Author Topic: 3458A - black edition  (Read 48828 times)

0 Members and 3 Guests are viewing this topic.

Offline voltampere

  • Regular Contributor
  • *
  • Posts: 58
  • Country: de
Re: 3458A - black edition
« Reply #200 on: March 29, 2020, 04:13:52 pm »
Does that mean i should expect a rather "fresh" version on my new A3?

How shall i interpret the printing on the sticker?

1NC1-0017
903-4/03


My very uneducated guess here would be -> April 2004!

Now i am scratching my head....

Regards

Joerg
 

Offline martinr33

  • Frequent Contributor
  • **
  • Posts: 363
  • Country: us
Re: 3458A - black edition
« Reply #201 on: March 29, 2020, 05:08:29 pm »
Handling marks and handwritten magic marker numbers would seem to support the idea of an older date code. However, poking through some pictures and boards, I don't see a consistent numbering scheme for U180 that would lead to a firm conclusion.

What I do see is a consistent pattern of hand-written "magic marker" numbers. That pattern goes back a long way, and I wonder if it isn't some part of a pre-screening process for the packaged chips.

Also, the scratch on the chip suggests a fair amount of manual handling.

I did find this piece of trivia:

http://www.icproto.com/ab-profile.html

This looks like the company that used to package the ASIC that HP was using, until they replaced the ASIC with the daughterboard. The ASIC appears with brands Promex and Orbit, both brands associated with Quik Pak. Maybe the original vendor (Fujitsu) dropped support for the part, or the package, or both.
 
The following users thanked this post: TiN

Offline Dr. Frank

  • Super Contributor
  • ***
  • Posts: 2425
  • Country: de
Re: 3458A - black edition
« Reply #202 on: March 29, 2020, 05:12:53 pm »
Would have been fortunate that you first had a look on xdevs site, there are several different pictures of different A3 vintage, to afterwards make an educated guess.

So let me google it your you.. Obviously, U180 rarely has a unique D.C., neither that the format is standardized.

As your A3 board is assembled in 2019, that 903 seems to be the D.C.

4/03 is probably never used as a date format on semiconductors.
 

Offline syau

  • Frequent Contributor
  • **
  • Posts: 369
  • Country: hk
Re: 3458A - black edition
« Reply #203 on: March 30, 2020, 12:13:17 am »
Photo for some recently acquired board in March 2019

959518-0
 
The following users thanked this post: BFX

Offline chuckb

  • Frequent Contributor
  • **
  • Posts: 345
  • Country: us
Re: 3458A - black edition
« Reply #204 on: March 30, 2020, 01:16:48 am »
Some defective U180 chips from bbs.38hot
 
The following users thanked this post: iMo

Offline syau

  • Frequent Contributor
  • **
  • Posts: 369
  • Country: hk
Re: 3458A - black edition
« Reply #205 on: March 30, 2020, 06:12:24 am »
Photo for some recently acquired board in March 2019

(Attachment Link)

This board was a "Test Fresh Board" with Compl Date 21 Dec 2018.

The marking on the U180 is 1NC1-0017 730-1/17 (Hand written 49) on another "test Fresh Board" with Compl Date 11 Jan 2019
 

Offline iMo

  • Super Contributor
  • ***
  • Posts: 5164
  • Country: bt
Re: 3458A - black edition
« Reply #206 on: March 30, 2020, 10:59:30 am »
My guess would be the 0017 or 0108 is the production date (17w 2000, 8w 2001), the string like 1NC1, 1SJ8 is the revision code, and the 3digits number there is the Serial number (like 014, 812, 040..) within the batch (like they cook 1000 hybrids in a batch).
« Last Edit: March 30, 2020, 11:17:21 am by imo »
Readers discretion is advised..
 

Offline voltampere

  • Regular Contributor
  • *
  • Posts: 58
  • Country: de
Re: 3458A - black edition
« Reply #207 on: March 31, 2020, 08:50:44 pm »
Wow....

Lots of guesses.

