan with wires attached for protocol sniffing)
(bottom board: DIY 20-chan solid state scan card)

The on state resistance is a kind of compromise: low resistance also means more leakage. So for just voltage measurements one would prefer even higher resistance (up to a few KOhms) and thus lower leakage types.

[KedasProbe]

It's alive!!!!

A low cost DIY (CozScan 2020) solid state relay solution for DMM6500 
I blame you that I just ordered a DMM6500 for home use    (I don't expect to get the 'free' kickstart license)
Will this be open hardware/software?

[h4X||]

I have decided on either a Keithley Tek DAQ6510 or Keysight 34465A. Really leaning towards the Tek for attempting to drive a bed of nails test rig in addition to daily hardware dev and test logging duties.
Anyone using the 6x8 DAQ module?

PM me any thoughts / links on bed of nails test rig driving (dont want to clog the thread). context: I have previously just used target device MCUs to self drive a test rig, but that has limitations similar to external MCU driver.

Keithley/Tek visibility and issue transparency in the forum on this thread is a HUGE reason for considering a Tek. That takes serious stand up character to manage. My Agilent/Keysight MSO7104 scope has been a workhorse when I depended on it mid career.

Anyone have any recommendations for quotes/pricing? I am in USA and want calibrated new (built in 2019).

[KedasProbe]

Feature request:
1) A fast way to store the stats of the measurement to USB, I know you can use a screenshot but then you need to store screenshots. You can also not just copy it in a excel file.
Two options:
  A) When you take a screenshot with home+enter also write a second file, a csv with the same name.

  B) Make it possible to link a script to a button, like maybe home+trigger runs a pre-selected script.
      and let the script do the needed actions.
      (This is maybe a feature on its own that could be useful for all kind of situations.)

2) A feature request for the current measurement mode, show the voltage drop! (burden voltage isn't a second measurement) Now I have to check to be sure it's low enough. (although if not much work I choose my own resistor value)

3) For math y=mx+b, support entering 2 points (like: 4mA=0 and 20mA=7) instead of entering m and b and let it calculate m and b. (maybe this should be a script)
Also an option to change the 'X' to an 3 char that has meaning. (sometimes you forget what you were displaying after reloading the settings)

[KedasProbe]

I made few comparisons: (only one voltage from battery)
new DMM6500 vs older DMM6500
and
new DMM6500 vs old DM3068:  diff 0.0017%

The DM3068 seems to be less noisy.  (I will measure the BW later but the DM3068 is also big)

[JxR]

Feature request:
1) A fast way to store the stats of the measurement to USB, I know you can use a screenshot but then you need to store screenshots. You can also not just copy it in a excel file.
Two options:
  A) When you take a screenshot with home+enter also write a second file, a csv with the same name.

Something like this would be nice.  If such a feature is considered: I would also like to see the ability to only copy the log data between the cursors, if the cursors were turned on.

[hwj-d]

I take screenshoots and csv's per LAN 'Virtual Frontpanel' and 'Extract Data' to Libre Office.
https://www.eevblog.com/forum/metrology/usa-cal-club-round-2/msg2314803/#msg2314803

Yes, that's a bit more work, but nothing essential. In my opinion, it would be important to be able to position the cursor exactly at all, instead of just using the imprecise touchscreen.

Edit: Csv's from within the cursor would be nice certainly. Especially since the data already stored on screen.

[KedasProbe]

Just a quick burden voltage test:
The front goes up to 3A but is only useful to go that high if the burden voltage isn't an issue.
At 1A you already have 0.27V so putting it in series with a battery supply isn't an option.

The 10A must be connected to the back then you can go to 2.8A for an 50mV drop. (17.6m\$\Omega\$)

[KedasProbe]

Beta storage service (initialstate): Anyone using this?
https://www.initialstate.com/keithley/

[hwj-d]

Beta storage service (initialstate): Anyone using this?
https://www.initialstate.com/keithley/

Don't really know what it is at all. Something to buy?

