Ultra Precision Reference LTZ1000

[Andreas]

Interesting.
Pardon my ignorance, but how did you come up with the TC resistor values? Purely empirical?
I have a non-A LTZ1000 and thinking what values I should start from.

Hello,
I would start with R9 = infinite.
you can trim T.C. up by +40 ppb/K for each 1 Meg resistor in parallel for R9 on a LTZ1000A.

https://www.eevblog.com/forum/metrology/ultra-precision-reference-ltz1000/msg907048/#msg907048

Hello
my heater is an old ion. (Bügeleisen)
Do not take it too serious.

Just a fast solution for me. :-)

Interesting trimming bolt on the side 

best regards

Andreas

[Kleinstein]

The TC compensation used the heater voltage. This means the strength depends on the used power level and set temperature. So the +40 ppb/K is only a ball part number. The actual number depends on the heater level and thus set temperature and insulation (e.g. cap).  A lower set temperature is expected to give less heater power and thus more change in voltage to compensate for temperature changes at a given thermal setup. The tendency should be a larger resistor for a lower set temperature.
Rather close to the set temperature things can get quite nonlinear - making TC compensation difficult.

Testing / selecting R9 should be done with the final thermal setup (especially cap over LTZ) and set temperature.

[alex-sh]

Hello,
I would start with R9 = infinite.
you can trim T.C. up by +40 ppb/K for each 1 Meg resistor in parallel for R9 on a LTZ1000A.

Thank you
The below is very helpful:

For R9 = 1 Meg I get a T.C. increasement of around +0.04ppm/K
100K give around +0.4ppm/K over 20 deg C.

I feel like to need to read all 100 pages of this thread. 

[Andreas]

I feel like to need to read all 100 pages of this thread. 

Otherwise you do not know that shortening the legs of a LTZ1000A also increases the T.C.
I recommend to start with long legs and shorten as necessary.

with best regards

Andreas

[TiN]

that shortening the legs of a LTZ1000A also increases the T.C.

Hmm, you have a link to data on this? I didn't see a difference in my boards.

[Andreas]

Hello Illya,

its all within this thread for my samples:

LTZ#3
Long legs -80 ppb/K
https://www.eevblog.com/forum/metrology/ultra-precision-reference-ltz1000/msg881955/#msg881955

Half shorted legs -40 ppb/K
https://www.eevblog.com/forum/metrology/ultra-precision-reference-ltz1000/msg924793/#msg924793

2-3 mm remaining legs -22 ppb/K
https://www.eevblog.com/forum/metrology/ultra-precision-reference-ltz1000/msg928435/#msg928435


LTZ #4
Long legs -230 ppb/K
https://www.eevblog.com/forum/metrology/ultra-precision-reference-ltz1000/msg882026/#msg882026

short legs -95 ppb/K
https://www.eevblog.com/forum/metrology/ultra-precision-reference-ltz1000/msg905152/#msg905152

so the last mm seems to change more than shortening by half.

Summary table
https://www.eevblog.com/forum/metrology/ultra-precision-reference-ltz1000/msg935102/#msg935102

with best regards

Andreas

[DC1MC]

Woah, this legs thing it's really woodoo magic  !!!

 DC1MC

[TiN]

Andreas
I'm lost with your posts now, as you contradicting yourself in here  .

[Andreas]

Hello,

sorry I do not see what you mean.
I think it is a straight forward observation:
the shorter the legs the more less negative the T.C. (so T.C. increases).

with best regards

Andreas

[TiN]

Per your results, going from -80 ppb/K to -22 ppb/K and from -230 ppb/K to -95 ppm/K means TC is decreased (less impact on output voltage from temperature), not increased.

[2N3055]

Per your results, going from -80 ppb/K to -22 ppb/K and from -230 ppb/K to -95 ppm/K means TC is decreased (less impact on output voltage from temperature), not increased.

It decreased in absolute value (mathematically) , but it increased in value (as in it was less negative)....
Terminology gap.

EDIT: Spellchecker typo. Value decreased, it didn't die ... 

[TiN]

Yup. I'm going by usual convention of zero TC = no change, ignoring the sign.

[GregDunn]

Freshly arrived from across the pond, thanks to branadic for this nice little package.  Aged LTZ1000 with log of voltage vs. time, measured at 10.000002 VDC.