Let me sum up:

a) my ( seemingly ) stupid guess -> april 2003 ( could be march 2004 as well...)

b) academic backed guess of "Dottore" Frank -> march 2019

c) imo's guess -> week 17 of 2000

Personally i would like Frank's guess the most, because it would mean that my chip is rather fresh.
Anybody who knows about the scheme for the marking?

Best regards

Joerg
 

Offline Dr. Frank

  • Super Contributor
  • ***
  • Posts: 2425
  • Country: de
Re: 3458A - black edition
« Reply #208 on: March 31, 2020, 11:38:21 pm »
Wow....

Lots of guesses.

Let me sum up:

a) my ( seemingly ) stupid guess -> april 2003 ( could be march 2004 as well...)

b) academic backed guess of "Dottore" Frank -> march 2019

c) imo's guess -> week 17 of 2000

Personally i would like Frank's guess the most, because it would mean that my chip is rather fresh.
Anybody who knows about the scheme for the marking?

Best regards

Joerg

Hi Jörg
Akademisch basiert? So ein Quatsch. Ich habe vielleicht ein bisschen mehr Erfahrung mit Date Code Formaten, auf Grund meines Jobs.
Das hier ist eine einfache Mustererkennung.. wenn überhaupt mal ein Anzeichen eines Date Code zu erahnen ist, in all den Beispielen, dann stimmt dieses dreistellige oder vierstellige Fragment immer ungefähr mit den Date Codes der anderen Bauteile auf der Leiterplatte überein, und diese ganzen anderen Ziffernfolgen eben nicht, wie zB 0017, das kommt doch schon seit 20 Jahren auf dem IC vor.
Dieses ewige Spekulatius hier ohne jegliche Logik, oder irgendeine Begründung warum das so sein sollte, nur ins Blaue rein, was den Leuten so gerade einfällt, das geht mir echt auf den Nerv, sorry.
Wäre toll, wenn das MM doch noch stattfinden würde, dann könnte man die authentischen Leute dazu befragen.
Frank
 
The following users thanked this post: Henrik_V

Offline martinr33

  • Frequent Contributor
  • **
  • Posts: 363
  • Country: us
Re: 3458A - black edition
« Reply #209 on: April 01, 2020, 03:17:20 am »
The real test will come when you check the drift on  the part.

Some of these are pretty much zero drift, but some are not. The part looks pretty much the same as it has been for decades, and it still goes through the magic marker process. The board says, "Made in Malaysia", so it is going to use the same production parts as the previous version.

I'd tag the whole board at somewhere in September 2019 (week 42).

Now - what did they do to replace that Elantec super high speed comparator? Looks like they might have used this part TL3016 - except it took 3 parts to replace one Elantec 3016. The TI is faster, but has higher offset currents.

Hmph - the TI part is 5V only, so it needs regulators, which is what the other parts look like. I still wonder if the TI part might offer better performance, although it seems unlikely from the spec. 

So there's an exercise to figure out why the black version has three of these in place of one elantec - making it equivalent in performance, or making it better? I'd assume that better performance results in less noise in the ADC which would be interesting.

If we could get a sharper picture of the bottom left corner of the board, that would be really helpful. I can't quite figure out what those other parts are. I thought they might also be 3016s, but the wiring is odd.


« Last Edit: April 01, 2020, 03:48:53 am by martinr33 »
 

Offline TiN

  • Super Contributor
  • ***
  • Posts: 4543
  • Country: ua
    • xDevs.com
Re: 3458A - black edition
« Reply #210 on: April 01, 2020, 04:17:33 am »
3016C comparators + 7805 and 7905 LDOs in SOIC, since original Elantec's were powered by +/-15V.
YouTube | Metrology IRC Chat room | Let's share T&M documentation? Upload! No upload limits for firmwares, photos, files.
 

Offline voltampere

  • Regular Contributor
  • *
  • Posts: 58
  • Country: de
Re: 3458A - black edition
« Reply #211 on: April 01, 2020, 06:55:48 am »
Frank:
No worries. I got your point...

I tried a quick shot with a bit more resolution.


Drift test is in progress.

30 minutes noise does look okay...
Plot is 500 samples with NPLC100.

Best regards

Joerg
 
The following users thanked this post: martinr33

Offline Dr. Frank

  • Super Contributor
  • ***
  • Posts: 2425
  • Country: de
Re: 3458A - black edition
« Reply #212 on: April 01, 2020, 07:14:08 am »
Frank:
No worries. I got your point...

I tried a quick shot with a bit more resolution.


Drift test is in progress.

30 minutes noise does look okay...
Plot is 500 samples with NPLC100.

Best regards

Joerg

Hi Joerg,
where did you get this A3 board from?
Did KS sell it to you as a spare part, or exchange for your old board?
So you dis-assemble it again for this photo?

Frank
 

Offline Dr. Frank

  • Super Contributor
  • ***
  • Posts: 2425
  • Country: de
Re: 3458A - black edition
« Reply #213 on: April 01, 2020, 07:16:47 am »
3016C comparators + 7805 and 7905 LDOs in SOIC, since original Elantec's were powered by +/-15V.

Hi Illya,
when will you reveal your photos of the Black Edition, and your review / characterization, or did KS not allow that?

Frank
« Last Edit: April 01, 2020, 08:14:56 am by Dr. Frank »
 

Offline Kleinstein

  • Super Contributor
  • ***
  • Posts: 14780
  • Country: de
Re: 3458A - black edition
« Reply #214 on: April 01, 2020, 07:51:40 am »
For the comparators they use the TL3016 powered from local +-5 V. In addition they limit the input voltage range with clamping diodes and limit the input side current (looks like 2 JFETs and 3 resistors in a rather conventional circuit). The comparator is not directly connected to sensitive nodes, so the slightly higher bias should not be a problem. If at all it would be the extra current from the clamps.
A different delay can have an effect as this effects the overshoot in the rundown and thus the times measured.
Besides the comparator they also changed the slope amplifier. There are now 2 diode for each direction. Chances are they also changed the old relatively noisy OP.
 

Offline voltampere

  • Regular Contributor
  • *
  • Posts: 58
  • Country: de
Re: 3458A - black edition
« Reply #215 on: April 01, 2020, 09:44:10 am »
The board is a spare part from KS germany.
Story is here:
https://www.eevblog.com/forum/metrology/hp-3458a-ad-board-spare-part-availibilty-and-cost/
I asked for exchange to get a lower price but they denied.

This shot was actually done shortly after i did the first pics with my mobile phone( the one i published before ).
So " ...tried a quick shot ..." is a bit misleading here.

Nevertheless is have no problems taking the board out again.
It is my second 3458 that is still work in progress.
I have done the basics already ( RAM upgrade, NVRAM exchange, line filter, elcetrolytics replace ).
Paid more for A3 spare than for the meter itself!!! ( but i somehow expected that... ).

Regards

Joerg
 

Offline Nixfried

  • Contributor
  • Posts: 18
  • Country: de
Re: 3458A - black edition
« Reply #216 on: July 03, 2020, 06:30:52 am »
Just found two new pictures showing the innards...





Look at that tiny digital board!
 

Offline Dr. Frank

  • Super Contributor
  • ***
  • Posts: 2425
  • Country: de
Re: 3458A - black edition
« Reply #217 on: July 03, 2020, 06:54:37 am »
Just found two new pictures showing the innards...

Look at that tiny digital board!

Illya has a complete picture session of all PCBs on his xdevs site, but for some reason, he is stuck publishing his review any further. :-//
Just sneak in his picture folders.

Frank
« Last Edit: July 03, 2020, 07:05:11 am by Dr. Frank »
 
The following users thanked this post: MegaVolt, syau, Nixfried

Offline matches

  • Contributor
  • Posts: 25
  • Country: de
Re: 3458A - black edition
« Reply #218 on: September 04, 2020, 03:16:52 pm »
When I look at the specs, it seams every value one can read in the 100 Ohm range has a higher specified tolerance compared to the 1 k Range.
Ppm of range is a factor of 10 higher in the 100 Ohm range and ppm of Reading is a little higher (12 compared to 10) in the 100 Ohm range.