[Brad O]

Beta storage service (initialstate): Anyone using this?
https://www.initialstate.com/keithley/

This is something we're testing right now.  Come to think of it, I probably should've mentioned it here before!  Initial State specializes in cloud based data visualization and monitoring.  The idea is that you can use all the Initial State features with a DAQ6510.  So you could access your data anywhere, set up custom, complex triggers and notifications, custom dashboards etc.  I've attached the datasheet/flyer.  An Initial State account is required and costs money (14 day free trial if you want to try it out and give me/Initial State feedback!) but is otherwise free.  You can then also use Initial State for any other projects, they have a ton of in-depth guides and walkthroughs on using their service with a wide variety of applications and data sources.  The full guide for the DAQ6510 is here if you want to try it out: https://support.initialstate.com/hc/en-us/articles/360024313832-1-Keithley-Datalogger-Integration-Beta-Summary-Getting-Started

Feature request:
1) A fast way to store the stats of the measurement to USB, I know you can use a screenshot but then you need to store screenshots. You can also not just copy it in a excel file.
Two options:
  A) When you take a screenshot with home+enter also write a second file, a csv with the same name.

  B) Make it possible to link a script to a button, like maybe home+trigger runs a pre-selected script.
      and let the script do the needed actions.
      (This is maybe a feature on its own that could be useful for all kind of situations.)

2) A feature request for the current measurement mode, show the voltage drop! (burden voltage isn't a second measurement) Now I have to check to be sure it's low enough. (although if not much work I choose my own resistor value)

3) For math y=mx+b, support entering 2 points (like: 4mA=0 and 20mA=7) instead of entering m and b and let it calculate m and b. (maybe this should be a script)
Also an option to change the 'X' to an 3 char that has meaning. (sometimes you forget what you were displaying after reloading the settings)

1) Hmm, so a csv of just the buffer stats?  We have thought about using buttons as shortcuts for scripts before but ultimately decided not to since the buttons are supposed to be universal across instruments.  Scripts are already pretty easy to access with the drop-down menu too, yes?  Or not fast enough for you? 

2) The DMM6500 doesn't have any internal way to find out what the voltage drop is.  The shunt resistors are internally calibrated, but there's also the path resistance which is pretty variable and can't really be measured by internal processes.  BUT, you can use the secondary measurement feature to measure voltage drop by shorting INPUT HI to AMPS, that would measure the internal voltage across the current leads.  If you choose the 1A or 3A range, there won't be any relay clicking on <100V ranges.  OR, you could go a step further and measure the voltage drop out to your DUT by connecting INPUT HI to the high side of your DUT.  You could remove the low side path voltage drop by using the rel feature between INPUT HI and your DUT low terminal before swapping to measure voltage between INPUT HI and your DUT high side.  Does that make sense?  I can draw the process out if you want.

3) That definitely sounds like a good scripting opportunity.  y =mx+b is the standard, but I can see the opportunity for cases where a 2-point definition is more convenient.  It would certainly be possible to use a script to find the values of m + b for you.

@cozdas: Very impressive!

[KedasProbe]

1) Hmm, so a csv of just the buffer stats?  We have thought about using buttons as shortcuts for scripts before but ultimately decided not to since the buttons are supposed to be universal across instruments.  Scripts are already pretty easy to access with the drop-down menu too, yes?  Or not fast enough for you? 

2) The DMM6500 doesn't have any internal way to find out what the voltage drop is.  The shunt resistors are internally calibrated, but there's also the path resistance which is pretty variable and can't really be measured by internal processes.  BUT, you can use the secondary measurement feature to measure voltage drop by shorting INPUT HI to AMPS, that would measure the internal voltage across the current leads.  If you choose the 1A or 3A range, there won't be any relay clicking on <100V ranges.  OR, you could go a step further and measure the voltage drop out to your DUT by connecting INPUT HI to the high side of your DUT.  You could remove the low side path voltage drop by using the rel feature between INPUT HI and your DUT low terminal before swapping to measure voltage between INPUT HI and your DUT high side.  Does that make sense?  I can draw the process out if you want.

1) Yes just the stats also cursors if there are any, can we write a custom csv file to usb with scripts?
Fast, well would be easy if it was present in the same screenshot function so they match an I guess this is little work or risk for firmware.

2) Hmm I'm going to give that a try, but I don't need to see the exact drop but more like a warning of expected max drop like 2 digits accuracy. Like text in a new tab you update when going there "The burden voltage is about 0.23V at max current in the buffer and 0.52V at full range."
I assume it's possible in a script, maybe with your own measured values.

I just did a frequency measure test and it's within specification although I did expect it to be much better than spec.
My Rigol DM3068 (spec 0.007%) is measuring 100kHz correctly up to the last digit 0.1Hz
My DMM6500 shows 1.3Hz too high, not planning to use it as an accurate frequency measurement device.

[Kleinstein]

Quite some part of the burden voltage for the highest current range comes from the fuse. Those larger high performance fuses have quite some resistance. Another point are switch contacts. So an estimate could be quite a bit off, especially if the fuse is replaced with a different brand / batch.

The information one the estimated / max burden is in the manual, but with so much screen area an information on the scree can be a good idea even if with some 25% uncertainty.

[KedasProbe]

2) The DMM6500 doesn't have any internal way to find out what the voltage drop is.  The shunt resistors are internally calibrated, but there's also the path resistance which is pretty variable and can't really be measured by internal processes.  BUT, you can use the secondary measurement feature to measure voltage drop by shorting INPUT HI to AMPS, that would measure the internal voltage across the current leads.  If you choose the 1A or 3A range, there won't be any relay clicking on <100V ranges.  OR, you could go a step further and measure the voltage drop out to your DUT by connecting INPUT HI to the high side of your DUT.  You could remove the low side path voltage drop by using the rel feature between INPUT HI and your DUT low terminal before swapping to measure voltage between INPUT HI and your DUT high side.  Does that make sense?  I can draw the process out if you want.

Tried this measurement method:
Seems to be a difference of 37m\$\Omega\$ less, I'm losing a bit of wire there.

edit: adding 100mA and 10mA range, it isn't working as expected. (relay switching)

[Brad O]

1) Hmm, so a csv of just the buffer stats?  We have thought about using buttons as shortcuts for scripts before but ultimately decided not to since the buttons are supposed to be universal across instruments.  Scripts are already pretty easy to access with the drop-down menu too, yes?  Or not fast enough for you? 
1) Yes just the stats also cursors if there are any, can we write a custom csv file to usb with scripts?
Fast, well would be easy if it was present in the same screenshot function so they match an I guess this is little work or risk for firmware.

Yes you can write custom csv files pretty easily using file.open() and file.write(). The next firmware will include support for getting buffer stats of a subset of the buffer, which is how cursor stats work, if not the ability to automatically create those csv files with screenshots. 

2) The DMM6500 doesn't have any internal way to find out what the voltage drop is.  The shunt resistors are internally calibrated, but there's also the path resistance which is pretty variable and can't really be measured by internal processes.  BUT, you can use the secondary measurement feature to measure voltage drop by shorting INPUT HI to AMPS, that would measure the internal voltage across the current leads.  If you choose the 1A or 3A range, there won't be any relay clicking on <100V ranges.  OR, you could go a step further and measure the voltage drop out to your DUT by connecting INPUT HI to the high side of your DUT.  You could remove the low side path voltage drop by using the rel feature between INPUT HI and your DUT low terminal before swapping to measure voltage between INPUT HI and your DUT high side.  Does that make sense?  I can draw the process out if you want.

2) Hmm I'm going to give that a try, but I don't need to see the exact drop but more like a warning of expected max drop like 2 digits accuracy. Like text in a new tab you update when going there "The burden voltage is about 0.23V at max current in the buffer and 0.52V at full range."
I assume it's possible in a script, maybe with your own measured values.

It sounds like you might just want to set up a limit then?  You could estimate at what current you'll start worrying about burden voltage and set up a measurement limit to warn you at that level.

