Keysight 34460A/34461A/34465A/34470A Maximum Allowable Voltage change notice.

[fonograph]

I want to apologize,I was making my first message about this when Dave posted just the first video,I didnt know that there is the long video and that this was just short part of it.It looked to me like Keysight send him just to do that 4 minut video,so I take my words back regarding acusation from my side that its "damage control bs".

That being said,I stand behind my logical analysis of this happening.It might seem from PC screen that I am hating on Keysight,actualy they are my number 1 favorite electronic test equipment brand with Rohde Schwarz close second.While this might seem like red flag,overall Keysight in my eyes have least red flags,no manufacturer is perfect 100 percent all the time,I still believe they might do the right thing and come clean and tell what part it is.

[_Wim_]

I think the following has happened (speculation only):
- QA/Safety got notified of a potential  issue
- They checked the specs (purely on numbers), and contacted the engineering department
- The engineering department did not have an "approved written rationale" or "documented evidence" why is was ok to use this part in this configuration (because they always used it like this, and the original design decisions were not documented or lost)
- According to procedure, QA/Safety has X days to issue a field alert (typically 3 days)
- That time was too short for the team from engineering to prove they were correct (like explained in the video, they were also benchmarking others who do this the same way, but this costs time to organize and document a benchmark)
- Hence the field alert was issued by QA/Safety
- After some hard work from engineering they succeeded in convincing QA/Safety that the design was ok after all
- The alert was cancelled, but the damage was already done...

But, as this issue has been in the public, and no changes were made by Keysight and the culprit part was not made public, they must be really sure about their entire product. Imagine a safety issue would happen in the future, everybody will say that they “knew” about this, and did not act on it (which is a lot worse than being “surprised” by an “unknown” fault)
So I would say their safety/quality system works as it should (safety first), and I have no doubt the product is safe.  But off course, this is only speculation on my part from seeing similar thing happen in other industries…

[Someone]

They are likely trying to stop people like you from going ape over a single spec sheet figure that may not be relevant without a ton of other engineering test and historical data that would likely be difficult to comprehend to Joe Average.

Or it could be that the stresses across the part(s) only exceed a specification in transient or fault conditions (bootstrapped etc) where the part is used beyond its paper specification but has no safety issue to the user. These are the sorts of difficult areas where a quick look at something could appear dangerous until all the requirements and failure modes are considered. That or as you say many parts don't have transient specifications unless you collect the data yourself which then becomes valuable proprietary information.

[EEVblog]

I am noob,I admit it,but I dont go ape for irrational reasons when I discover some product uses under spec parts in it.

Right there is your problem. If you were actually experienced in the industry then you would know that over-speccing parts happens a lot, and for very legitimate engineering reasons.
In my 121GW multimeter for example we are using a chip below it's rated voltage rail specification. We have the OK from the manufacturer to do this and have done extensive testing to ensure it works fine. It is most definitely not for cost reduction reasons.
There can be many reasons to do this, too many I couldn't even explain them all, but the most likley in this case is that either:
1) There is no other part that does the job
or
2) This is best part that does the job based on a variety of specs. It's not always possible to find the best component.

And the engineering cost of evaluating all this would be oodles more than any supposed "cost saving" you are hounding on about.

For example Rigol uses under spec ADCs in their scopes,I didnt go "ooo!!! Rigol uses under spec ADCs ooOH my god WORST T&M MANUFACTURER EVER NEVER GOING TO BUY FROM THEM AGAIN OOOOOOO!"

You do realise that by Rigol doing that (and testing the parts to ensure they worked) changed the entire low cost test and measurement industry!

I hope that in time you'll learn some respect for what's possible in engineering and the challenges faced by companies in product design. It's not all about the bottom line.

And yes yes,Keysight doesnt care about what I think,but I am not alone who is curious what part it is.

Sure we'd all love to know, but once again your industry inexperience has prevented you from knowing their would be some very legitimate reasons why a company would not say this publicly.

[EEVblog]

- After some hard work from engineering they succeeded in convincing QA/Safety that the design was ok after all

And having had to have done this myself I can assure you the amount of work needed to convince a QA/Safety team within a large organisation to go with something "outside the banner spec numbers" is phenomenal.
People didn't just wave their hands here, extensive testing and investigation would have gone into this.
Those who have never done this would have no idea how much work is involved. I'd trust Keysight 100% on this.

Also, a huge part of this would be that this hasn't been an issue in at least 25-30 years they have been using this part in this way. Clearly it comes from the 34410A days and as John said, they have shipped 25,000 of those a year and it's the industry standard. If there was an issue it would have been found by now. So for anyone to simply come along and point at the datasheet spec and say "gotcha" is clearly demonstrably wrong.

[_Wim_]

Also, a huge part of this would be that this hasn't been an issue in at least 25-30 years they have been using this part in this way. Clearly it comes from the 34410A days and as John said, they have shipped 25,000 of those a year and it's the industry standard. If there was an issue it would have been found by now. So for anyone to simply come along and point at the datasheet spec and say "gotcha" is clearly demonstrably wrong.

But try to explain that to your QA/Safety department!

I can imagine the frustration of the engineering team when QA/Safety said they will go public, but are not involved in cleaning up the mess afterwards...

[s8548a]

BTW: for logging there are much better alternatives than BenchVue

Hi,

Can you/anyone please explain about the better alternatives available and on what model DMM's will they work for logging?

I am looking to buy a DMM and searching for various models like K2000, Agilent34401a, Rigol or Solartron but it is very confusing.

Thanks.

[emax]

After all, the whole fuzz has apparently turned out to be a no-problem, nice.

I am after a DMM and the 34465A is in my focus, but before I order, I would like to know whether this statement is still current (maybe I've overseen a dementi somewhere):

...
 
Next steps
Keysight has begun redesigning the 3446XA and 34470A digital multimeters to reinstate the original maximum input ratings of 1000VDC/750VAC. We have set a goal to ship the redesigned products by December of 2017, and will make every effort to begin shipment sooner than that.
 

Are there any official news about that?

[HKJ]

Are there any official news about that?

In David's interview with Keysight it is explained that it was some misunderstanding/missing documentation that was the reason, not any fault in the meter.

[HighVoltage]

I am after a DMM and the 34465A is in my focus, but before I order, I would like to know whether this statement is still current (maybe I've overseen a dementi somewhere):
Are there any official news about that?

Based on what John Kenny said in the interview with Dave, there was no problem, so there was no need to make any changes.
You should be totally alright, ordering a 34465A

[emax]

Thank you all, didn't watch the interview.




Will be my christmas gift - made by myself, though.

[anotherlin]

Just purchased a 34465A DMM.
It's a very recent one, factory calibrated on Nov 7th.
Serial starts with MY5750xxxx.
I just can't find the app-note pdf from Keysight stating which serial numbers starting with, have the max allowable voltage issue.

[Dr. Frank]

This technical note has been withdrawn... if it ever existed.
In my archive I only found the reduced specification (600V dc max.)

There has never been an issue, in reality.

Frank

[anotherlin]

This technical note has been withdrawn... if it ever existed.
In my archive I only found the reduced specification (600V dc max.)

There has never been an issue, in reality.

I know well that it is a non issue.
I just wanted to know if my device is one of the newer revision.

[HighVoltage]

You probably have to contact Keysight and ask them to be sure.
(Most likely they will not let us know)

Or open your 34465A DMM and look if the PCB has a new revision number.

[anotherlin]

Ok, I've found it again:

http://literature.cdn.keysight.com/litweb/pdf/34461A-04.pdf

So with a serial number starting with MY57xxxxxx, I have one of the new revised model.

Note in question 5 (Q5 page 3), the devices having the max allowable voltage "issue" are offered an extra year of warranty (for a total of 4 years).

[wraper]

Ok, I've found it again:

http://literature.cdn.keysight.com/litweb/pdf/34461A-04.pdf

So with a serial number starting with MY57xxxxxx, I have one of the new revised model.

Note in question 5 (Q5 page 3), the devices having the max allowable voltage "issue" are offered an extra year of warranty (for a total of 4 years).

I don't think that revised model was ever made. There was no issue to begin with, that was false alarm. Multimeters with later serial numbers had stickers over voltage ratings with reduced ratings. Later there was another service note which showed how to remove those stickers.

From that service note

3. A service note will be released when the enhanced 3446XA/3447XA Digital Multimeters is
available, tentatively in December 2017. For customers that still require 1000VDC/750VAC,
please contact Keysight Customer Contact Center and reference to the service note number.

[anotherlin]

I don't think that revised model was ever made. There was no issue to begin with, that was false alarm. Multimeters with later serial numbers had stickers over voltage ratings with reduced ratings. Later there was another service note which showed how to remove those stickers.

The "enhanced" or "revised" model may just be one with the "offending" part swapped with one with appropriate ratings.
Or indeed, just the same, as it is claimed (by Keysight) to be a non-issue. Or maybe, they've taken the opportunity to make some other unrelated changes.

I just wanted to know out of curiosity.