So one would be better of skipping the 100 Ohm range? Of course just pure calculation from the spec sheet. 

Am I missing something?
 

Offline matches

  • Contributor
  • Posts: 25
  • Country: de
Re: 3458A - black edition
« Reply #219 on: September 04, 2020, 03:18:03 pm »
Sorry, double post
 

Offline Kleinstein

  • Super Contributor
  • ***
  • Posts: 14780
  • Country: de
Re: 3458A - black edition
« Reply #220 on: September 04, 2020, 05:20:45 pm »
The ppm of range error part is mainly the offset. This offset does not improve much going from the 1 V range (1 K) to the 100 mV range (100 Ohm range) so no surprise with the fixed error / drift part.
Also with voltage measurements the 100 mV range has only limited advantage over using the 1 V range - the absolute noise does not improve much, as the ADC is noise is quite good and already the 1 V range has about as much noise from the amplifier as from the ADC. Going to the 100 mV range does not help much with the amplifier noise (it helps a little with a lower impedance in the feedback network - thus a tiny bit lower noise there).

The scale factor for the 100 mV range does have a little more uncertainty (one more step away from the reference). So the 100 Ohms range is of limited value, at least for slow measurements. In some aspects the 1 K range may be the better choice.

With temperature sensitive resistors (e.g. RTD) even the 10 K range may be better as it uses a lower test current, that could be better for heat sensitive resistors. Other meters may have a special low power resistance more to measure 1 K with a lower test current.
 

Offline matches

  • Contributor
  • Posts: 25
  • Country: de
Re: 3458A - black edition
« Reply #221 on: September 05, 2020, 07:46:59 am »
Ahh, now I see it. Thanks  :D
The 100 mV range even cuts of one digit.

So I could have asked the same question here.

Regarding RTD measurements, that is where I noticed it.
I use Pt100 together with the ‘58 as temperature reference for calibration of thermocouples.
They are calibrated at 1 mA so the self heating error is included.
I should make a measurement... now I’m curious  ::) But it’s not the big source of error anyway.
 

Offline miro123

  • Regular Contributor
  • *
  • Posts: 216
  • Country: nl
Re: 3458A - black edition
« Reply #222 on: May 04, 2022, 12:31:26 pm »
Just a status update about 3458A Black review progress. I have collected lots of data and had nice photoshoot, and currently busy checking data and doing additional tests on my own 3458A's and other meters used in comparison. Black meter is back at Keysight now, and I plan to release review this month, including nice juicy thread in EEVBlog Metrology.

Hello
I'm trying to find the KS3458a review from Ilja. Where can I find it?

Best regards,
Miro
 
The following users thanked this post: MiDi

Online ScoobyDooTopic starter

  • Regular Contributor
  • *
  • Posts: 85
  • Country: fr
Re: 3458A - black edition
« Reply #223 on: May 07, 2022, 06:13:45 am »
Hello Voltnuts,
                        The link below gives an impressum about the performance of the KS3458A - B.E.         
                         (Kudos: Marco Reps) ...



Here is a link to TIN's article: https://xdevs.com/article/ks3458b/

Herzliche Grüße/Meilleures salutations/Best regards

Best regards
ScoobyDoo

« Last Edit: May 11, 2022, 08:10:50 pm by ScoobyDoo »
 
The following users thanked this post: miro123

Offline miro123

  • Regular Contributor
  • *
  • Posts: 216
  • Country: nl
Re: 3458A - black edition
« Reply #224 on: May 07, 2022, 09:19:10 am »
Thanks for sharing.
As usual Marco reps videos shows good combination of deep techncal knowledge, structured explanation and professional video processing. This 35min video is good example.
 
The following users thanked this post: egonotto


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf