Author Topic: Ultra Precision Reference LTZ1000  (Read 1330114 times)

0 Members and 2 Guests are viewing this topic.

Offline Dr. Frank

  • Super Contributor
  • ***
  • Posts: 2377
  • Country: de
Re: Ultra Precision Reference LTZ1000
« Reply #2350 on: October 09, 2018, 07:59:05 pm »
Hello,


the current consumption of my LTC2057 buffered devices is listed here:
(20-24 mA including a LT1763 voltage regulator which should be well below 1 mA)

https://www.eevblog.com/forum/metrology/ultra-precision-reference-ltz1000/msg874560/#msg874560

For ageing: I think it depends more on the individual device or datecode than on the "A" or non A version.
But since we do not have enough samples and a JJA at hand we will never know.

with best regards

Andreas

Andreas,
thank you very much for your hints, which support my rough estimations.
Therefore, the LTZ1000A really has no big advantage regarding power consumption over the non-A device.

Regarding ageing, I think it's possible to make at least a well educated guess, even w/o a JJA.

If you have a history, and a big enough amount of LTZ1000, you can identify these minute deviations.

Frank
 

Offline Dr. Frank

  • Super Contributor
  • ***
  • Posts: 2377
  • Country: de
Re: Ultra Precision Reference LTZ1000
« Reply #2351 on: October 09, 2018, 08:09:04 pm »
Whatever QA testing means  :-//
Usually: you take from each lot a number of samples and test them.

There's different philosophies, or requirements on testing, which I have seen at different manufacturers at their production lines...

If you have commercial grade devices, you would only take some random samples and test them to the limits (e.g. over temperature), and rely on predictive statistics.
If you have high grade devices, like military, or critical automotive parts (e.g. non-volatile memory), you would really test them 100% (each device) at the limit temperatures.

I don't know, how serious LT takes the expensive LTZ1000, but I could imagine that because of the price tag, they really test it to 100%.

Maybe we should contact them, which tests they really perform on the devices.
And maybe they are not really aware of the pitfall of such handling.
 

Offline Dr. Frank

  • Super Contributor
  • ***
  • Posts: 2377
  • Country: de
Re: Ultra Precision Reference LTZ1000
« Reply #2352 on: October 09, 2018, 08:45:39 pm »

...as measured and proven by Pickering, hp in their drift - AN18A,

Would you please link the publications/pdfs you are refering to? Thanks.  :-+

-branadic-

I refer to the patent of Pickering, and his explanations on the 7001 reference.
AN18a should be well known, so I don't put a link here. You just have to extract from that AN18a, that the 3458A may basically suffer from hysteresis, if left unpowered (which is not included in the specifications).

branadic, when you have measured my LTZ #3 last January, it was quite cold, below 0°C, during the transport. You see the graph before, and after this ring comparison, and over the following year after initialization, referenced to my home 3458A and 5442A. Latter seem to be constant within < 1ppm.
This #3 reference made a hysteretic jump of about +1ppm, due to the low temperatures, which is obviously a permanent, and constant one.
You also see two baseline points for the 3458A, the first is from acbern, the 2nd is taken recently at the Hannover Maker Faire, using this same #3 compared to the PTB 3458A.
The ring comparison before/after the transport to the Maker Faire shows no change, in contrast.

This event, and several others, support my own experience with hysteresis behaviour of the LTZ1000 device, at sub-ppm level, even w/o JJA.

Frank
« Last Edit: October 09, 2018, 08:56:05 pm by Dr. Frank »
 
The following users thanked this post: TiN, 3roomlab, msliva, alm, MiDi

Offline martinr33

  • Frequent Contributor
  • **
  • Posts: 363
  • Country: us
Re: Ultra Precision Reference LTZ1000
« Reply #2353 on: October 10, 2018, 12:05:42 am »
So... ife we could add a little bit of logic to the 3458a to deliver that ramping waveform to the LTZ1000A on startup, it would improve the drift on the 3458a?

I think that the circuit could be noise-free - it would have a "suicide mode" that cut all power once the warmup was completed. There would certainly be some digital noise in the window. Problem is, it is 4 hours long which seems a bit much for day to day operation. but is the spec for true precision.

