Ultra Precision Reference LTZ1000

[quarks]

What about equipment imported from another EU country?

Very stupid but within the EU you can buy gear without CE Mark with no problems, probably because this is not handeled as an import. 

[splin]

Coincidentally I asked for a quote for Edwin's resistors a few days ago and thought I would post the reply (with permission) as it does provide some more information and pricing which may be of interest to others here:

Hello Mr. xxxxx,

Thank you for your inquiry, your request is a bit complicated for a simple reply but I'll try to keep it to a short novel.  All of my resistors are rated to military temperature range so a lab environment would be quite soft and 0 ±3PPM/°C TCR is standard for nearly all of my resistors.  I presume you would prefer the resistor values called out by Linear Tech as these have proven to be the best values that work, only the heater value may need to be tweaked if the units are operated at elevated temperatures, the LTZ must always be above the maximum expected ambient temperature to maintain Vref.  I am currently manufacturing these sets for other customers with minor variations in heater resistor values.  ±0.1% tolerance is quite sufficient for those resistors.

Your home lab standards, 1 ohm, at the moment is a bit difficult as very low TCR alloy in larger diameters is a hit or miss even today, I just made some 10 ohm resistors for a customer with the 'best' low TCR wire I have at that big of a size, its TCR is slightly higher than I normally accept (by about 10%) but I can't always get the TCR I want.  Basically what this means is that the final TCR batch range may be shifted by 0.3PPM/°C, not much really.  While it is not impossible to make a 1 ohm resistor from this stock, it will be quite difficult as the length of wire will be quite short, 1.5" and at 0.010" diameter, not much to work with.  Unfortunately, trying to parallel multiple low ohm resistors and keep reasonably close tolerance and particularly low TCR is very tricky, a problem I was working on for the Bob Pease project that was mentioned in the forum threads.  I can try and make a 1 ohm resistor from this stock but I'm not real sure how it would turn out.  The 100 ohm is no problem nor the 10K, I have made working standards with 0 ±10PPM tolerance from 100 ohms up to 10M.  At low ohm values, lead resistance becomes very significant, small variations in the lead spacing will cause significant measurement errors.  As a point of interest, transfer standards (also known as working standards) have a TCR of 0 ±5PPM/C° at ?100?, higher than my resistors.

I haven't made any real low ohms in several years, I work mainly with the Evanohm alloys and they usually are not available (easily) in larger sizes and low TCRs unfortunately, I'm told it is a quirk of the alloy, the smaller the diameter, the lower the TCR possible, an inverse relationship.  Given the current circumstances, any resistance below about 10 ohms is considered on request, it is not so much that I don't know how to make them, it is a matter of getting the necessary materials.  At this time I generally decline making resistors below 1 ohm.

A dozen 1K resistors is easy, tolerance is no problem, quite easy, even ±0.005%.

Long term stability is specified at 0 ±5PPM/year, data indicates the actual 'drift' is < 2 PPM/year.  At rated power, rated temperature range, drift is 0 ±10PPM/year or less.

A website is under construction as are new data sheets.

Approximate pricing, standard TCR, 0 ±0.1%:

Value    1-9       11-24
1R        $8.10    $6.08
100R   $6.85    $5.14
120R   $6.85    $5.14
1K        $6.34    $4.76
10K      $7.28    $5.46
12.5K   $7.43    $5.57
70K      $8.25    $6.19

For 1K,     ±0.1%, <= 1PPM/°C, quantity of 12, each: $6.76
For 1R,     ±0.1%, <= 1PPM/°C, quantity of 1, each $10.10
For 100R, ±0.1%, <= 1PPM/°C, quantity of 1, each $8.85

Note, selection of TCR is labor and time intensive, it is likely that the batch will be close to your TCR limit anyway, unless you are sure you really need such selection, I would recommend accepting the straight production run.

My 242D resistance bridge at 1 ohm, has an accuracy of better than 5 PPM, an uncertainty of 0.5PPM.

Quote based on using my 805 style which is the same as the often noted 8e16 from Rhopoint, 0.250" D x 0.500" L, measurement spacing is standard 0.375" from resistor body (1.25" C-C).  There are no exact equivalent resistors to mine, only in general terms and physical dimensions, electrically no one else can claim the same specifications or performance.

If the envelope meets requirements, the International first class package service rate would be approximately $10, the International priority package rate would be approximately $25.  The paperwork requires an address and specific details to get a more exact rate.

Best regards,

Edwin G. Pettis

Pettis Engineering / Ultrohm Plus
125 Vista Grande Dr.
Grand Junction, CO  81507-1427
1-970-242-4929
pettiseng@q.com

Hi,

I'm interested in purchasing some of your precision resistors. I understand that firm pricing probably depends on the precise order so approximate pricing will be Ok. for now.

These would all be for lab use so TCs below 10C or over 35C may be greater.

I'd like between 1 and 4 sets of resistors for an LTZ1000 reference; standard 0 to +3ppm parts should be OK.

I am also interested in some resistors for use as home lab 'standards'. The number obviously depends on pricing but at least 1R, 100R and 10K would be required. Are you still able to supply selected, < 1ppm, parts? If 1R is not available in < 3ppm TC, could you quote for the minimum number of paralleled resistors that would achieve 1R, < 3ppm and < 1ppm? 1% tolerance would be acceptable but I would like to know the measured value. What would be the measurement uncertainty please?

Are you able to supply low value, < 10mohm 10W resistors? If so what would the be the price and specification?

Could you supply a set of 12, approximately 1k resistors matched to 1% or better with TCs matched to < 1ppm? What if they were matched to .1%?

How much would shipping to the UK cost for 20 resistors?

Do you have a web site?

Do you have any data available on long term, room temperature stability?

Thanks,
    xxxxxxx

[HighVoltage]

What about equipment imported from another EU country?

Very stupid but within the EU you can buy gear without CE Mark with no problems, probably because this is not handeled as an import.

Correct, anything from an EU country is not handled by customs. Sometimes I have items shipped to a friend in the  Netherlands, where the customs is much more relaxed than in Germany. But I found a different way for gear without CE marking. If you have a customs import number in Germany (EORI) and you make a statement that this gear is for research and development (Forschung und Entwicklung) then the German customs will release it to you on your own risk and you have to sign a release form. But in any case it is a hassle.

Another great way to bypass German customs has worked a few times for me in the last few month. Just tell the seller to write on the custom forms: "Adult Plastic Toys" and suddenly it is delivered to your house without anyone from customs touching it.
 

[Edwin G. Pettis]

High Voltage,

Adult plastic toys.....that is hilarious!  My resistors would fit into just the right size box for that trick!

[macfly]

Hi voltnuts,

problems with the customs at import to the EU ?

That's really unbelievable for me. During the past 14 years I bought nearly all my
test gear in the U.S. and never had any problem with CE marking.
My last buy was october last year.

Perhaps I am a lucky guy ....

Regards,

macfly

[bingo600]

I haven't had any probs. with CE markings either here in DK.

Only customs , and i don't even think "Adult toys" would slip by , as DK-Mail can claim the 20€ in fee for every package that's above 10€.

I even had a package once bought for 12$ from an ebay seller , but DK-Mail (Customs) claimed i had to prove the value by sending them a print of the paypal transaction.

They really want the 20€ in fee.

/Bingo

[janaf]

The Edwin P resistors, I'm in for three sets if there is a group order.
3x120R
3x1K
3x12K5
6x70K
If it gets too complicated, I can order directly from him.

Jan

[Andreas]

Hello Ken,

do you have a higher resolution photo of the chip?
From the first view it looks rather different from the LTZ chip.

With best regards

Andreas

[babysitter]

<guessing>
By the looks of it I believe that it is a stacked, combined and averaged bunch of sources until its figured out better.
Even probing the single contributors and disconnecting the bad apples is possible IMHO, or the single cells might be prepared to have different tempcos and then select them so they cancel.
</guessing>
Or does somebody have a better interpretation of this l.q. photo?

[Andreas]

<guessing>
By the looks of it I believe that it is a stacked, combined and averaged bunch of sources until its figured out better.
Even probing the single contributors and disconnecting the bad apples is possible IMHO, or the single cells might be prepared to have different tempcos and then select them so they cancel.
</guessing>
Or does somebody have a better interpretation of this l.q. photo?

For me it looks like having 16 references all in parallel on LTFLU.

Similar to the LTZ1000 where the transistors Q1 + Q2 are splitted
into 4 elements each staggered in 45 degree arangement.

The tempco on LTFLU will most probably be adjusted by adjustment of the zener current.
Usually these parts are sold as a complete RefAmp-Set (the reference + the proper adjusted resistors).

With best regards

Andreas

[richiem]

For those interested who haven't visited the D-105 voltage standard thread, the one I ordered arrived, and I posted some brief results of testing there; look for my post near the bottom of page 15:

https://www.eevblog.com/forum/testgear/calibratory-d-105-dc-precision-voltage-reference-standard/210/

[TiN]

In case anyone miss, I have two references for sale.

[Dr. Frank]

Just in case anyone is interested, here is a photo of a Linear-Tech LTFLU-1ACH with the top of the can cut off:

Hi Ken,

I've compared this chip with the known LTZ1000 chip..
This one, of the presumed LTFLU, looks very strange, and I especially could not find any resemblance to the LTZ1000 chip, although the buried zener should be identical, acc. to Bob Dobkin.

Therefore, are you sure, that this is not a photo of a fake chip?

Frank

[babysitter]

I am in the process of setting up a dedicated webspace used for voltage and resistance standards. I am not yet behind the idea stage, but maybe I am making progress with the community.

http://wildvolttaming.schaffenburg.de is the temporary location.

I think about using a Wiki-Style to have a "public device history file", with a unique ID assigned to every device.
The idea is to stick a QR code plaque to the Device so the hosts can see the history of their guest devices and add their own results to it.
The home labs can have a Wiki page there, too. Maybe later online comparison tools will be available.
This will support cross-comparison of hobbyists standards.

Does this idea get some love here?

[quarks]

I've heard stories [or fables] that *some* very old Fluke 732A's hold the volt to +/-0.1ppm per year.   I don't know if this is true, but if I owned one of those, you would have to use dynamite to get it out of my hands...

So, some projects that have been "stewing" in my lab are:

• Poor-man's primary voltage transfer standard [based on LTZ1000(A)]
• Poor-man's voltage transfer device
• Poor-man's multi-junction thermal converter for AC/DC transfers
• Poor-man's resistance standards [1R, 10R, 100R, 1K, 10K, 100K, 1M, 10M, 100M, 1G]
• Poor-man's resistance synthesizer [that uses the above resistors to get any value]
• Poor-man's direct current comparator bridge for low resistance measurement
• Poor-man's electrometer for high resistance measurement
• Poor-man's calibrator for DCV, ACV, DCI, ACI
• Poor-man's automatic LCR bridge
• (and a few others that are still in the idea stage)

There's enough there to keep me busy for a very lonnnnnnnng time...

Almost anything on your list is/was also on my list and probably on many others list too.
But I wonder what you mean by "Poor-man's ...".
Can you share what accuracy and stability you like to achive or find good enough for your listed projects?
If some is DIY, what are your candidates for the parts?

To make a start, for DIY resistance standard I started to select and test parts.
The top canditates (I already have) are from Isabellenhuette, Burster, VPG and Caddock. Because I can measure the values with quite good accuracy myself, I did not look for much better than 0.01% types.  I mainly care for low TCR and long time stability.

So far I have not really managed a good 1Ohm solution and nothing adequate for 100 M and 1 GOhm.

Maybe we should open a new thread for this.

[janaf]

Here are some results from measurements I made on the heaters of two LTZ1000A

[quarks]

Anyway, yes, I agree, these are probably better discussed in another thread.  I planned on launching some threads with the "[Metrology] " prefix, so that they would be easier to search for.

Very good, please do so

[janaf]

Here is a diagram for setting the temperature of the LTZ1000A. The values are calculated from datasheet, not measured. I hope it's useful to someone.

[richiem]

I have an HP 3458A ref board, REV A 03458-66509. Can anyone confirm whether this board rev has an LTZ1000 or 1000A? The HP part number for the LTZ is 1826-1860 -- I have a good HP X-ref, but this part # isn't in it. I've put a 12.1K metal foil resistor in the place of the 15k Vishay for temp control. This value would be OK for the 1000, but possibly marginal for the 1000A. Solid info will be very welcome!

[Dr. Frank]

Hi Dick,
As far as commonly known, HP used the A version only, due to the nominal 95•C.
12.1k sets  the oven to about 45•C but the interior temp already will be 35.. 40•C at least, so the difference is much too small, as the A version requires 10•C overhead plus some additional regulation margin, at least.
The self heating of the 3458A adds  13•C to room temperature ,plus 5 •C more for a dirty filter, plus 10•C if placed in a rack.
Therefore, I always recommend to set the temp to 65•C , ie 13k, or 100k in parallel to the 15k, if the lab is not warmer than 30•C

The reference ages only, when it is powered.
Therefore, drifts < 1ppm/year can be expected on 65•C,  if the 3458A is not running continuosly.

If you use the ref board outside the 3458A, then 12.5k is the correct value, for about 55•C.
45•C is a little bit critical for the A version, otherwise you have to make sure, that the ambient temperature will not exceed 30•C. And there should be a thermal shielding around the TO 99, which should allow heat exchange, ie not too good an isolating material.





Frank

[ltz2000]

I've heard stories [or fables] that *some* very old Fluke 732A's hold the volt to +/-0.1ppm per year. I don't know if this is true, but if I owned one of those, you would have to use dynamite to get it out of my hands...

True. Sometimes it happens that the drifts of the zener and the amplifier (resistors) are approximately the same magnitude but opposite direction resulting a very low net drift. However that is not "real" stability because the individual drifts can be and usually are still quite high. Would be risky to assume the net drift to stay constant forever because it is very sensitive to changes in individual drifts.

[quarks]

I have an HP 3458A ref board, REV A 03458-66509. Can anyone confirm whether this board rev has an LTZ1000 or 1000A? The HP part number for the LTZ is 1826-1860...

that is an interesting question. I have several boards with the same revision as yours and I read original LTZ1000ACH and also 1826-1860 on older ones (which is probably only printed on for HP). In the schematic (see att.) it says clearly LTZ1000 and there also is R417 200k which should be 400k, according to LT reference design. But even more interesting is, that it should not be populated when LTZ1000A is used, but the resistor is on all boards I have seen so far.

https://www.eevblog.com/forum/projects/ultra-precision-reference-ltz1000/msg440601/#msg440601

I wonder if there is a good explanation.  Would be great if anyone could share knowledge.

[branadic]

I found 0.3-0.8ppm/°C Precision zener-based references today. Maybe of interest for someone?

[Mickle T.]

I have a similar 10 V chip - APEX VRE310J. It is the worst reference I know Low-frequency noise is ~60 uV p-p (0.01-10 Hz).

[Andreas]

Thanks Mickle,

the first who can confirm my results.


Although this is off topic in the LTZ1000 tread:

I thought I was the only one who got 2 mondays devices (VRE3050AS) from APEX.
They are noisy as hell too, have large temperature hysteresis, bad PSRR.
The only good thing is T.C. even when one device was maginal at 25 deg C.
For the price I can get easily a hand full of other good references and select out the best.

By the way Thaler was acuired from APEX so the company name has changed.
The components are still called VRExxxx (available at DigiKey).

With best regards

Andreas