Author Topic: Forum Crowd Project - Simple 10V LTZ1000 Plug'n'Play Box (with 1k/10k hermetic?)  (Read 1708 times)

0 Members and 1 Guest are viewing this topic.

Offline AG7CK

  • Regular Contributor
  • *
  • Posts: 131
  • Country: th
In my view, there seems to be quite a few members who want to buy already built, tested and measured references.

I hope that by keeping it simple, a project like this could be finished in a few months. So who wants to buy, and who wants to take our money and deliver the boxes?
 

Offline AG7CK

  • Regular Contributor
  • *
  • Posts: 131
  • Country: th
Status post
 

Offline AG7CK

  • Regular Contributor
  • *
  • Posts: 131
  • Country: th
When I write 1k/10k resistor in the title, it could be in the same box, or a different box (option). I am quite sure there is a need for measured resistor references also.

Alternatively it could be a divider from 10v to 1v/100mv (as in the Fluke 732C).

Just an idea.

EDIT

I forgot to say that in my opinion, it should be able to run from a 12v SLA UPS battery. But I am open for anything. I just want two more refs in addition to the one I have.

« Last Edit: October 06, 2019, 10:12:34 pm by AG7CK »
 

Online ArthurDent

  • Frequent Contributor
  • **
  • Posts: 962
  • Country: us
Are you suggesting that someone else should build these or are you going to step up and do it yourself?

If you check previous threads you will find that perhaps 10 units were made and sold some time ago as a one-shot deal and just the cost of materials for those was quite high. Then the aging and calibration of these references takes months on top of that, not to mention the specialized equipment you would need to fully test and characterize each reference. With the high build cost and small market I see this as a lose-lose proposition.

With apologies to anyone who has or would consider doing this-you’re crazy. This is really only practical and economical if you build one for yourself. The one I put together uses battery back-up to keep it running for a couple of days in case I lose power. It has been on power for almost a year now and seems quite stable judging from the meter that it has been connected to all this time.   
« Last Edit: October 07, 2019, 02:28:47 am by ArthurDent »
 

Online Mechatrommer

  • Super Contributor
  • ***
  • Posts: 9378
  • Country: my
  • reassessing directives...
if something can select, how cant it be intelligent? if something is intelligent, how cant it exist?
 

Offline AG7CK

  • Regular Contributor
  • *
  • Posts: 131
  • Country: th
Are you suggesting that someone else should build these or are you going to step up and do it yourself?

If you check previous threads you will find that perhaps 10 units were made and sold some time ago as a one-shot deal and just the cost of materials for those was quite high. Then the aging and calibration of these references takes months on top of that, not to mention the specialized equipment you would need to fully test and characterize each reference. With the high build cost and small market I see this as a lose-lose proposition.

With apologizes to anyone who has or would consider doing this-you’re crazy. This is really only practical and economical if you build one for yourself. The one I put together uses battery back-up to keep it running for a couple of days in case I lose power. It has been on power for almost a year now and seems quite stable judging from the meter that it has been connected to all this time.   


1. I am hoping that someone of the more experienced members will build 10 or 20 boxes because they think it is an interesting project. It might give them gratitude from the buyers, experience, prestige and payment.

2. I have read previous threads for many years. Then 10-lot you are talking about at USD 500 a piece  could today probably have sold 10 more. It however is - IMO - a bells and whistles project. I propose something KISS at USD 2-300.

3. After the boxes have been built and are working electrically, a month or so is enough for a first testing, aging and measurement. The refs will probably drift for years, so we send them back to the builder or someone else with a calibrated 3458A for new measurements against payment.

4. Quote: "With the high build cost and small market I see this as a lose-lose proposition." Well, that is your opinion. And nothing but that. The same goes for "With apologizes to anyone who has or would consider doing this-you’re crazy."

5. Quote: "This is really only practical and economical if you build one for yourself." I am not really concerned about what is practical and economical for you. I am more interested in seeing how many members (if any) who are interested in buying a ready made simple ref, and to see if any of the experienced members here are willing to make a batch of 10 against a small but honorable payment.

All it really takes is a handful of PCB's, LTZ1000s, Vishay resistors, boxes and binding posts. The rest is a question of someone with a 3458A and a temperature chamber / logger to characterize them. There are quite a few experienced builders here from Europe, Asia and USA. I would not be surprised if one or more of them got together to realize this or a similar project.

And if there are no buyers and no builders / testers, well - then I know that.