Or maybe the 3458a devices were selected for low hysteresis.
« Last Edit: October 10, 2018, 12:15:28 am by martinr33 »
 

Offline branadic

  • Super Contributor
  • ***
  • Posts: 2377
  • Country: de
  • Sounds like noise
Re: Ultra Precision Reference LTZ1000
« Reply #2354 on: October 12, 2018, 09:33:41 am »
UPDATE: Got the info today, that the boards have been shipped on Wednesday, October 10, 2018. What should I say, they today already received at my location. Attached is a picture on how they looked like, before they where shipped.

Time to send out some mails this weekend.

-branadic-

EDIT: Just for sanity check. Everyone should have received an email with all the further details. Hopefully we will see some pictures of assembled LTZ-boards soon.
« Last Edit: October 13, 2018, 01:30:17 pm by branadic »
Computers exist to solve problems that we wouldn't have without them. AI exists to answer questions, we wouldn't ask without it.
 
The following users thanked this post: bck, Macbeth, MiDi, chekhov

Online Vgkid

  • Super Contributor
  • ***
  • Posts: 2710
  • Country: us
Re: Ultra Precision Reference LTZ1000
« Reply #2355 on: October 13, 2018, 06:53:29 pm »
Received email , monies will be sent.
If you own any North Hills Electronics gear, message me. L&N Fan
 

Offline branadic

  • Super Contributor
  • ***
  • Posts: 2377
  • Country: de
  • Sounds like noise
Re: Ultra Precision Reference LTZ1000
« Reply #2356 on: October 13, 2018, 08:44:57 pm »
All extra boards are sold.  :-+

-branadic-
Computers exist to solve problems that we wouldn't have without them. AI exists to answer questions, we wouldn't ask without it.
 

Offline Andreas

  • Super Contributor
  • ***
  • Posts: 3207
  • Country: de
Re: Ultra Precision Reference LTZ1000
« Reply #2357 on: October 14, 2018, 10:01:01 am »
Hello,

when I see the effort I would rather make something with a small processor and a DAC (LT1257).
similar to the cirquit here:
https://www.eevblog.com/forum/metrology/ultra-precision-reference-ltz1000/msg844021/#msg844021

In any case I would use a relay to disconnect the hysteresis cirquit after the full time.

with best regards

Andreas
 

Offline MiDi

  • Frequent Contributor
  • **
  • Posts: 599
  • Country: ua
Re: Ultra Precision Reference LTZ1000
« Reply #2358 on: October 16, 2018, 11:41:43 pm »
First impression of branadics (Dr. Franks) design - could not wait and bodged test-pcb together - thank you very much!

« Last Edit: October 17, 2018, 10:54:53 pm by MiDi »
 
The following users thanked this post: branadic, bck, flittle

Offline branadic

  • Super Contributor
  • ***
  • Posts: 2377
  • Country: de
  • Sounds like noise
Re: Ultra Precision Reference LTZ1000
« Reply #2359 on: October 17, 2018, 06:52:35 am »
Wonderful, hope it is alive  :-+

BTW: The second bunch with boards went to the post office this morning and should receive their owners within the next days.
Thanks to all those of you sending a beer or two extra.

-branadic-
« Last Edit: October 17, 2018, 08:22:34 am by branadic »
Computers exist to solve problems that we wouldn't have without them. AI exists to answer questions, we wouldn't ask without it.
 

Offline Andreas

  • Super Contributor
  • ***
  • Posts: 3207
  • Country: de
Re: Ultra Precision Reference LTZ1000
« Reply #2360 on: October 17, 2018, 07:36:08 am »
First impression of branadics (Dr. Franks) design - could not wait and bodged test-pcb together - thank you very much!

