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

[Alex Nikitin]

 

are you able to measure the current source in 4W mode? I'd be interested in this failure mode, as KS probably won't give a feedback.

Unfortunately, I couldn't measure the current as our HP3458A is out for calibration, however from the data I've collected on other ranges it looks like the "jumps" are proportional to the measured value for a particular range (i.e. measuring 500 Ohm on 1K range gave me jumps of ~10ppm from 500 Ohm), and that reasonably confirms the current source problem. And I have the serial number recorded  .

Cheers

Alex

[Alex Nikitin]

An update. I've received the replacement 34465A last week and it was on continuously for about 100 hours so far. I've run a log over the weekend measuring the LTZ1000 and the meter drifted about 4ppm down, however ACAL this morning returned the results to within 1ppm of measured by HP3458A previously. The resistance ranges are now nice and stable, the attached graph shows 1 hour log for two 34465A and HP3458A  - all for NPLC 100 measuring my 10K LTC450C reference resistor (taken at different times, obviously, but with a fairly constant temperature in the lab for each run). Keysight UK was very helpful in resolving this problem. I aslo should get the 3458A Opt 002 back from them later this week so I will have a freshly calibrated reference.

Cheers

Alex

[TiN]

Thanks for followup, Alex Nikitin. That was hell of a wacked meter before. Sorry for doubting your measurement skills  .

[HighVoltage]

What a huge difference between those two 3465A meters. Great that Keysight solved this for you.
But also a surprise, that your first meter has left the factory like that.

[Alex Nikitin]

What a huge difference between those two 3465A meters. Great that Keysight solved this for you.
But also a surprise, that your first meter has left the factory like that.

Yes, it was an unpleasant surprise. The second meter meanwhile has settled and does not drift as in first few days. Here is an overnight run using the same LT450C 10K resistor "standard", with the room temperature log as well. Vertical scale 1ppm/1C/div

Cheers

Alex

[Alex Nikitin]

Another update. The replacement K34465A was running 24x7 for last two months and today I've paired it with the Fluke 731B. Below is the result of a 3 hours run at a fairly constant temperature in the lab (about 24C). I am rather surprised at a good noise performance and stability. It looks like the reference in the unit is a pretty decent one. Measurements are done at 10s intervals, NPLC100, vertical scale 0.5ppm/div.

Cheers

Alex

[upover]

Keysight have terminated their offer of a free MEM upgrade for the Truevolt DMMs.
Anyone know if this is a returning offer?
I'm buying mine NOW from Farnell Norway 
After reading a ridiculous amount of posts, I have concluded that a 34465a is the sweet spot.

[F64098]

Keysight have terminated their offer of a free MEM upgrade for the Truevolt DMMs.
Anyone know if this is a returning offer?
I'm buying mine NOW from Farnell Norway 

The registration period ends february 28nd.
You don't have to present the bill or other papers.
You just have to claim, you bought it on december 29nd...

HTH

Frank

[Messtechniker]

Just for the heck of it:
I know the 34465A is capable if measuring its internal temperature whereby
the sensor is located at the front terminals. This internal temperature can
be displayed externally too. Great. But can it be shown on it's own display?
And even processed as any other displayed value (included in dual display mode, add offset, histogram, record etc.)?
If not - this might be a contender for the next firmware update.

Yours Messtechniker

[LaurentR]

The measured temperature is displayed on the calibration screen. I haven't seen any way to use it elsewhere.

[LaurentR]

Note that I don't know how good the internal temp sensor is. I have a 34461A and a 34465A and they are about 2C different...
I asked Keysight a while ago about this and they said:

"The temperature shown in the 34461A calibration window is the internal temperature of the instrument. The temperature sensor is located on the PC board. This temperature/sensor is not calibrated. Changes in the internal temperature should correlate to changes in the ambient temperature."

[kj7e]

It looks like our 34465A is failing the noise spec on 10K. I'll talk to Keysight on Monday. It would be nice if somebody with the 34465A would repeat this resistance test (10K, 100NPLC, 10 sec logging interval).

Cheers

Alex

I know your meter was replaced already, but I was curious to try this myself.  10K, 10NPLC over 12 hours with temp plot overlay.
VPG, Z201 in a shielded enclosure;
https://www.eevblog.com/forum/metrology/resistance-standard-shielding-how-important-is-it/

So if I'm reading this correctly, this shows 1.4ppm Std dev over 12 hours, much of that is tempco;

[kj7e]

An update. I've received the replacement 34465A last week and it was on continuously for about 100 hours so far. I've run a log over the weekend measuring the LTZ1000 and the meter drifted about 4ppm down, however ACAL this morning returned the results to within 1ppm of measured by HP3458A previously. The resistance ranges are now nice and stable, the attached graph shows 1 hour log for two 34465A and HP3458A  - all for NPLC 100 measuring my 10K LTC450C reference resistor (taken at different times, obviously, but with a fairly constant temperature in the lab for each run). Keysight UK was very helpful in resolving this problem. I aslo should get the 3458A Opt 002 back from them later this week so I will have a freshly calibrated reference.

Cheers

Alex

My attempt to duplicate your 1 hour, 100NPLC, 10K 4W, 2ppm full vertical scale with lines at 0.5ppm;
Edit, typo in the chart title, should be 34465a.

[Faith]

Hey everyone. So I've finally managed to get my new lab set up after finally securing, designing and renovating the interior of my home.

I did however have everything in storage for several months during my housing woes to make moving around easier.

So anyhow now that everything's back up and running I did a quick check on all of my equipment today and I realised that my 34465A is a lot noisier than I remember on the ohms range (particularly 10K-100K).

When measuring a 100K resistor on the 100K ohms range at 10PLC the reading jumps after each update with a variance of up to 5 ohms.

At 100PLC it's a little less noisy (though I suspect due to averaging?) but I still see a variance of several hundred milliohms following each update.

I don't have a proper sealed resistance standard but I do have a pile of Vishay PTF56 resistors lying around which I used for testing (during the afternoon, no lights, and everything else powered down).

Is this normal? I don't recall this behaviour when I last used this instrument.

It's been running for an hour now so I've just cleared its statistics and will let it gather more data for the next couple of hours.

[floobydust]

Try put the cap/resistor under test right in the input jacks, compared to at the end of the test leads.

I find the LSD moving around so much on 34461's on resistance and worse on capacitance, I had to laugh at all those supposed extra digits.

I traced it down to some sensitivity to hum or common-mode noise on the 34461's. Even on a "quiet" bench with no other equipment. You new lab might have more noise/different grounding and layout.

[Faith]

Try put the cap/resistor under test right in the input jacks, compared to at the end of the test leads.

I find the LSD moving around so much on 34461's on resistance and worse on capacitance, I had to laugh at all those supposed extra digits.

Thanks for your quick reply!

I'll give that a try later once I've gathered all the data for my current test run!

You new lab might have more noise/different grounding and layout.

Yep that came to mind >,<"... though every other measurement (volts & amps) are pretty quiet.

I'll also point out that I do not by any means require super-accurate measurements (hobbyist and all), but my 34465A is my best DMM so I treat it as my reference point for every other meter and hence I like to make sure it works properly.

So just want to make sure that there isn't any other underlying problem with the meter that's causing this or potentially any other issue.

[HighVoltage]

Do a 4 wire resistance measurement of a 1 Ohm resistor and compare the noise level.

Any kind of RF transmitter, phone, wireless router, neighbors router and so on can have a significant influence on the digits, especially when you measure a 100k resistor with 2 wires. I don't think there is something wrong with your 34465A.

[floobydust]

As far as I investigated, something for others to keep an eye on, these meters use a mains SMPS and associated Y-capacitors.
So they generate their own common-mode noise- that seemed to be what I was seeing.

I would be curious if it's the lab/setup noise or the meter's have a sensitivity to RF.

[TheSteve]

As far as I investigated, something for others to keep an eye on, these meters use a mains SMPS and associated Y-capacitors.
So they generate their own common-mode noise- that seemed to be what I was seeing.

I would be curious if it's the lab/setup noise or the meter's have a sensitivity to RF.

What leads you to believe these meters use a SMPS?

[HighVoltage]

As far as I investigated, something for others to keep an eye on, these meters use a mains SMPS and associated Y-capacitors.
So they generate their own common-mode noise- that seemed to be what I was seeing.

These meters and previous models of Agilent meters have all a linear power supply and no SMPS !

[floobydust]

Really?! After the cooling fan roar, as my first multimeter ever with a fan    I assumed Keysight went SMPS for small size.
This would be good news. So it probably has a line-filter with Y-caps then.

[TheSteve]

Really?! After the cooling fan roar, as my first multimeter ever with a fan    I assumed Keysight went SMPS for small size.
This would be good news. So it probably has a line-filter with Y-caps then.

So you just decided to post that they have a SMPS without actually looking? That is a pretty bold statement to make, a little research wouldn't hurt next time. The Keysight frequency counters, AWG's etc use a SMPS, the DMM's are all linear.
The fan on my 34461A is pretty darn quiet, it is actually the least noisy of all of my fan equipped gear - but I'd still prefer it didn't have a fan and it doesn't seem to need it in a cool environment.

[floobydust]

I'm wrong, Dave's 34461A teardown pics show a power transformer and looks like POL buck converters.

After seeing a simple capacitance measurement noisy and then perfectly quiet with no test leads, I wrongly assumed these have mains-SMPS contributing noise. What else could it be, as no other gear in the lab suffers from this.

Bought two 34461A's to replace aging 34401's and thoroughly disappointed.  Would not purchase again.
Sorry, I am not a fanboy of these meters.

[TheSteve]

It's too bad they don't meet your expectations. When the 3446x series was released they didn't even have capacitance measurement, it is quite possible it was a bit of an after thought. Also noteworthy is that the 34460A does not have a fan in it, only the higher models do.

[MrW0lf]

Just as side note... Even mains power today is mostly much more than power. Here we have this:



Can imagine it can create all sorts of weirdness due to intermittent nature.