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

[dadler]

But changing from DC to AC, at least on the 34465a, does not cause any relay click.

I have been told (in the last hour or so) otherwise - I'm not sure if the relay will click when the meter is in its stopped mode (i.e. not taking measurements), I think it probably would, but it might be a reason why you are not hearing it.

Ah you are indeed correct. I think because it's a singular click almost in concert with the button press that I psychologically filtered it out. I guess that's why all the clicky-clacky during AC+DC was jolting to me.

[HighVoltage]

Today I tested how long the 34470A need after turn ON, before it reads stable
The instrument was hooked up to a stable and warmed up 10V reference and then turned on.

Lab temperature was about 21 to 22 °C

Here are my screen shots:

34470A_after-04min.png
Screen shot after 4 min of turn on

34470A_after-37min.png
Screen shot after 37 min of turn on

34461A_Warm.png
Proof, that the reference was pretty stable during the 37 min, shown here with the 34461A in parallel to the 34470A
Only the 34461A was on for several hours already.

It seems that the 34470A needs about 30 min to stabilize the readings and shows reliable numbers.
 

[LA7SJA]

Have you done this test on your 34461A? And if so, do you care to share the result?

Johan-Fredrik

[DJ]

Today I tested how long the 34470A need after turn ON, before it reads stable
The instrument was hooked up to a stable and warmed up 10V reference and then turned on.

Lab temperature was about 21 to 22 °C

Here are my screen shots:

34470A_after-04min.png
Screen shot after 4 min of turn on

34470A_after-37min.png
Screen shot after 37 min of turn on

34461A_Warm.png
Proof, that the reference was pretty stable during the 37 min, shown here with the 34461A in parallel to the 34470A
Only the 34461A was on for several hours already.

It seems that the 34470A needs about 30 min to stabilize the readings and shows reliable numbers.

That is an extremely useful post! It has talked me out of that extra digit, which was pretty much a "that would be neat!" thing.

6 1/2 should be plenty to see drift, things leaking, discharging etc.

I owe you a beer!

[HighVoltage]

Have you done this test on your 34461A? And if so, do you care to share the result?

Johan-Fredrik

Sure, here you go....

The 34461A was turned off over night.
The 10V reference has been on for a few days now.

It seems the 34461A is a little faster in getting stable and reliable readings

[HighVoltage]

That is an extremely useful post! It has talked me out of that extra digit, which was pretty much a "that would be neat!" thing.

6 1/2 should be plenty to see drift, things leaking, discharging etc.

I owe you a beer!

Especially since the 34461A now shows one more digit after the firmware update, we can see a drift much easier than before. For me that is probably the best feature of all the firmware enhancements.

[rosbuitre]

It seems the 34461A is a little faster in getting stable and reliable readings

Thanks for the data, very useful 
Regards

[LA7SJA]

That is an extremely useful post! It has talked me out of that extra digit, which was pretty much a "that would be neat!" thing.

6 1/2 should be plenty to see drift, things leaking, discharging etc.

I owe you a beer!

Especially since the 34461A now shows one more digit after the firmware update, we can see a drift much easier than before. For me that is probably the best feature of all the firmware enhancements.

I owe you two beer at least, you have convinced my that my 34461A will serve me good for a few year forward and I would rather spend my money on more frequent calibrations of all my gear. A full calibration of my Rohde & Schwarz CMS 54 is > 750$ so I keep my 34461A until it breaks (or I) and spend my money on calibration and more stuff I don't need.

Johan-Fredrik

[HighVoltage]

Johan-Fredrik, I think you will be very happy with the 33461A after the firmware update. Both of my 33461A's are working exceptionally well.
I take you guys up on the beer offer, next time I am traveling to your country.


Here is another good comparison between the 33461A and the 33470A

Conditions:
- Both instruments ON for several hours
- 10V reference on for a few days
- Both instruments on: 100 PLC, Auto-Ranging, Auto-Scale-Once (pressed once), parallel connected to 10V reference,
- Run time about 2h:18 min as shown in "elapsed time" and 2077 samples taken

 

[dadler]

Which voltage reference are you using?

[HighVoltage]

This 10V voltage reference was the Geller Labs SVR Series (Not the SVR-T)
I will hook up the SVR-T to the 33470A in the next few days to see, how it performs, will be an interesting comparison.

[Vgkid]

It will be interesting to see the actual drift rates after a year. On paper the 34470a is rather close to my 3456a.

[Andrew]

Upgraded my 34461 to firmware v2.08 and tried out the new capacitance feature.

I noticed the "numeral" display moving about quite a bit depending on the range and value, that is, not a particularly stable reading. Made a couple of very short banana-to-croc leads (100mm tip to tip) and connected a 1nF 1% foil cap.

The attached images are what I found using manual range switching on the "trend" display starting on the 1nF range then switching through 10nF ("kissing fish"!), 100nF and 1uF.

Trying different capacitor values from about 470pF through to 1000uF, and different capacitor types - ceramic, polyester, electro etc, I noticed the cyclic behaviour mostly in the 1nF to 1uF range.

Anyone experience similar variation with v2.08 or is this a very bad "pebkac" error?

And...

...first thing I noticed in the new 34461A firmware is that they screwed up the "4" character in the font...

...I don't like the altered "4" either.

Cheers
Andrew

[6thimage]

I don't think it's a pebkac - I've tried a poly cap and I get the same results (using sprung clips at the end of a metre long cable). I originally thought it was going to be an issue with certain capacitor types, but I don't think it is. Unfortunately, I don't have many through hole capacitors, which makes testing harder.

I've attached my version of what Andrew did, but condensed into a single trend plot - this is with a WIMA MKS-4 poly. The second attachment is the data (via CSV output) with the error margins as per the meter's specifications (61A) - I was wondering if the oscillations were within the error margins, which they are just for the 1 nF, but they aren't for the 10 nF range. There is no extra data in the gnuplot graph, the different ranges are done at different apertures / PLCs.

I did also notice that the measured value changes quite significantly between the 1 and 10 uF ranges, which is what the third image (screen 2) shows, with me creating a square wave by switching between the ranges (this was a tantalum cap).

I think part of the issue with the capacitance mode is that it is designed as an extra utility feature, rather than a main measurement mode - hence the lack of options with it. It also measures the capacitance by charging and discharging the capacitor (essentially using the ohms current source and voltage measurements), which means it will differ from meters those that use the frequency method. It also means that the measured value changes as the capacitor is charging (screen 3 of the attachments), I do wonder if this is considered in the accuracy specifications.

[rosbuitre]

My 34461A

VISHAY-RODERSTEIN MKP-1837 10nF / 100VAC/160VDC, MKP-1837, Metalized Polypropylene, 1% Tolerance

Russian_NOS_Silver Mica - 1nF / 350V, Silver Mica Capacitor (SGMZ-A), 5%, Each

[bson]

I upgraded my 34465A to firmware 02.08 and discovered the DCV peak reading is very useful - I used it to quick check the performance of my HP 33120A function generator.  Result: within spec, with only a 0.4mV output offset; less than 1/2 digit of the 1mV resolution offset setting.  The peak is about 2 digits better, 1.39998V P-to-P for a 700mV setting.  It's pretty neat that this DMM alone can be used to run through either the Q (quick) or P (performance) verifications since they're done at 1kHz/100Hz (depending on function), and this DMM is precise enough!  Is it possible to measure DCV peak as a primary measurement?  I poked around and couldn't find a way off hand...

I also noticed that holding down a key, including soft keys, for a couple of seconds brings up the help section for that function!  Brilliant!

[HighVoltage]

I also noticed that holding down a key, including soft keys, for a couple of seconds brings up the help section for that function!  Brilliant!

Yes, all new Keysight gear has that built in help function and it is really neat, also on the new scopes.
I don't think you can make the DC Peak to the primary measurement, at least I have not figured it out, if it should be possible

Yesterday, I compared my 34470A to a calibration lab just for fun and hooked it up to a calibrated Fluke 732A voltage standard.
The meter was warmed up for a few days in the cal lab at 23°C and the CAL button was pressed, before we took some measurements.
It was really nice to see 9.999,998 VDC on the 34470A and an almost flat line in the trend chart.
Actually the value written on the 732A was 9.999,999,99
The labs 3458A showed 10.000,000,4

So far, my 34470A seems to be a really reliable meter and I am very happy with it.

[Lizerd]

How do I buy the 34465A with Digi option ?,
Do I have to request/order the digi "unlock code" from Keysight direct ?
Any one ordered a extra option for your 344xx ?

[HighVoltage]

I ordered the 2M Reading Memory and the unlock code was sent to me by email

The fast DIG option is only available if you place it with your order.
Based on Keysight information, it is not user installable.

[6thimage]

Really? Both the DIG and MEM options are just unlock codes (well license files) - there are no physical changes to the meter. Even the GPIB option is designed to be fitted by the end user. It would be mad of them to restrict you to buying an option only when your purchase the device.

[dadler]

That's unfortunate -- I was hoping to add the digitize option eventually.

Newark will sell it to you it seems: http://www.newark.com/keysight-technologies/3446digu/digitizing-advanced-triggering/dp/30Y4300

[kwass]

That's unfortunate -- I was hoping to add the digitize option eventually.

Newark will sell it to you it seems: http://www.newark.com/keysight-technologies/3446digu/digitizing-advanced-triggering/dp/30Y4300

Page 27 of the manual (http://www.farnell.com/datasheets/1898969.pdf) says it's "Distributor or End Customer" installable -- it's just a license.

[Lizerd]

Yeep it's a license

34465A-DIG Digitizing and advanced triggering license
34465A-MEM 2 MB memory license

So I have to order it direct from KeySight, was hoping a could get it from the same distributer as the main unit.

But is't Newark is the same as Farnell ?

[HighVoltage]

All the German distributors have something like this on their website for the DIG option:

Option zum Keysight Digital-Multimeter 34470A:
Digitizing und erweiterte Triggerung ab Werk installiert.


Usually this means an installation by Keysight, before it is delivered to the enduser.
But who knows, may be it is a mistake.

I will call a German distributor tomorrow and will find out .

[bson]

The 3446DIGU option is listed as customer installed...
http://www.keysight.com/main/editorial.jspx?cc=US&lc=eng&ckey=2340971&nid=1119130.0.00&id=2340971
Maybe there are different product codes depending on whether it's factory or customer installed?