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

[HighVoltage]

In dmm noise thread,High Voltage posted comparsion where both 34470 and keithley 7510 set to 10V DCV range and he is measuring warmed up stable reference.It shows 7510 gets stable in 40 seconds while 34470 needs 20 minutes to reach stability due to warm up.

Can someone do the same tests? I am specificaly curious if 34465 with its lm399 reference warms up faster.

Hello Fonograph
Sorry, I dont have a Keysight 34465A (not yet) but I compared here a 34461A (LM399) with a 34470A (LTZ1000). And there is almost no difference in warmup time.

- 34461A, increase of 210 mV during 30 min warmup
- 34470A, increase of 172 mV during 30 min warmup

So, I would suspect the 34465A to be in the same range.

[fonograph]

Thank you very much.I looked again and I made mistake,it was maxwell3e10 who posted that 65 vs 7510 graph.I think the 61 is goid enough,its obvious that it isnt the reference.

I dont know what to do,34470 have normaly working autozero and datasheet spec says it have 0.5ppm linearity but it turns on for 20 minutes and its temperature coefficient is double of 7510.
7510 is better in everything except that horrible useless autozero and as result its averaging almost doesnt work at all,also the linearity is 2ppm,thats probably not a issue,but 34465 is 1ppm and cost 1/3.

[HighVoltage]

If you don't need the resolution of the 7 1/2 digit 7510 or 34470A, then the 34465A should be perfect for you.

[maxwell3e10]

DMM7510 has the potential to be a great meter.  Its temperature coefficient is specified at 0.15ppm/degree, a factor of 10 better than for 34470. That is probably the reason it reaches steady state much faster. In my tests for dynamic temperature changes at 1000 sec it was about a factor of 2 better than 34470. That comes out to TC of about 0.4ppm/degree for 7510, worse than spec, although its not exactly the same kind of test.  DMM7510 also has much lower noise on 100mV and 1V ranges.  The interface is nicer and it has more capabilities.

That is why its so frustrating that the 7510 autozero firmware problem has not been solved. Also the interface is still buggy (my latest success in crashing it involved using temperature secondary measurement). So there is no clear winner.

[HighVoltage]

DMM7510 has the potential to be a great meter.  Its temperature coefficient is specified at 0.15ppm/degree, a factor of 10 better than for 34470. That is probably the reason it reaches steady state much faster. In my tests for dynamic temperature changes at 1000 sec it was about a factor of 2 better than 34470. That comes out to TC of about 0.4ppm/degree for 7510, worse than spec, although its not exactly the same kind of test.  DMM7510 also has much lower noise on 100mV and 1V ranges.  The interface is nicer and it has more capabilities.

That is why its so frustrating that the 7510 autozero firmware problem has not been solved. Also the interface is still buggy (my latest success in crashing it involved using temperature secondary measurement). So there is no clear winner.

I totally agree with all your points.

In comparison, I like both meters a lot.

The 7510 is repeatedly more stable than 34470A (in the ppm range) and I am not even using the ACAL feature on the 7510! The ACAL on the 34470A is sometimes flaky and over corrects and then needs few min to correct itself again. But the really amazing feature of the 7510 is the ultra fast warmup time.
Here is a new picture from today below.
After turn on, I cleared the memory after 10 seconds because of the weird saw-tooth graph at the very beginning. And then in 120 seconds after that, the instrument is stable.

Another very important thing to consider is service. (within a warranty or outside of warranty). Try to get any kind of service with Keithley / Tektronix is a horrible experience, to say the least. If it comes to service alone, I would never buy a Keithley / Tektronix instrument again.

If you have any problems with a Keysight instrument, it is being taken care of immediately and so far always to my fullest satisfaction. This summer I had both of my 34470A serviced at the Keysight  Böblingen / Germany facility and I was (again) astonished about the really amazing service attention I got.

[dr.diesel]

Another very important thing to consider is service. (within a warranty or outside of warranty). Try to get any kind of service with Keithley / Tektronix is a horrible experience, to say the least. If it comes to service alone, I would never buy a Keithley / Tektronix instrument again.

I will second this statement.  It is unlikely I will choose Keithley/Tek in the future, service (and new product follow-up) is just plain horrible.

[dr.diesel]

That is why its so frustrating that the 7510 autozero firmware problem has not been solved. Also the interface is still buggy (my latest success in crashing it involved using temperature secondary measurement). So there is no clear winner.