Obviously, I don't have a DMM good enough to check the LSDs; but this newly repaired HP3490A seems to think it's pretty close at 5.5 digits worth of precision.  This is after 4 hours of warmup; I'm now in the process of letting some other DMMs settle so I can cross-check them with it.

[AG7CK]

I like this "no voodoo" clean simple functional 10V box based on the green PCB from recent posts and a "classic" DIY-electronics style enclosure.

If Branadic or someone else has one more for sale or will produce these characterized plug-and-play reference units for other members, I would also like to buy one or be put in a queue for future purchase. Thanks.

[alex-sh]

My board is almost done. Still waiting for precision resistors to arrive...

[MiDi]

My #1 branadic / Dr. Frank



Did quick check on noise for 7V & 10V with linear bench PSU:
Noise floor 10min: ~115nVpp

7V buffered
10min: 2.3µVpp
12s best: 1.5µVpp
12s typ: 1.7µVpp
12s worst: 2.1µVpp

10V
10min: 3.2µVpp
12s best: 2.0µVpp
12s typ: 2.3µVpp
12s worst: 2.8µVpp

[Magnificent Bastard]

You might get even better noise readings if you add an air-draft excluder to the top and bottom of the LTZ1000.

[RandallMcRee]

I suspect that you have some unidentified noise source.

Both of my DrFrank boards (sorry for not posting!) showed ~0.9uV of noise. There were two different LTZs, one from Arrow, one from Mouser, A and non-A.

Certainly air drafts, etc. are a possibility. My noise floor is similar to yours.

[MiDi]

Thanks for the suggestions, seems it is worth to investigate. 
First thought was: ok, it is in spec, but bit on high side (DS: 1.2µVpp typ, 2µVpp max).

[Dr. Frank]

I'm very pleased that Mr. Pettis' resistors fit well on the PCB, that looks really nice, also in the final enclosure.

I also suggest more thermal isolation, as that noise level is a bit high, but dependent on the sample.
That measurement is also dependent on the bandwidth under observation, so identical methods and BW limits should be obeyed.

For a more practical use, you may determine StDev with your 'new' 3458A, at NPLC 100, so you may get 100..200nV rms only, that's equivalent to maybe < 1Hz BW.

And you might measure and trim its T.C. (if not already done?)
Frank

[MiDi]

Seems it was a misunderstanding what was measured (clarified in the post):
I measured 7V after buffer and the noise added by the LTC1052 is quite high (DS: 1.5µVpp typ, 7Vdirect -> 7Vbuf: ~0.7µVpp), did not realized that it adds so much noise...

Battery powered 0.1-10Hz noise:
7V buffered: ~1.7µVpp (12s, typ) - same with linear bench psu
7V direct: ~1µVpp (12s, typ)

Seems it is in the ballpark of other users 

There was no measurable difference in noise with or without air-draft excluder in 0.1-10Hz range for this unit, but maybe it affects noise below 0.1Hz 

TC compensation leads to next project: thermal chamber
Or even two for direct comparison of temperature stabilized one against DUT?

[Insatman]

Your results are generally consistent with mine.  Typical 0.1Hz to 10Hz noise is around 0.8 to 1.4uV with a noise floor of the 10K:1 amp used around 150nV.   

[GregDunn]

I've been using my reference to check calibration of my DMMs (particularly newly repaired ones) and as such I have had to move the output cables around to connect to different units.

Something I've seen happen: when removing and re-connecting cables, sometimes the reference output will drop to about 2.9VDC and stick there.  Cycling power to the reference brings it back immediately.  I've been careful to ground myself before touching the reference's case and not touch the bare leads but it happened anyway.

Is this a characteristic of the LTC1052 going into protection mode?  The datasheet only talks about overload recovery and doesn't address any sort of latch-up considerations...

[branadic]

Haven't observed something similar yet, neither on your reference nor on mine.

-branadic-

[chuckb]

From the LTC1052 data sheet.

Note 2: Connecting any terminal to voltages greater than V+, or less than
V–, may cause destructive latch-up. It is recommended that no sources
operating from external supplies be applied prior to power-up of the
LTC1052/LTC7652.

It usually takes several ma to induce SCR latchup. Is there a transorb on the output to protect against ESD causing latchup? Or it may just be a bad Op-amp.