2) The DMM6500 doesn't have any internal way to find out what the voltage drop is.  The shunt resistors are internally calibrated, but there's also the path resistance which is pretty variable and can't really be measured by internal processes.  BUT, you can use the secondary measurement feature to measure voltage drop by shorting INPUT HI to AMPS, that would measure the internal voltage across the current leads.  If you choose the 1A or 3A range, there won't be any relay clicking on <100V ranges.  OR, you could go a step further and measure the voltage drop out to your DUT by connecting INPUT HI to the high side of your DUT.  You could remove the low side path voltage drop by using the rel feature between INPUT HI and your DUT low terminal before swapping to measure voltage between INPUT HI and your DUT high side.  Does that make sense?  I can draw the process out if you want.

Tried this measurement method:
Seems to be a difference of 37m\$\Omega\$ less, I'm losing a bit of wire there.

edit: adding 100mA and 10mA range, it isn't working as expected. (relay switching)

Ah, my bad on not explaining this, the relay switching will cause brief discontinuities in the AMPS terminal so the Rigol that's not synced to the switching would measure a higher resistance.  If you hook a scope up to the terminals you should be able to see this.  The discontinuity happens because voltage measurements happen while the current range relay is in the 1A/3A position.  So if the relay isn't already in that position it will be forced there.  I'm not totally sure all the reasons that position is required, but I believe part of it is to prevent noise from reaching the voltage measurement circuitry. 

So anyway, it's best to use the 1A/3A ranges if you want to continuously measure burden voltage with this method.

[JxR]

The discontinuity happens because voltage measurements happen while the current range relay is in the 1A/3A position.  So if the relay isn't already in that position it will be forced there.  I'm not totally sure all the reasons that position is required, but I believe part of it is to prevent noise from reaching the voltage measurement circuitry. 

So anyway, it's best to use the 1A/3A ranges if you want to continuously measure burden voltage with this method.

Thank you for this magic number.  I've been scratching my head on why couldn't get relay switching to stop happening in certain ranges.  If this information isn't in the manual, it deserves a special place with bold font.

[cozdas]

Another graphing bug report:

While scanning multiple channels (in this case 19 channels), the channels are not always displayed correctly on the x axis. As you can see in the attached images (where X-axis is in "show all" and y-axis is in "swim lanes" mode), when the x-axis is re-scaled to fit all, the tracks are aligned nicely but they will progress in different speeds and will get misaligned in time until the next re-scale.

[KedasProbe]

Let's check what 1nA looks like, 10Volt over a 10GOhm resistor.
Looks better than I thought it would.  I tried not to move while pressing the screenshot buttons

Edit
P.S. my kickStart license is finally shipping from farnell (tek UK) for work.  (the conrad.be (tek DE) email game is just started for home)

Edit2 I received my KickStart license for home already via "Tektronix France and Italy"    (5 Days after my email)
Different approach though: via farnell I got a package delivery, for home I got an email from tek with everything I need.
Both at the same day only the one from work I emailed about it almost 2 months ago.
(farnell did send a big box with an A4 paper in it with the code on it that I needed  )

[hwj-d]

Did you use any filter on dmm6500? What's the input impendance of both? Try to use "repeat, count 10", input z 10M , 5 NPLC

[KedasProbe]

Did you use any filter on dmm6500? What's the input impendance of both? Try to use "repeat, count 10", input z 10M , 5 NPLC

No filter, just 1PLC it's DC current.
The DMM6500 spec says 10k so that is 1/1000 000 of the big resistor or it's measuring 10µV. STD: 0.455µV

edit: ah not for me, that makes more sense

[hwj-d]

Did you use any filter on dmm6500? What's the input impendance of both? Try to use "repeat, count 10", input z 10M , 5 NPLC

No filter, just 1PLC it's DC current.
The DMM6500 spec says 10k so that is 1/1000 000 of the big resistor or it's measuring 10µV. STD: 0.455µV

Oh, excuse me, KedasProbe, that's to shodan. My posting had overlapped with yours.

[hwj-d]

The representation of the characteristic curves reminds me rather of different interpolation procedures to determine and represent individual measured values.

[Kleinstein]

There was a detailed measurement on the noise around reply  #95 in this thread. The DMM6500 seems to be relatively noisy.
The low temperature effect looks impressive though. Chances are they use a numerical compensation, a little like in the DMM7510.

The Keithley meters tend to use a slightly different way of AZ mode, that includes some averaging. This gives slightly improved performance at low PLC, however as a downside the 100 PLC and similar data may be not that good. Form the old data is looks like the 6500 is also effected by this, though it looks better than the 7510 in this respect. Anyway the DMM6500 does not look like it is really low noise even at 1 PLC.

It is kind of sad, the Keithley does not offer a different, more conventional AZ filtering for slow measurements. This could be quite an improvement for slow measurements. With some luck it may be possible to get this via a script, if there is access to the raw data from a 1 or 5 PLC Az mode.

[KedasProbe]

I tried:
https://github.com/tektronix/keithley/blob/master/TSP_Apps/Resistance_Tolerance_Meter.tspa

- When starting I get an error, like not correctly initialized.
- When I exit the app the custom buffer stays the selected buffer, maybe this should be set back to the default buffer to avoid this next error.
- I got a few bad reading (high impedance) due to bad contact, so had to press retry but if two follow each other I will not be able to do that.
- It also shows rear terminals in the csv file while I used the front.

Not related to this app, can I make saving the measurements so that the column time is always the first column and the measured value always the second?  (makes it easier in excel)

[Brad O]

I tried:
https://github.com/tektronix/keithley/blob/master/TSP_Apps/Resistance_Tolerance_Meter.tspa

- When starting I get an error, like not correctly initialized.
- When I exit the app the custom buffer stays the selected buffer, maybe this should be set back to the default buffer to avoid this next error.
- I got a few bad reading (high impedance) due to bad contact, so had to press retry but if two follow each other I will not be able to do that.
- It also shows rear terminals in the csv file while I used the front.

Not related to this app, can I make saving the measurements so that the column time is always the first column and the measured value always the second?  (makes it easier in excel)

Alas!  I blame myself for not testing on the correct firmware.  Updated version here: https://github.com/tektronix/keithley/blob/master/TSP_Apps/Resistance_Tolerance_Meter.tspa

- I fixed the indexing error that caused that first error, the MENU will now render correctly as a result.
- Currently, there's no way to execute specific code when you exit an app.  The development firmware allows for resetting the instrument when an app exits, which would fix this.  Otherwise, since it's a writable buffer, it doesn't really need to be active, but having it active simplifies small things like going to the reading table (so that the results buffer is already up). 
- Hmm, should unlimited re-tests be allowed?  Was the bad contact just due to it taking a reading while you were still setting up or something else?
- Fixed the mislabeled terminals.  Front terminals are actually required for this app... I'm not sure why but I didn't remove that requirement in case there's a reason I didn't see (I didn't write this one). 

On saving buffers in a specific way: kinda. There's quite a few different ways to skin a cat export a buffer.  The options are fed as parameters to the buffer.save() function and are documented on page 722 (15-21) of the Reference Manual.  The option you probably want is buffer.COL_BRIEF, which saves only the reading and the timestamp (though the timestamp comes second).  For the ultimate customization though, writing your own csv is the way to go, that's typically what I do.

Let's say you want to save the relative timestamps and readings from defbuffer1 to a file named "savedbuffer.csv".  Here's a code block to do that:

Code: [Select]

fileVar = file.open("/usb1/savedbuffer.csv", file.MODE_WRITE)
file.write(fileVar, "Relative Time, Reading\n)
for i = 1, defbuffer1.n do
     file.write(fileVar, defbuffer1.relativetimestamps[i]..","..defbuffer1.readings[i].."\n")
end
file.close(fileVar)

In the Resistance Tolerance Meter app, the code to save data is at line 83, in the function save_data_btn_press_event(), should you want to customize it yourself.