 

Offline TiN

  • Super Contributor
  • ***
  • Posts: 4112
  • Country: us
  • xDevs.com/live - 24/7 lab feed
    • xDevs.com
Without specifications targets or expected performance numbers, this talk is just one of those many previous threads of "poor man's LTZ" and such, and does not have any value.
Even if take simplest spec of them all - tempco, reference that does 0.2ppm/K or 0.02 ppm/K would have very different price, even though both devices would have LTZ1000A chip.
1K/10K hermetic resistor reference included? There are some resistors that cost more than outlined 200 USD budget  :=\
YouTube | Metrology IRC Chat room | Live-cam | Share T&M documentation? Upload! No MB limit, firmwares, photos.
 

Offline 3roomlab

  • Frequent Contributor
  • **
  • Posts: 901
  • Country: 00
  • can you tell me how to cal sesame street meter?
everything in components rise in costs every quarter, havent anyone noticed? even a stupid TEKO box could become more expensive.
the way certain business moves is to corner the entire market to themselves so they control the price (and therefore have license to gloat at others?)
of cos, the best person to gloat at others here is tin. nobody else not even dave have that license esp when playing with LTZ.
« Last Edit: October 07, 2019, 10:31:23 am by 3roomlab »
overclocked CPU and GPU are a waste of energy and time, it is highly energy + calculation inefficient vs watts. there is an entire influencer industry milking users off it, they call it "premium" but lifespans are short, oxymoronic crap , more like single use.
 

Offline branadic

  • Super Contributor
  • ***
  • Posts: 1518
  • Country: de
Without making promotion for a specific website, what you want is a ready to go board and the only source I currently know of is this: https://islandlabs.eu/ltz1000.html
In any case you shouldn't be surprised if you hear what all the effort costs you.

I'm not entirely sure what  you are after, a 732C clone with 10V/1V/100mV or a 10V reference with 1k and 10k reference as the approaches to realize are different. To create 1V  out of 10V a better way to go is with a stable resistor network with focus on ratio only, but not necessarily precision 1k/10k resistors.

-branadic-
« Last Edit: October 07, 2019, 07:30:33 pm by branadic »
Metrology Meeting 2020 on 06.06.2020
 

Offline TiN

  • Super Contributor
  • ***
  • Posts: 4112
  • Country: us
  • xDevs.com/live - 24/7 lab feed
    • xDevs.com
That source is not decent enough even to write own circuit operation description or at least give a link to my article used as a source. Shameless copycats.  :--
YouTube | Metrology IRC Chat room | Live-cam | Share T&M documentation? Upload! No MB limit, firmwares, photos.
 
The following users thanked this post: Ysjoelfir

Online NANDBlog

  • Super Contributor
  • ***
  • Posts: 4554
  • Country: nl
  • Current job: ATEX certified product design
1. I am hoping that someone of the more experienced members will build 10 or 20 boxes because they think it is an interesting project. It might give them gratitude from the buyers, experience, prestige and payment.

2. I have read previous threads for many years. Then 10-lot you are talking about at USD 500 a piece  could today probably have sold 10 more. It however is - IMO - a bells and whistles project. I propose something KISS at USD 2-300.
That is not enough money to keep the 3458A calibrated + to buy the reference.
If you want to KISS, buy an REF102, trim it to be 10V, and slap it in an Altoids box. You will get better results than an LTZ1000 with cheap components.
 
The following users thanked this post: TiN

Offline Andreas

  • Super Contributor
  • ***
  • Posts: 2473
  • Country: de
Without making promotion for a specific website, what you want is a ready to go board

There are other suppliers of complete devices even with the 1V output:
http://www.friedrich-messtechnik.de/

If it has to be cheaper then I would consider either IANs PDVS2 or PDVS2mini.
And there are also other 10V references in the wild: E.g. the heated LT1021 (DVR1) of Adrian (A.B.)
And on entry level the DMMCheck PLUS.

with best regards

Andreas

 
The following users thanked this post: TiN

Offline niner_007

  • Regular Contributor
  • *
  • Posts: 141
  • Country: us
Status post
There is nothing simple about the LTZ1000, for $200, an LM399 based reference might be more relevant.

Another way is to build a voltage standard around the 3458 DC reference, 03458-66509 (8ppm) is $700, 03458-80003 (4ppm) is around $996, and 03458-66529 (2ppm) is $2,000
« Last Edit: October 08, 2019, 04:56:30 am by niner_007 »
 

Offline TiN

  • Super Contributor
  • ***
  • Posts: 4112
  • Country: us
  • xDevs.com/live - 24/7 lab feed
    • xDevs.com
3roomlab, not here to gloat, just to set expectations reasonable. I'm not even using any LTZ references in commercial application (to make any meaningful statistics you gotta work thru hundreds of LTZ refs), so just a guy tinkering with pieces at home.

Also about readily available references, I can recommend ab-precision DVR and resistance boxes. More than enough for hobby purposes.
Other option is indeed Ian's PDVS2 series. I like the box, will need some time to finish review. It's programmable and noise/tempco are impressive for the money.

And A9 refs are not 10V ;)
YouTube | Metrology IRC Chat room | Live-cam | Share T&M documentation? Upload! No MB limit, firmwares, photos.
 
The following users thanked this post: IanJ, ap

Online SilverSolder

  • Super Contributor
  • ***
  • Posts: 1112
  • Country: 00
The makers of the DMMCheck PLUS also makes another product at the economy end of the spectrum called VREF10-002,  claimed to be 0.003% accurate (actually 0.0025% in the latest improved version).    http://www.voltagestandard.com/-.html

I purchased one of those recently,  connected it to a Fluke 8505A in 7.5 digit mode and found there wasn't really much to complain about...   the seller will calibrate to a temperature that you specify, so I chose the prevailing temperature in the "man cave". 

This result is of course pure luck and lasted only barely long enough to take a photo, but still - nice!




« Last Edit: October 08, 2019, 04:56:29 pm by SilverSolder »
 
The following users thanked this post: TiN, notfaded1

Offline Magnificent Bastard

  • Regular Contributor
  • *
  • Posts: 124
  • Country: aq
Status post
There is nothing simple about the LTZ1000, for $200, an LM399 based reference might be more relevant.

Another way is to build a voltage standard around the 3458 DC reference, 03458-66509 (8ppm) is $700, 03458-80003 (4ppm) is around $996, and 03458-66529 (2ppm) is $2,000

The LTZ1000 performance is much better when powered all the time (24/7).  Without this, some form of hysteresis recovery will be needed for maximum performance-- and even then, there will always be some unavoidable retrace.  For travelling purposes, a reference made up of a number of LM399 (say, 4 of them) averaged to reduce LF noise, followed by a PWM boost circuit might be more appropriate for shipping with power off.  The PWM could also be used to correct for drift due to temperature and barometric pressure, and (possibly) time-drift.  Because of the lack of high-performance resistors, this design could be made within the $200 price range.
 

Online Kleinstein

  • Super Contributor
  • ***
  • Posts: 6776
  • Country: de
With precision resistors for the 7 to 10 step, the accuracy of these resistors would be the major point. So it would be more about using expensive parts (and still no big mistakes in the layout / thermal design) than a clever design. For a 10 V source the LTZ1000 is only a small part of it all.

When using  2 or 4 x LM399 as reference there would be no need to use PWM as a boost circuit. One could start with 2x7 V in series and use PWM to go down as a divider, which would allow for a linear control with also the option to also reach other points like 1 V or 5 V.
 
The following users thanked this post: TiN, razvan784

Offline doktor pyta

  • Frequent Contributor
  • **
  • Posts: 286
  • Country: pl
    • RF Scientific
I must admit, that I've been working on similar project for few months. The progress is slow due to the fact that I became half time babysitter for my children :)
My draft project in the attachment. Probably I shouldn't reveal schematics at this stage, but this thread can motivate me to dig further in this direction.
Main features that are planned:

10V
1V
100mV
1mA
10mA
10k
100R
1R

Using above features with null-voltmeter could provide coverage of not all - but some very useful ranges.

The design takes advantage of NOMCA statistical arrays. In critical places they could be paralleled to improve TC/ noise.
To maintain temperature performance whole device should be in the oven (Peltier or heated only).

This is obviously inspired by Mr Pickering's CERN transfer source so it's nothing new, but doesn't meat this will be easy.
It won't be and may end it's life in a drawer.

Please comment.
« Last Edit: October 09, 2019, 02:29:16 pm by doktor pyta »
 
The following users thanked this post: 3roomlab

Online SilverSolder

  • Super Contributor
  • ***
  • Posts: 1112
  • Country: 00

[...] I became half time babysitter for my children :)  [...]


My wife once (rather angrily) explained (ruled) that "it is not called babysitting when it is your own children"!  :)
 
The following users thanked this post: RandallMcRee


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf