Keysight's new 34465A (6.5 digit) and 34470A (7.5 digit) bench multimeters

[maxwell3e10]

It maybe interesting to repeat this test on 100 NPLC setting. There are some data from OldNeurons on TiN's website that suggest that 34461A is better than 34465A/34470A on 100 mV scale and 100 NPLC.

[dmm2018]

25.41°C would be the internal temperature and not the lab temperature

Thank you e61_phil.

If 25.41°C is the internal temperature, I guess the lab temperature (ambient temperature) should be lower than 25.41°C.

I note that the temperature readings shown on the Cal menu of the meter keep changing (see below). If I adjust my room aircond temperature setting until the "Temperature change from last calibration" = 0°C, is it ok to guess/tell that the ambient temperature of my room is now equal (or very close to) to the ambient temperature of the Keysight Cal lab when the meter was calibrated at the Keysight Cal lab?

[dmm2018]

Wished to access the built-in Web Interface of my 34470A via a Web browser. My Web browser (Google Chrome, Version 71) has no problem to access the page shown below. However, when I clicked the "Instrument Monitor and Control" tab (the 2nd tab, see below), an error message popped up: "Java not detected".

I followed the troubleshooting steps described at the following Keysight site: https://www.keysight.com/main/editorial.jspx?cc=MY&lc=eng&ckey=2573860&nid=-11143.0.00&id=2573860 but still cannot solve the problem.

My OS is Windows 7 (32-bit).
Java version 8 is installed.

Kindly advise.

[HKJ]

Is it a old firmware version, my meter looks different and I can select a HTML5 control that works. It do also have a Java applet, but it states that Java applets are not supported in Chrome.

[JDubU]

Some info on Java applets in Chrome:

https://www.java.com/en/download/faq/chrome.xml

[dmm2018]

Is it a old firmware version, my meter looks different and I can select a HTML5 control that works. It do also have a Java applet, but it states that Java applets are not supported in Chrome.

Thank you HKJ.

Later when I tested with IE-11, it works !!!

Then checked the firmware version. Yes, it's an old firmware: A.02.06. I borrowed this unit from a friend.

[HighVoltage]

It maybe interesting to repeat this test on 100 NPLC setting. There are some data from OldNeurons on TiN's website that suggest that 34461A is better than 34465A/34470A on 100 mV scale and 100 NPLC.

OK, I will make another run at 100 NPLC of all 4 instruments
May be I will add a second 34461A

[dmm2018]

Some info on Java applets in Chrome:

https://www.java.com/en/download/faq/chrome.xml

Thank you JDubU.

Tried using Firefox version 64 >> Not Successful

Tried using IE version 11 >> Working

[Dr. Frank]

25.41°C would be the internal temperature and not the lab temperature

Thank you e61_phil.

If 25.41°C is the internal temperature, I guess the lab temperature (ambient temperature) should be lower than 25.41°C.

I note that the temperature readings shown on the Cal menu of the meter keep changing (see below). If I adjust my room aircond temperature setting until the "Temperature change from last calibration" = 0°C, is it ok to guess/tell that the ambient temperature of my room is now equal (or very close to) to the ambient temperature of the Keysight Cal lab when the meter was calibrated at the Keysight Cal lab?

The 34465/470s internal temperature usually rises about 3°C above room temperature, therefore this instrument had been calibrated at about 22.5 °C.
In the cal certificate, they do not give the actual cal temperature, only a range around 23°C.

My instrument has 'date 23C' as a cal string, so I assume this to be the original room temperature (as the internal temperature is displayed on that specific screen in a different manner).

So you should set your R.T. accordingly but also use ACAL, to get best accuracy.

And your friend should update the firmware to the recent 3.0, as this old one still contains a lot of errors, and 3.0 adds the DIG option, and HTML5 operation of the browser.

Frank 

[HighVoltage]

OK, repeated the histogram noise test again, this time on 5 instruments at the same time over night.
The temperature span was about 1.5 degrees.

I am not sure what happened to the 34465A this time.

All standard settings, 100 NPLC

1. Room temperature measurement with 5k thermistor on a 34461A
2. Noise 34461A #1    = 70 nV StdDev
3. Noise 34461A #2    = 90 nV StdDev
4. Noise 34470A #1    = 90 nV StdDev
5. Noise 34470A #2    = 150 nV StdDev
6. Noise 34465A #1    = 120 nV StdDev

[maxwell3e10]

I wonder if some of the standard deviation is due to long term drift. One can clearly see a double-peaked distribution on the 34465A.
One would have to save the raw data and calculate the Allan variance to separate short term noise from long term drift. 

[Andreas]

I am not sure what happened to the 34465A this time.

Temperature drift or even worse: popcorn noise?
Do you have the data over time?

with best regards

Andreas

[TheSteve]

Thought I'd give this a try on my 34461A. It ran overnight and through the day. The twin peaks are the cooler temp of night and the warmer temp of day in the house. It would be nice to see no change with temperature but overall it seems the two temps are maintained quite well. There is a 3 degree C temp shift.

[dmm2018]

Thought I'd give this a try on my 34461A. It ran overnight and through the day. The twin peaks are the cooler temp of night and the warmer temp of day in the house. It would be nice to see no change with temperature but overall it seems the two temps are maintained quite well. There is a 3 degree C temp shift.

I also got similar plot (2 peaks) sometimes... I believe it's due to temperature drift... in my case, when I placed the meter is a slightly different orientation, I could see it started to drift then settled at a new peak...

[Messtechniker]

And another one.
Ambient temp. was 20.5 +- 0.5 during measurement

[zzattack]

I've read some references in this thread about keysight moving away from WinCE for these units. Is there any concrete info about this? What would the replacement be?

[Kjelt]

Had luck found an affordable 34465A last week, couldn't resist.
Installed the new firmware 3.0 and now have the DIG option enabled.

I an now thinking of the MEM option but can only find that it extends the internal storage to 2M events instead of 50k.
So what is the usecase for this extra storage if I insert a USB stick and log to external storage then this is not needed or what am I missing ?
Second where can you buy this, I saw it on Farnell for $250 but is that the normal way of purchase, not directly at Keysight, any special discounts going on presently that anyone know of ?

Thanks. Great meter, enjoyed playing with it this weekend.

[Dr. Frank]

Had luck found an affordable 34465A last week, couldn't resist.
Installed the new firmware 3.0 and now have the DIG option enabled.

I an now thinking of the MEM option but can only find that it extends the internal storage to 2M events instead of 50k.
So what is the usecase for this extra storage if I insert a USB stick and log to external storage then this is not needed or what am I missing ?
Second where can you buy this, I saw it on Farnell for $250 but is that the normal way of purchase, not directly at Keysight, any special discounts going on presently that anyone know of ?

Thanks. Great meter, enjoyed playing with it this weekend.

As soon as you really use (need) digitizing @50kHz, for longer intervals than 1sec, you'll also need the MEM option.

The instrument does not support fast data transfer to USB, or USB memory.. I didn't find the exact numbers of these speed limitations, right now.
This is mainly caused by the firmware, or the data crunching / transfer, possibly limited by the operating system WIN CE, because, in a similar manner, there's a limit of direct data transfer over the LAN and GPIB, to about 25 kSa/s.

You can observe that, if you are using BenchView trying to digitize at maximum rate and for longer period of time.

The internal buffer of the 34465A is used for that. As soon as the 50kSa, or 2MSa are exceeded, the whole system slows down.

They still have two bugs for digitizing, inside the command set, which prevent that the 344465A / 470A are really compatible with the 34411A (which probably was able to do direct high speed digitizing over its buses), and this also prevented my attempts to program digitizing at full speed.

1) There is no dedicated command to configure for 20µs or 50kHz w/o delay.
2) The smallest data format (50kHz delivers 16Bit only) for fast transfers is 8 Byte, that means bus data transfer rates of at least 400kB/sec, instead of 4 Byte / 200kB/sec, like in the 34411A. (*)

If you search the 344465A/470A for substantial data transfer rates over USB/LAN/GPIB, you'll only find incomplete information.

Anyhow, I have used this digitizing feature quite often in DCI digitizing mode, and was really happy about the 2MSa depth... got it for free, at that time.

Frank

(*) There's a detailed AN for the 410A/411A, "Agilent 34410A/34411A 6 1/2 Digit High-Performance Multimeters", which describes the digitizing state machine, and capabilities.
I'M missing that for the "successor instrument", maybe they simply replaced this state machine by SW, which is too slow.

[Mr. Scram]

I've read some references in this thread about keysight moving away from WinCE for these units. Is there any concrete info about this? What would the replacement be?

The four channel 1000X oscilloscope uses Linux where it's two channel brother uses CE. I've heard Linux mentioned in interviews too so that seems to be the most logical candidate.

[Kjelt]

@ Dr. Frank, thanks for the explanation. 

[idolclub]

Keysight Releases New Firmware v3.01 for 3446xA/70A series.

Firmware Update 3.01 Release Notes:

34460A/34461A/34465A/34470A Revision 3.01 Enhancements and Fixes:

1. Improvements to reduce the occurrence of a memory leak associated with LAN traffic that occurs when the DMM is connected to a LAN network but remains idle for several days.

2. Changed the 100 kΩ range calibration to internally select 100 NPLC rather than 10 NPLC to better reject noise associated with selected Fluke 5730 calibration values.

3. Removed Option AVG from the list of available options (Help – About window) as it is not available to all customers.

[HighVoltage]

3. Removed Option AVG from the list of available options (Help – About window) as it is not available to all customers.

Too bad not a major update with enhancements.
But good that the LAN problem was fixed.

Did we ever find out what this AVG option was all about?
As its seems now, it was only for selected customers.

[Dr. Frank]

Hello HighVoltage,
yes, about the AVG option, it's been answered here:
https://community.keysight.com/thread/26870

Frank

[HighVoltage]

Thank you Dr.Frank, I had not noticed this.

This was the answer:

The 34465AVG  AC average function is a special option that was done for the Navy that you shouldn't want.  The Navy wanted to get matching results that are given by some very old analog voltmeters before good RMS converters were available.  The 34465A true RMS AC voltage function gives a much more accurate RMS voltage.  In any case, the AVG option is only available to the Navy.  It was a mistake to have this show up in the Help > About window.

So, it seems, Keysight has a special firmware for the Navy. And may be other large customers.
Very interesting.
 

[babysitter]

Just to give back to the eevblog community:
A little python script leaking measurements into the clipboard.

Requirements:

• DMM which understands SCPI and is supported by pyvisa
• Python 3 intepreter with visa, pyvisa and pyperclip packages installed
• LAN to your PC
• clipvolteev.py file with your DMM IP Adress ath the DMM=rm.open_... line

You enter the query, a multiplier (if needed) and if you want CR-terminated string in the clipboard.

Your trusty DMM will start cyclic measurement and you can e.g. write reports and get the most recent measurement into your report by pressing Ctrl-V.

With CR termination (j) its  -0.000
 -0.000
 -0.001
 -0.001
 -0.001
 -0.001

Without CR termination (n) its  -0.000 -0.000 0.000 0.000 0.000 0.000 0.000

Tested with a 34461A.

BR
Hendrik