New Keysight 3458A - in black

[Kleinstein]

I think is does not makes much sense to do larger changes to the 3458 than just the replacement of obsolete parts and may be some changes is cases to make it easier to produce. Many users want the backwards compatibility - so being close to the original is a big plus.  In some areas the available parts have changed a lot (e.g. much better CMOS switches, better integrated ADCs) and a modern DMM would thus be different quite a bit. So it's either those more cosmetic changes or an essentially new design. I don't think the new black edition would be good for another 30 years.

[MiDi]

There are lots of changes inside.
We had a very interesting talk @Maker Faire with PTB about this and they know what they are talking about.

Edit: link corrected

[TiN]

I would hold on to golden aged 3458As and wait to see if new black design/refresh units are any good. We are spoiled by aged well-cared 3458A performance , that expectations are much higher that just same never updated specifications from 1989. Even something like replacing few parts from obsoleted stock to new RoHS parts could cause issues and worse performance.

We could do some guesswork bit more once photos of internal guts are revealed. Or Keysight could team up and send few random units to some voltnuts or researchers to do a public comparison tests. Chances of that happening? I expect zero, very niche meter this is, and customers who buying these don't need reviews on some blog anyway

[iMo]

TiN and all the voltnuts who glorify the old 3548A are the precision DMM's progress inhibitors actually. The DMM metrology grade industry assumes TiN and all volnuts are happy with the existing 3458A base and they are not motivated to develop something new 

[Kleinstein]

The market for the metrology grade DMMs is small. There are actually 2 new Fluke meters that may do the job.
With instruments  already that good, one can not expect large improvement. So even if a new, really good one comes out, a lot of labs would still stick to there old meters - at least as long as the old ones are still running. A meter with a well documented history gets more valuable with time.  A new type would also need to earn it's reputation over time. At this level even coming from Fluke or Keysight is not a guarantee for good long time stability. Time will tell (in some 3-5 years) how good the new Fluke ones and the black edition actually are.

[floobydust]

TiN and all the voltnuts who glorify the old 3548A are the precision DMM's progress inhibitors actually. The DMM metrology grade industry assumes TiN and all volnuts are happy with the existing 3458A base and they are not motivated to develop something new 

Progress in electronics is all about lowering cost, reducing size and speeding up design cycles etc. - any of which are enemies to metrology.
You need gold and lead, you need big ugly through-hole parts with less physical/electrical/thermal stress and fine dielectrics, ceramic substrates, you need JFETS galore, and long term testing. TO-99 forever!

I would say analog electronics at this level has deteriorated. The old guard has died off and fresh engineers who think plopping in SMT op-amps is going to be a piece of cake here, are in for a colossal waste of time. Even the RoHS-driven metallurgy changes alone, are a science in and of itself that few people in the world would have expertise in.

"new" would be funny as Keysight will never match or beat the old design. The old dogs here know and I know because the 3458A was a massive R&D expense, like landing on the moon.

[NoisyBoy]

It is official in the U.S., just got this from Keysight today:


The 3458A has been the standard 8.5 digit DMM for over 30 years, with leading-edge DCV linearity accuracy and lowest internal noise performance.

The refreshed Keysight 3458A is RoHS compliant, has a new industrial design, includes extended memory as standard, and is 100% code compatible with legacy versions. You will continue to enjoy the superior stability and accuracy that this product is known for, whether you use it for precision measurements with the highest degree of accuracy, as a transfer standard for instrument calibration, or for high throughput DMM measurements in manufacturing.

It is nice to see them updating the benchmark, the $10,987 price tag is reasonable considering how much old ones fetch in the used market.

[Keysight DanielBogdanoff]

A few comments

Even the RoHS-driven metallurgy changes alone, are a science in and of itself that few people in the world would have expertise in.

"new" would be funny as Keysight will never match or beat the old design. The old dogs here know and I know because the 3458A was a massive R&D expense, like landing on the moon.

Yes, it was definitely a challenge to get the new design out, we basically didn't ever update the 3458A design because it worked! And it was an absolutely massive project like you say - it took many, many, many years to get out the door (so I've been told). Like, maybe the longest R&D project we've had as a company, easily top 5-10.

There are lots of changes inside.
We had a very interesting talk @Maker Faire with PTB about this and they know what they are talking about.

Yup, pretty significant overhaul.

We could do some guesswork bit more once photos of internal guts are revealed. Or Keysight could team up and send few random units to some voltnuts or researchers to do a public comparison tests. Chances of that happening? I expect zero, very niche meter this is, and customers who buying these don't need reviews on some blog anyway

Challenge accepted. 

[0culus]

You could send one to me for uhhh, "review"... 

[kj7e]

We could do some guesswork bit more once photos of internal guts are revealed. Or Keysight could team up and send few random units to some voltnuts or researchers to do a public comparison tests. Chances of that happening? I expect zero, very niche meter this is, and customers who buying these don't need reviews on some blog anyway

Challenge accepted. 

Yes, please send one to TiN for a review!

[Zucca]

Yes, please send one to TiN for a review!

#sendittoTIN

+1

[coromonadalix]

#sendittoTIN

+1

[NoisyBoy]

Great call Daniel,

With its venerable reputation, there is always skepticism when changes are made.  If I were in the market for one, my intention would be to go with a new unit with a couple options rather than buying used and inheriting someone else's issues and no warranty.  And that is unless the new design is inferior to the old models, which I hope would not be the case.  Such a comparison would be invaluable to people who share my feeling, as well as labs who are ready to retire the fully depreciated 3458A workhorses and upgrade to the new version.

The only change I wish Keysight had made was to do away with the old VFD display and update it to one without burn-in issues.  For equipment that are left on pretty much permanently, the VFDs are killers.  While it was good 30 years ago, there are so many new technology that are more durable and easier to read, even if it does not provide any additional information.

A few comments

Even the RoHS-driven metallurgy changes alone, are a science in and of itself that few people in the world would have expertise in.

"new" would be funny as Keysight will never match or beat the old design. The old dogs here know and I know because the 3458A was a massive R&D expense, like landing on the moon.

Yes, it was definitely a challenge to get the new design out, we basically didn't ever update the 3458A design because it worked! And it was an absolutely massive project like you say - it took many, many, many years to get out the door (so I've been told). Like, maybe the longest R&D project we've had as a company, easily top 5-10.

There are lots of changes inside.
We had a very interesting talk @Maker Faire with PTB about this and they know what they are talking about.

Yup, pretty significant overhaul.

We could do some guesswork bit more once photos of internal guts are revealed. Or Keysight could team up and send few random units to some voltnuts or researchers to do a public comparison tests. Chances of that happening? I expect zero, very niche meter this is, and customers who buying these don't need reviews on some blog anyway

Challenge accepted. 

[IanJ]

There can only be one.........anybody else in terms of metrology will just comment on it's pretty colour.

#sendittoTIN

+1

Ian.

[IanJ]

The only change I wish Keysight had made was to do away with the old VFD display and update it to one without burn-in issues.  For equipment that are left on pretty much permanently, the VFDs are killers.  While it was good 30 years ago, there are so many new technology that are more durable and easier to read, even if it does not provide any additional information.

I agree, but kinda dissagree also!........my 3458A is 20yrs old next year, and yes it's dimming and burning in a little but it isn't bad. For it's 20th birthday it's going to get a new VFD next year.......:-)

Ian.

[MiDi]

The only change I wish Keysight had made was to do away with the old VFD display and update it to one without burn-in issues.  For equipment that are left on pretty much permanently, the VFDs are killers.  While it was good 30 years ago, there are so many new technology that are more durable and easier to read, even if it does not provide any additional information.

I agree, but kinda dissagree also!........my 3458A is 20yrs old next year, and yes it's dimming and burning in a little but it isn't bad. For it's 20th birthday it's going to get a new VFD next year.......:-)

Ian.

Best age as a classic car or Youngtimer - as we call it in Germany 
Mine is becoming a modern classic (Oldtimer) and far a thing from the past.
Received it in nearly factory condition including original equipped nvram with healthy calram.
Men, it was build when I was a innocent 10 year old boy.
Around the time where east & west united and end of cold war...

[Keysight DanielBogdanoff]

The people have spoken! I'm working on the #sendittoTIN angle

[NoisyBoy]

Daniel, that’s excellent news, can’t wait to see the comparison.  Thanks for your constant support for the community, it makes me feel good as a Keysight shareholder and long time Keysight/Agilent equipment owner.

[EE-digger]

Hate is a strong word but I hate black instruments.  Part of it may be my aging eyesight but the greater part is the lighting required to see nomenclature adequately.

I hate the LeCroy 3054's at work for that reason (and MANY others).  I carry my Keysight (in its cool looking black bag).

Black is very cool looking but then some smart people add either grey text or purple text (i.e. Sony Vaio and other keyboard 2nd and special functions).  Then you need sunlight to see it well      I do have a shiny black car and love it !

[Berni]

But hey at least they didn't go for the old "Keithley brown" color.

[TiN]

Perhaps Keysight could offer white-color front panel for older engineers in large business customers  Benefits of full compatibility between old and new meter. 

The people have spoken! I'm working on the #sendittoTIN angle

EEVBlog folk will get me into trouble. I guess season for test suggestions and wishes is open. I'd do usual comparison suite, like noise performance, stability vs external LTZ/732 references, ACAL performance and INL tests. ACV tests would be also interesting, as it is yet another underestimated part where 3458A really shines.

[Berni]

Perhaps Keysight could offer white-color front panel for older engineers in large business customers  Benefits of full compatibility between old and new meter. 

Hmm they have been boasting how its all still compatible with old SCIPI scripts, but they never said anything about being compatible with old engineers. Clearly Keysight has some room for improving compatibility of this new instrument

[TiN]

3458A does not support SCPI. At least old one. I'd expect new one is same in this regard?

[Berni]

Sorry i meant to say GPIB 

I got into a bad habit of calling test gear automation SCPI since modern stuff uses all sorts of interfaces (Serial,USB,LAN...) for remote control but all talk SCPI over those. So i subconsciously threw GPIB in the same bucket even tho old gear does not talk according to the SCPI standard (Like the 3458A).

[2N3055]

... ACV tests would be also interesting, as it is yet another underestimated part where 3458A really shines.

That is very good point. It would be interesting to see.

Off topic, but it would be also interesting to see comparison of 34465A (or any other TrueVolt, sampling type AC RMS ) meters compared to older generations with RMS converter chip, especially for linearity around zero...