Vishay Z202 Foil Resistors for LTZ1000 Reference Board - Where to get?

[Martin72]

Hi,

I wanted to build the reference from xDevs, I've already ordered the board and all the components for it.
Now I'm still missing the Vishay resistors from the BOM:
https://xdevs.com/article/kx-ref/



The problem is, it seems they no longer exist...
Are there successors or alternatives to these resistors, ideally matching the layout?
Because I received the board today....

Martin

[TimFox]

Element14 and Newark have some stock of Z202 resistors in UK.
I didn't check for your exact values.
(They are expensive, something like $30 USD each in small quantities.)
They are technically from a different company:  "Vishay Precision Group"  https://foilresistors.com/docs/63131/z202.pdf
See press release about spin-off  https://www.vishay.com/docs/99916/99916.pdf

[tatel]

Perhaps here?:
https://foilresistors.com/foil-resistors/how-to-order/

[Martin72]

I had already seen that, and it seems that the website was last updated in 2017.
I don't want to believe that there are no alternatives to these resistors, maybe I'll ask Vishay directly.

[chuckb]

Another way to help availability is to relax the tolerance requirements. The tempco is important, however, the tolerance and actual value of the resistors is not important. There is no "sweet spot" for resistor values.

For the 120 ohm current setting resistor I have used values between 50 and 200 ohms and the voltage reference still works fine. Use a 100 ohm 1% tolerance part.
For the 70k resistors, anything from 50k to 200k will work. The 200k will reduce the unheated tempco a little. If a 1% tolerance 60k or 100k is avaulable, use it.

You will want the keep the ratio of the 1k and 13k tight to set the chip temperature. With 0.1% ressitors the chip temperature will be within a +-3 deg C window. The chip itself has a +-5 deg C tolerance for temperature.

The precision and accuracy is needed in the voltage measurement after you build it, burn it in for awhile and calibrate it. There is no need to make things hard on yourself paying for high accuracy, hard to find resistors.

Good Luck!

[Andreas]

Hello,

those metal foil resistors are usually built on order. So usually you have 6-13 weeks delivery time.

I would either order them from Powertron (Teltow, Germany)
or Texas Components.

Besides tempco also long term drift is essential.
So I personally would use at least 3-5 ppm/K wire wound resistors or S102 resistors to do the job.
By the way: the 70K resistors are those with the lowest influence on stability.
So using resistors in metal can is a waste of money here. (one of the S102 footprints should fit)

The most critical is the 1K/13K divider and then the 120 Ohms current setting resistor. (see also measurements in the long LTZ1000 thread).

with best regards

Andreas
 

[maat]

The German distributor changed to BADER GmbH & Co. KG (https://bader.net/). At least that was the state of things in 2023 when I last got resistors from Vishay. They officially moved their sales from Powertron to Vishay Precision Foil GmbH in Heilbronn, but Bader was the one that replied to my request when I ordered 100 resistors or so.

[Alex Nikitin]

I have some surplus resistors for LTZ1000 circuit, but not exact values (though in some respect better and certainly well aged  ) Vishay S102C 87K15 0.01% for 70K position, very stable W/W Ultrohm 110 Ohm 1% for 120 Ohm position and a 3-pin Vishay resistor array 10K + 770 Ohm 0.02% as the temperature setting divider with the ratio of 13.99 (for my three samples, measured). PM me if that is of interest. 

Cheers

Alex

[TimFox]

In my past work, I often used Texas Components Corporation  https://www.texascomponents.com/store/home.asp  as an expensive source of small quantities of high-end Vishay resistors with custom values.

[Martin72]

Thanks for the many replies!
I'll summarize it for myself:
- The resistors are not freely available, but have to be specially manufactured (do private individuals even have a chance of having them manufactured?)
- It doesn't necessarily have to be 0.01% tolerance and 0.2PPM drift.
- Other values can also be used
I think I need to read up on this further to be able to classify it correctly.
As I said, I already have the board (5 of them because of the minimum quantity), the rest of the components should arrive next week.
These 5 resistors are missing, although their price is not that important to me unless they cost around €100.

[TimFox]

Texas Components made custom values of standard Vishay parts for me (in small quantities), but that was before the Z202 series.
I think the ones I bought were S102 hermetic packaged devices, but their website I cited discusses which ones they make.
Vishay supplied them with "chips" that could be laser-blasted and trimmed to get precise values.

[Martin72]

Hi,

Someone has read this thread and updated the parts list on the website...

I have now ordered these resistors, totaling €80, which is still okay.

[Kleinstein]

The LTZ1000A does not need the 400 K resistor. The 400 K resistor for the non A version is also only a crude starting point. The suitable actual value can change with the set temperature and termal condidtions, roughly in a range of some 200 K to 1 M.

[Martin72]

The LTZ1000A does not need the 400 K resistor.

I know, it´s also mentioned in the circuit description and in the LTZ1000 datasheet.

[TimFox]

R3, R4, and R5 are high-precision low-tempco devices, but R6 is a normal metal-film resistor with a somewhat better tempco than average.

[Martin72]

Yes, let's see how that turns out later.

[TimFox]

Where did you find the Vishay foil resistors?

[Martin72]

My last picture of the BOM is a screenshot of the xDevs.com website.
The blue highlighted changes are links that all lead to Digikey:


https://www.digikey.de/en/products/detail/vpg-foil-resistors/Y0007120R000T9L/2609876

https://www.digikey.de/en/products/detail/vpg-foil-resistors/Y00071K00000T9L/1888094

https://www.digikey.de/en/products/detail/vpg-foil-resistors/Y006213K0000F0L/4218402

https://www.digikey.de/en/products/detail/vishay-dale/PTF5675K000QXEB/2273701

[aronake]

One of the 70K resistors is more critical than any of the other resistors.

See this post:

https://www.eevblog.com/forum/metrology/ultra-precision-reference-ltz1000/msg4817900/#msg4817900

You may want to upprioritize that one.

[The Soulman]

I've used these PTF56 resistors through-out my LTZ1000 build, using multiple resistors for one position in the hopes of reducing tempco and improving overall stability.
The original plan was to measure each individual resistor for temp-co and make combinations for optimum performance but I didn't.
What I've used: 

70K are 3pcs 20K in series (60K) + 2pcs parallel (10K) = 70K

120 Ohm I've changed to 200 Ohm and is constructed from 4pcs 200R in series/parallel arrangement (to reduce zener-current for the benefit of long term stability at the expense of higher noise)

12k/1K didvider I've changed to 13/1K, 13K are 6pcs 2K in series (12K) + 2pcs parallel (1K) = 13K and 2pcs parallel for 1K (lowers the oven temperature, again to improve long term stability)

10V output divider is 10/4K, 10K are 2pcs 20K in parallel and 4K are 5pcs 20K in parallel.

Total costs for these resistors was about 50 euro's and were (are?) in stock at digikey.

The reference is now been powered on 24/7 for around five months, in the first few months the 10V output drifted up by a total of 10ppm, the last month it has been stable to 1ppm (for what I can tell).

Not to bad for a freshly assembled v-ref, and as of yet I have no regrets using these "cheap" resistors.
 

[The Soulman]

My last picture of the BOM is a screenshot of the xDevs.com website.
The blue highlighted changes are links that all lead to Digikey:


https://www.digikey.de/en/products/detail/vpg-foil-resistors/Y0007120R000T9L/2609876

https://www.digikey.de/en/products/detail/vpg-foil-resistors/Y00071K00000T9L/1888094

https://www.digikey.de/en/products/detail/vpg-foil-resistors/Y006213K0000F0L/4218402

https://www.digikey.de/en/products/detail/vishay-dale/PTF5675K000QXEB/2273701

The top three links are for S102 resistors not Z202, some have shown excessive hysterisis after a thermal cycle:
https://www.eevblog.com/forum/metrology/t-c-measurements-on-precision-resistors/msg624790/?topicseen#msg624790

Something to be aware off.

[Andreas]

Something to be aware off.

It´s 10 ppm over 30 deg C so 0.3 ppm/K -> much better than the specified +/-3.5 ppm/K (+/-(1+2.5)ppm/K)
Absolute error over 30 deg C is comparable to the Z201 measurements.

With best regards

Andreas

[The Soulman]

Sorry for the off-topic but am I reading this graph wrong?



Is it thermal lag between resistor and sensor or actual a shift in resistance value as the result of thermal cycling?
It was speculated it had something to do with mechanical tension between the epoxy package and the substrate of the resistor?

[TimFox]

Many Vishay foil resistors achieve low tempco by the mechanical expansion of the substrate cancelling the resistivity temperature co-efficient, so adding an external package around all of the above can certainly affect the temperature performance.

[Martin72]

The top three links are for S102 resistors not Z202

Yes, because:

https://www.eevblog.com/forum/metrology/vishay-z202-foil-resistors-for-ltz1000-reference-board-where-to-get/msg5659729/#msg5659729

Someone had changed the resistors in the parts list in the last few days.


I assume that the person knows what they are doing. I will find out soon, as all the components should arrive this week.

[Andreas]

Is it thermal lag between resistor and sensor or actual a shift in resistance value as the result of thermal cycling?
It was speculated it had something to do with mechanical tension between the epoxy package and the substrate of the resistor?

Hello,

in my opinion it is mainly mechanical lag (creeping) between the substrate the cement and the metal foil bonded to the substrate
+ influence from the epoxy package and absorbed humidity.

What I typically see with plastic molded resistors with low T.C.:
The resistance somewhere starts "in the middle".
With temperature excursion there is some hysteresis.
After half a day at initial temperature the resistance returns to the middle.
Pure thermal lagging would have much smaller time constants (in the minutes range).

Sometimes you can also see that on subsequent temperature cycles the hysteresis gets smaller.
Perhaps with some drift.
(the absorbed humidity gets out of the epoxy package).

with best regards

Andreas

[Martin72]

If you rely on parts lists...
C1, C2, C3 and C9 are Kemet 100nF/63V film capacitors and “huge”, but the footprints on the board are only for 1206, I could still fit one capacitor halfway, the other positions are simply impossible in terms of space.
There seems to be a discrepancy between the parts list, the circuit diagram and the board.
C9 is specified as X7R type in the circuit diagram and not as a film capacitor.
C1 to C3 are connected in parallel to the large tantalum capacitors.
The question is, do they have to be film capacitors at all, or would 100nF X7R in a 1206 package size suffice for all positions?
Personally, I see no electrical reason for this, but then I'm no metrologist.

[Alex Nikitin]

I would avoid using X7R dielectric as much as possible, it is leaky, unstable and microphonic. NP0 (C0G) capacitors are much better for this kind of application, preferably from TDK.

Cheers

Alex

[Martin72]

The resistors from Digikey arrived today.
It looks unspectacular for the money...
The LTZ1000ACH also arrived today, so if I can find another 100n NP0, I can finish the assembly.
I had also brought an insulating washer from work that fits the LTZ; it holds the component at a slight distance from the circuit board – there's no reason not to use that, is there?

[Andreas]

Hello,

my LTZ1000A samples had different T.C. depending on lead length. (see also LTZ1000 thread)
So I could trim the T.C. partially by shortening the legs.
(with the non-A version I was not so successful).

I personally would keep the leads (KOVAR) as long as possible to keep the soldering joints more equal in temperature.
Further you need to thermally isolate the leads (on both sides of the PCB) from air drafts to reduce noise.

with best regards

Andreas

[Korchahin]

....
Absolute error over 30 deg C is comparable to the Z201 measurements.
...

On the Mouser.de website I noticed the Z201T and VART series resistors.
Z201T
https://www.mouser.de/datasheet/2/428/VIPGS01681_1-2572870.pdf
VART
https://www.mouser.de/datasheet/2/428/zseries-3247554.pdf

Are they suitable for assembling LTZ1000 references?

[Andreas]

Hello,

The Z-series is suitable.

The VAR series is more a "gag" for the audiophiles.
They are very brittle and have less humidity protection than the z-series.

I think I measured 2 samples in the Resistor thread:
https://www.eevblog.com/forum/metrology/t-c-measurements-on-precision-resistors/msg462296/#msg462296

with best regards

Andreas

edit:
https://www.eevblog.com/forum/metrology/t-c-measurements-on-precision-resistors/msg879467/#msg879467