Even more annoying is that I believe they are simply ignoring the AZ issue.

FW 1.6.4c is better, but I also still experience lock-up issues.

[Dr. Frank]

How linear are the 34460,61,65 and 70 adc? I read on 65,70 brochure that they are 0.5ppm.Keithley 7510 datasheet shows 1ppm + 1ppm of range.

1. What is adc linearity of 34460?
2. What is adc linearity of 34461?
3. What is adc linearity of 34465?
4. What is adc linearity of 34470?
5. What is this "1ppm +1ppm of range"? I dont understand it,how is it different to just 2ppm linearity?

edit: I found another brochure,this one is from Testequity and it says 61 = 2ppm,65 = 1ppm,70 = 0.5ppm
Interesting,I thought they 65 and 70 have same adc,I tought only difference is that 70 have ltz1000,correct me if I am wrong but voltage reference doesnt have anything to do with adc linearity?

http://www.testequity.com/documents/pdf/keysight/34461A-34465A-34470A-ad.pdf

34460 isnt mentioned there,what is its linearity?

I determined the linearity on two 34465A, and on one 34470A (from HighVoltage): https://www.eevblog.com/forum/testgear/keysight's-new-34465a-(6-5-digit)-and-34470a-(7-5-digit)-bench-multimeters/msg889219/#msg889219, and found no obvious difference between both models, apart from the reference.

Usually, the integral linearity is defined as the maximum deviation from a straight line (measurement compensated for offset and gain errors), related to full scale.
The datasheet parameter is therefore 1ppm of 10V for all models, whereas the 'ppm of reading' is not so clear.

Anyhow, all 3 instruments were well below the specified limits.
The 34460A is specified identical to the 461A.

Frank

[Tom45]

For those that did a warmup test of the Keysight meters, was the test from a cold start (plug it in), or from a warm start (in standby for some time and then turned on)?

Does the 7510 have a standby mode like the Keysight meters?

[Dr. Frank]

For those that did a warmup test of the Keysight meters, was the test from a cold start (plug it in), or from a warm start (in standby for some time and then turned on)?

Does the 7510 have a standby mode like the Keysight meters?

The Keysight meters (DMM) all do NOT have a standby mode!
The new ones do have a SW switch only, instead of a mechanical mains switch.
So even if the StandBy LED is lit, they are completely cold inside.

The initial drift in the first about 1/2 - 2h (depending on your stability criterion) is caused mainly by warming up of the resistor networks, but NOT of the voltage references.
The ovens of both the LM399H, and also the LTZ1000A warm up within 15 .. 30sec, and the output of the (more complex) LTZ1000 circuit is stable to < 1ppm within about 10 min.
See also my latest contribution in the 'Ultra Precision Reference  LTZ1000' thread.

Frank

PS: I attached a warm-up curve of an LTZ1000 module, running at about 50°C.
The output really settles within less 10 minuntes, so that should not be the problem in the 34470A, even if this reference is cooked at 90°C.

[dr.diesel]

Does the 7510 have a standby mode like the Keysight meters?

The 7510 doesn't either.

[Tom45]

The Keysight meters (DMM) all do NOT have a standby mode!
The new ones do have a SW switch only, instead of a mechanical mains switch.
So even if the StandBy LED is lit, they are completely cold inside.

Interesting. The Keysight manual refers to it as standby mode. What is the point of doing it the way they do?

[HighVoltage]

Interesting. The Keysight manual refers to it as standby mode. What is the point of doing it the way they do?

No benefit that I have found. Except the Keysight is showing a little yellow light, so you know the mains is plugged in.

Cold with unplugged mains or in standby, the boot-time and warmup-time does not make a difference on both instruments.

Boot-Times:
- Keithley DMM7510:  10 seconds
- Keysight 34470A:    19 seconds

[fonograph]

Can someone make that warm up graph with 3458 please?

[maxwell3e10]

I am curious why warm-up and start-up time are important in your application? Or is it just a measure of general temperature sensitivity?

[fonograph]

I am curious why warm-up and start-up time are important in your application? Or is it just a measure of general temperature sensitivity?

Becose I like to turn off my multimeter completly when I am done using it.Its big deal if I have to wait half hour every time I turn it on,especialy when 7510 exist that gets ready in 60 seconds.

Its not so much important,its more like I want to know the truth about how these instruments act,how they work,what to expect.Imagine that I have multimeter that warms up in 1 minute but I dont use it till 30 minutes after start becose I expect it to be still warming up,thats bad becose it waste time and electricity,or imagine I start using multimeter 1 minute after start and get bad readings becose I am not aware its warming up long time,knowing exact warm up behavior of the multimeter is is in my opinion good thing to know.

[nctnico]

It is not just the reference but the rest of the equipment needs to warm up as well. If you want to do really accurate measurements you'll need to wait until the test leads reach thermal equilibrium too. From what I have seen most test equipment needs about 30 minutes warm up time to meet the accuracy specs.

[Someone]

I am curious why warm-up and start-up time are important in your application? Or is it just a measure of general temperature sensitivity?

Becose I like to turn off my multimeter completly when I am done using it.Its big deal if I have to wait half hour every time I turn it on,especialy when 7510 exist that gets ready in 60 seconds.

Its not so much important,its more like I want to know the truth about how these instruments act,how they work,what to expect.Imagine that I have multimeter that warms up in 1 minute but I dont use it till 30 minutes after start becose I expect it to be still warming up,thats bad becose it waste time and electricity,or imagine I start using multimeter 1 minute after start and get bad readings becose I am not aware its warming up long time,knowing exact warm up behavior of the multimeter is is in my opinion good thing to know.

What absolute measurements are you making of that accuracy? Most devices under test themselves will need time to thermally soak before they are stable enough to have the meter dominate the error.

[HighVoltage]

Hello Fonograph
Sorry, I dont have a Keysight 34465A (not yet) but I compared here a 34461A (LM399) with a 34470A (LTZ1000). And there is almost no difference in warmup time.

- 34461A, increase of 210 mV during 30 min warmup
- 34470A, increase of 172 mV during 30 min warmup

So, I would suspect the 34465A to be in the same range.

OK, now I have a new Keysight 34465A, (made and calibrated in 2016).
And here I did a fresh comparison of the warmup-time for three instruments.
Interestingly the 34465A needs about the same time as the 34470A but the graph looks a little different.

It seems that 30 min warmup is not enough for the Keysight instruments and they really need one hour to stabilize completely. And the Keithley 7510 is almost stable after only 3 min.

Here are the graphs from today.

[0xfede]

Hi all folks.

In the past weeks I had to take a ton of measurement from my 34465A and I needed a fast and convenient method to store both screenshots and csv data. So I wrote a small C# program that use socket for connectivity (completely driverless) that I like to share.
It should work with both 34465A and 34470A, I tested it on WIN 7 64bit and it downloads 2Mpts in just 35 seconds. As a bonus I've added a few graph tweaking tools like deleting points and adding calibration factors for use with shunts and other stuff.
Please let me know if I have to change something.

Best,
0xfede

[HighVoltage]

It should work with both 34465A and 34470A, I tested it on WIN 7 64bit and it downloads 2Mpts in just 35 seconds. As a bonus I've added a few graph tweaking tools like deleting points and adding calibration factors for use with shunts and other stuff.

Please let me know if I have to change something.

Best,
0xfede

Hello Oxfede
Your program works really well
I just entered my instrument IP address, and it worked.
Nice and easy.
I like it as well, that I can copy the image to the clipboard, well made

I would like the following change:
Somehow I don't like the instrument serial number shown on top of the data screen
May be you can add in the "settings", if the serial number will be shown or not or may be, give the instrument a name instead of the serial number.

Otherwise it will work perfectly for such a task of quickly collecting data.
Thanks so much for sharing this.

Here I tested a 10V Reference

[0xfede]

Thanks for your feedback HighVoltage.
As you requested I added the 'Display Serial Number' checkbox in settings.

Best,
0xfede

[HighVoltage]

Wow, you are quick!
Thanks

Just a few more suggestions:
1. How about also a check in the settings for time and date (Sometimes I don't want it shown)
2. And time instead of "6-08-01" in 24h: "18:08:01"
3. And date instead of "11_22_2017" in "22.11.2017"
4. And may be make the x-axis more clear of the unit meaning, may be choosing sec, min, h?

BTW, your little program is faster than BenchVue!
And I don't need the IOLib installed, runs on my work laptop under Win7

[HighVoltage]

0xfede

Your program works perfectly on both of my 34470A but not on the 34465A and not on the 34461A
May be you can extend it to those instruments as well?

[Messtechniker]

BTW, your little program is faster than BenchVue!

That doesn't take much