But why do you use IC-sockets in a precision cirquit?
I had horrible drifts (several ppms) in my LM399#3 until I removed the sockets.

with best regards

Andreas
 

Offline borghese

  • Regular Contributor
  • *
  • Posts: 70
  • Country: si
Re: Ultra Precision Reference LTZ1000
« Reply #2361 on: October 17, 2018, 09:01:53 am »
Boards arrived
High resolution image https://islandlabs.eu/_media/Black%20Boards.jpg
« Last Edit: October 17, 2018, 09:11:27 am by borghese »
Cheers
Borghese
 

Offline MiDi

  • Frequent Contributor
  • **
  • Posts: 599
  • Country: ua
Re: Ultra Precision Reference LTZ1000
« Reply #2362 on: October 17, 2018, 10:02:21 am »
Short test was successful  :-+

But my heart went into my boots after power it on  :o

The output of +REFbuf was sitting at ~6.9V for a felt eternity until it suddenly went to stable ~7.10V  :phew:

Is this expected behaviour?

I triple checked everything before power it on and went step by step through initial operation:

1) supply check (only components for this installed)
2) chopper check (simple zener & for safety 330Ohm heater resistor installed, missing LT1013 & LTZ1000)
3) final test with all components

But why do you use IC-sockets in a precision cirquit?
I had horrible drifts (several ppms) in my LM399#3 until I removed the sockets.

I knew someone would ask that  8)
This pcb is just to practice and for testing purposes.
In the productive units there will be no sockets and no jellybean (0815) resistors wacked together  ;)
 

Offline dkozel

  • Regular Contributor
  • *
  • Posts: 116
  • Country: gb
Re: Ultra Precision Reference LTZ1000
« Reply #2363 on: October 17, 2018, 10:25:53 am »
Has anyone created a bill of materials from the schematic? If not I'll do that tonight and post it up here.

Many thanks to branadic and MIDI!
 
The following users thanked this post: Macbeth

Offline MiDi

  • Frequent Contributor
  • **
  • Posts: 599
  • Country: ua
Re: Ultra Precision Reference LTZ1000
« Reply #2364 on: October 17, 2018, 10:42:39 am »
Got BOM with schematics & layout from branadic, ask him.
 

Offline Andreas

  • Super Contributor
  • ***
  • Posts: 3207
  • Country: de
Re: Ultra Precision Reference LTZ1000
« Reply #2365 on: October 17, 2018, 04:09:15 pm »
The output of +REFbuf was sitting at ~6.9V for a felt eternity until it suddenly went to stable ~7.10V  :phew:

Is this expected behaviour?


No,

except if you count eternity in micro-seconds.
Here a cold-start of my LTZ#5

blue: 14V power supply after regulator (switching on is done via "on"-Pin of the LT1763)
yellow: zener voltage (unbuffered) but the buffered output follows within ms (2nd picture)
green: heater voltage (stable after ~400 ms)
red: zener current (voltage on 120 ohms resistor).

When I zoom in for the zener voltage the output is stable (within some mV) after 200 ms.

with best regards

Andreas




 

Offline Dr. Frank

  • Super Contributor
  • ***
  • Posts: 2377
  • Country: de
Re: Ultra Precision Reference LTZ1000
« Reply #2366 on: October 17, 2018, 04:39:35 pm »


The output of +REFbuf was sitting at ~6.9V for a felt eternity until it suddenly went to stable ~7.10V  :phew:

Is this expected behaviour?



That depends. And you have to be precise.
The A version is instantly on , < 1 sec for the oven, seems that branadic uses one.
The non A takes between 10 to 30sec  to stabilize. Last year, I have published several warm up graphs.
So which version do you use? That's not visible on your picture.

And there's a 50ppm/K T.C., so it also depends on the oven set point.

For that you have to check the warm up directly at the reference, then at the chopper output.
Latter may also take some time to operate correctly, but also less than 1s.

If you determine the proper time scales where things happen , 'felt eternity' is useless, it's much easier to track potential failure.
 
The following users thanked this post: Andreas

Offline Andreas

  • Super Contributor
  • ***
  • Posts: 3207
  • Country: de
Re: Ultra Precision Reference LTZ1000
« Reply #2367 on: October 17, 2018, 06:24:57 pm »
Forgot to mention: mine is a LTZ1000A version

with best regards

Andreas
 

Offline branadic

  • Super Contributor
  • ***
  • Posts: 2377
  • Country: de
  • Sounds like noise
Re: Ultra Precision Reference LTZ1000
« Reply #2368 on: October 17, 2018, 06:52:44 pm »
Quote
The A version is instantly on , < 1 sec for the oven, seems that branadic uses one.

You mean Andreas ;)

Quote
So which version do you use? That's not visible on your picture.

He's using LTZ1000CH, so "longer" warm-up time to live with.

-branadic-
Computers exist to solve problems that we wouldn't have without them. AI exists to answer questions, we wouldn't ask without it.
 

Offline Dr. Frank

  • Super Contributor
  • ***
  • Posts: 2377
  • Country: de
Re: Ultra Precision Reference LTZ1000
« Reply #2369 on: October 17, 2018, 08:03:35 pm »
branadic,
sorry, that I confused Andreas and you.

Currently being on an engineering meeting in Dresden, that seems to take its toll. Skål.

Frank
« Last Edit: October 17, 2018, 08:44:51 pm by Dr. Frank »
 

Offline MiDi

  • Frequent Contributor
  • **
  • Posts: 599
  • Country: ua
Re: Ultra Precision Reference LTZ1000
« Reply #2370 on: October 17, 2018, 11:36:01 pm »
Apologies for my unprecision on this ultra precision thread  ::)

Updated picture of first impression to better resolution.

It is a non A with 12k, so ~45° heater set point.

Attached are the shots for startup - Houston we have a problem!

Powersupply goes to 12V and then drops to ~10V before going to 12V again when heater reduced power.
At same time the REFbuf (dark blue) of chopper steps to the correct value - naturally as 10V supply do not allow for 10V output and REFBuf is to low.
 :wtf:

Checked all traces and connections on PS lines, all well below an Ohm.
Measured input voltage at jack on power up -> constant @15V.
Load is ok, ~50mA initial to ~20mA steady state.

LT1763 damaged?
Oscillation? Possible, but has to be quite low in amplitude, there is nothing in the plots.
Is missing local decoupling for chopper intended?
« Last Edit: October 17, 2018, 11:57:06 pm by MiDi »
 

Offline Dr. Frank

  • Super Contributor
  • ***
  • Posts: 2377
  • Country: de
Re: Ultra Precision Reference LTZ1000
« Reply #2371 on: October 18, 2018, 08:45:26 am »
I can only guess about the LT1763 circuit.. but it seems to be not OK.

Did you use the variable version (not any of the fixed ones)?

Does the device heat up extensively on power on? => 50mA at warm up is perfectly fine for the LTZ1000 circuit, settles to ~25mA.
P.D. is 150mW, maybe that's too much???

As it's an LDO, stability may be an issue, test for oscillation and check stabilization caps.

Frank
 

Offline MiDi

  • Frequent Contributor
  • **
  • Posts: 599
  • Country: ua
Re: Ultra Precision Reference LTZ1000
« Reply #2372 on: October 18, 2018, 12:00:03 pm »
Ordered LT1763IS8PBF from voelkner aka conrad, should be adj. version - see attachments.

Will do further investigation likely this evening...
 

Offline Macbeth

  • Super Contributor
  • ***
  • Posts: 2571
  • Country: gb
Re: Ultra Precision Reference LTZ1000
« Reply #2373 on: October 18, 2018, 08:01:58 pm »
Has anyone created a bill of materials from the schematic? If not I'll do that tonight and post it up here.

Many thanks to branadic and MIDI!

It would be great if you have curated a decent one.

Of the important ICs besides the LTZ1000/A my own BOM of actual manufacturer part numbers with the correct foot prints is:

LT1763CS8#PBF
LT1013ACN8#PBF
LTC1052CN8#PBF     (certainly not MAX420)

Correct me if I'm wrong?

I've had trouble with the 70K MF resistors not fitting perfectly. It looks like Cellularmitosis version of the board fixed this with the extra pads. Bent leads it is. Mustn't grumble.

A BOM of preferred caps would be good too if you've done it  :-+
 

Offline Macbeth

  • Super Contributor
  • ***
  • Posts: 2571
  • Country: gb
Re: Ultra Precision Reference LTZ1000
« Reply #2374 on: October 18, 2018, 08:52:06 pm »
Thank you branadic  (and Dr.Frank) :-+

Got my boards today, just a test fit for the most important parts...
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf