Author Topic: [FT] xDevs.com KX LTZ1000A fairy-tale, or the story of little jumper.  (Read 68945 times)

0 Members and 1 Guest are viewing this topic.

Online Kleinstein

  • Super Contributor
  • ***
  • Posts: 6226
  • Country: de
Re: [FT] LTZ1000A fairy-tale, or the story of little jumper.
« Reply #125 on: November 17, 2016, 08:46:08 pm »
Using chopper OPs for the LTZ1000 circuit can get tricky. They produce quite some RF spikes can have higher noise in kHz range, where the transistor in the LTZ is not helping anymore. There is no real need to replace the LT1013 or similar OP. The LT1013 is not that expensive. But at least dual OPs use the same pinout.

For the extra butter, the AZ OP is suitable, as here the error from the OP fully enters.
One might think about having some RC filtering here - to reduce higher frequency noise from the reference and keep spikes from the AZ OP away from the reference.

With a relatively high current, there was a good reason to have a transistor / JFET to drive the reference current. If one really wants to save on this, one could still use a bridge or diode like shown, but I would really have the space for the transistor on the board.

 

Online Andreas

  • Super Contributor
  • ***
  • Posts: 2392
  • Country: de
Re: [FT] LTZ1000A fairy-tale, or the story of little jumper.
« Reply #126 on: November 17, 2016, 09:16:21 pm »
Hello,

Using chopper OPs for the LTZ1000 circuit can get tricky. They produce quite some RF spikes can have higher noise in kHz range, where the transistor in the LTZ is not helping anymore. There is no real need to replace the LT1013 or similar OP. The LT1013 is not that expensive. But at least dual OPs use the same pinout.

Do you have founded measurement values on this?
I have not measured any differerence between LT1013 and LTC2057 on my latest 4 references.

For the extra butter, the AZ OP is suitable, as here the error from the OP fully enters.
One might think about having some RC filtering here - to reduce higher frequency noise from the reference and keep spikes from the AZ OP away from the reference.
+1
I fully agree I measured large differences between filtered and unfiltered buffer regarding EMI sensitivity.

With a relatively high current, there was a good reason to have a transistor / JFET to drive the reference current. If one really wants to save on this, one could still use a bridge or diode like shown, but I would really have the space for the transistor on the board.
The main intention of the FET-Buffer was LT1013-related together with reduced power supply voltage.
keep heating away from LT1013 with relative high offset drift
+ extend output voltage range of the LT1013 (level shift) for reduced battery voltage.

So with a rail to rail chopper I see no reason to use a (now obsolete) FET.
The initial design is now already 6 years old.

With best regards

Andreas
 

Offline mimmus78

  • Supporter
  • ****
  • Posts: 667
  • Country: it
Re: [FT] LTZ1000A fairy-tale, or the story of little jumper.
« Reply #127 on: November 18, 2016, 12:30:43 am »
are you shure that C10 is on the "right" side of the resistors?

Well yes, better to move it before. The idea is to have a low pass filter to reduce the noise, and maybe also EMI. Don't know how and if it will works.

I added also some load on the buffered output ... what I have to check on datasheet to confirm 47K is sufficient (maybe Peak to Peak Current Noise)?

I added also limit to output swing of the protection circuit ... this will protect from almost any negative or something like > 11V + couple diode drop.
Do not expect any CAT rating here, it may be not sufficient to save the last opamp but at lest rest of reference should be safe if some low voltage is applied to the binding posts by error.
Anyone have a good example of 10V reference protection? Don't know how much this will upset the result.
 

Online Kleinstein

  • Super Contributor
  • ***
  • Posts: 6226
  • Country: de
Re: [FT] LTZ1000A fairy-tale, or the story of little jumper.
« Reply #128 on: November 18, 2016, 08:50:32 am »
The "filter" around R15,R16 C10 is there to make the buffer amplifier tolerant to capacitive load.

I don't really understand the protection circuit shown at the output -  to me this makes little sense. If needed at all I would more expect kind of a beefy 8 V zener diode or equivalent circuit.

One point where an extra protection might make some sense would be the unbuffered output. This is known to be sensitive to a short or overload (e.g. large uncharged capacitor), as this would cause the heater to activate all the way. A protection could check for a drop of the output by more than an few millivolts and in this case turn down the heater. It could take as little as a tl431 and a few resistors and a diode.
 

Online Andreas

  • Super Contributor
  • ***
  • Posts: 2392
  • Country: de
Re: [FT] LTZ1000A fairy-tale, or the story of little jumper.
« Reply #129 on: November 18, 2016, 10:48:11 am »
Hello,

that´s what I have planned for my Rev "C" board.
C23 EMI filtering.
R25 C24 (wideband) noise filtering and limiting back current to LTZ1000 zener via the OP-Amp input protection diodes in case of output shortening.

(I am still looking for good ideas for a temperature stable 7 to 10V transfer)

with best regards

Andreas
 

Offline mimmus78

  • Supporter
  • ****
  • Posts: 667
  • Country: it
Re: [FT] LTZ1000A fairy-tale, or the story of little jumper.
« Reply #130 on: November 18, 2016, 11:57:01 am »
UPDATED 2016 Nov 25 - 22:30 Italy time

We still have: 8 PCBs with 1mm guard missing

Waiting for payment:
 
 - msliva               3pcs
-  bingo60            2pcs

=========================================================

Shipped 25/11/2016:

 - z01z                  3pcs (new pcb)
 - SvanGool          2pcs (new pcb)
 - fanOfeeDIY       2pcs (new pcb)
 - villas                  2pcs (new pcb)

=========================================================

Shipped 18/11/2016:

 - Nuno_pt             3pcs      -     PM SENT - Paid - Priority
 - SvanGool           2pcs      -     PM SENT - Paid - Priority
 - kutte                   2pcs      -     PM SENT - Paid - Priority
 - dr.diesel           5pcs       -       PM SENT - Paid - Registered

=========================================================

Shipped 11/11/2016:

 - VK5RC              3pcs       -      PM SENT - Paid - Registered
 - t2kv                    2pcs       -      PM SENT - Paid - Registered
 - klaus11              2pcs       -      PM SENT - Paid - Priority
 - HighVoltage       2pcs       -      PM SENT - Paid  - Registered - overbooking list
- pelule               3pcs       -        PM SENT - Germany zone 1 - priority mail 3.50 EUR - registered 7.90 EUR
 - Mimmus78         3pcs       -      that's me

=========================================================
« Last Edit: November 25, 2016, 09:50:44 pm by mimmus78 »
 

Offline mimmus78

  • Supporter
  • ****
  • Posts: 667
  • Country: it
Re: [FT] LTZ1000A fairy-tale, or the story of little jumper.
« Reply #131 on: November 18, 2016, 12:03:55 pm »
I don't really understand the protection circuit shown at the output -  to me this makes little sense. If needed at all I would more expect kind of a beefy 8 V zener diode or equivalent circuit.

Yes I rushed a little bit the design, and this is ok for just ~ 0.6V. Considering sense is attached to the binding posts maybe just some zeners is enough.

One point where an extra protection might make some sense would be the unbuffered output. This is known to be sensitive to a short or overload (e.g. large uncharged capacitor), as this would cause the heater to activate all the way. A protection could check for a drop of the output by more than an few millivolts and in this case turn down the heater. It could take as little as a tl431 and a few resistors and a diode.

I don't plan to expose this network outside the reference ... maybe this ca be still useful if some op-amp will die. I will check it.

How to calculate what minimum load I need to present to the op-amp?


 

Offline villas

  • Contributor
  • Posts: 6
  • Country: br
Re: [FT] LTZ1000A fairy-tale, or the story of little jumper.
« Reply #132 on: November 18, 2016, 05:14:42 pm »
I would like to build one reference, but I probably will not be able to find the resistors here and I don't like the idea getting these from eBay.
Do you guys plan to group buy the low TC resistors from VPG or some distributor?
 

Offline mimmus78

  • Supporter
  • ****
  • Posts: 667
  • Country: it
Re: [FT] LTZ1000A fairy-tale, or the story of little jumper.
« Reply #133 on: November 18, 2016, 07:43:41 pm »
Flintstone: this R is used for temperature compensation in LTZ1000A and is not necessary in not A version.

Inviato dal mio Nexus 6P utilizzando Tapatalk

 

Offline d-smes

  • Regular Contributor
  • *
  • Posts: 88
  • Country: us
Re: [FT] LTZ1000A fairy-tale, or the story of little jumper.
« Reply #134 on: November 18, 2016, 08:12:09 pm »
this R is used for temperature compensation in LTZ1000A and is not necessary in not A version.
It's the other way around per data sheet-  don't use resistor for "A" version.
 
The following users thanked this post: mimmus78

Offline mimmus78

  • Supporter
  • ****
  • Posts: 667
  • Country: it
Re: [FT] LTZ1000A fairy-tale, or the story of little jumper.
« Reply #135 on: November 18, 2016, 08:22:18 pm »
Ohh guys I was 100% sure was the other way around ...  |O
 

Online Andreas

  • Super Contributor
  • ***
  • Posts: 2392
  • Country: de
Re: [FT] LTZ1000A fairy-tale, or the story of little jumper.
« Reply #136 on: November 18, 2016, 09:13:59 pm »
Hello,

I use R9 also for the A version to fine adjust the T.C.
But not always 400K is the right value:

https://www.eevblog.com/forum/metrology/ultra-precision-reference-ltz1000/?action=dlattach;attach=253898

with best regards

Andreas
 

Offline bingo600

  • Super Contributor
  • ***
  • Posts: 1360
  • Country: dk
Re: [FT] LTZ1000A fairy-tale, or the story of little jumper.
« Reply #137 on: November 18, 2016, 10:03:48 pm »
@mimmus78

I'd like to get 2 boards from the "overbooking list" if possible

/Bingo
 

Offline dr.diesel

  • Super Contributor
  • ***
  • Posts: 2124
  • Country: us
  • Cramming the magic smoke back in...
Re: [FT] LTZ1000A fairy-tale, or the story of little jumper.
« Reply #138 on: November 18, 2016, 10:52:30 pm »
Did you already ordered resistors? They have extra long leadtime or you are going to use some thin film like PTF56/UXB series?
Good choice can be Rhopoint ( they will fit to TiN boards )
https://www.rhopointcomponents.com/components/resistors/precision-through-hole/wirewound-miniohm-3ppmc-5e10-5g10-series.html

Wow, they stock a big range of 3PPM parts!  I don't have any experience with these guys unfortunately, might be a good alternative depending on how busy Edwin's shop is.

Offline mimmus78

  • Supporter
  • ****
  • Posts: 667
  • Country: it
Re: [FT] LTZ1000A fairy-tale, or the story of little jumper.
« Reply #139 on: November 19, 2016, 12:38:29 am »
Here are options I found:

Vishay VHP101
VHP101 is the most stable and expansive resistors you can get (few PPM for years or °C) ...
if you have big pocket and a very special LTZ1000 you can use this two for the heater divider.
Using VHP101 also for the other resistors to me is just a waste of money.
If you like Vishay name S102C and Z202 are good too, and they can be used for
the other resistors at half the price of the VHP101.

TE Connectivity UPW50

UPW50 has TC 3PPMx°C and 25PPM x year stability.
You can buy at rs components or other store at 7 to 10 EUR each one.
Not all values are directly available, so you must buy multiple items to form the 70K and the 12K resistors.
I brought two sets and I spent less than 150 EUR for both @ rs components. You can find this on findchips.

ECONISTOR 8G16D @ rhopointcomponents
Very similar specs and price to UPW50 but you can request for custom values of "70K" and "12K" values.
They have higher EMF from UPW50. I get a quotation as low as 50 EUR per set. Cannot find this on findchips.

EDWIN
It's the cheapest of all others, specs are similar to the others (UPW50, ECONISTOR) but you have few weeks of lead time.
He had some problems in past with this resistors. I ordered 3 sets from him too, let's see when they arrive.
« Last Edit: November 19, 2016, 12:44:53 am by mimmus78 »
 
The following users thanked this post: Echo88, dr.diesel, bck, lukier

Offline SvanGool

  • Regular Contributor
  • *
  • Posts: 128
  • Country: nl
Re: [FT] LTZ1000A fairy-tale, or the story of little jumper.
« Reply #140 on: November 19, 2016, 12:06:38 pm »
I would like to build one reference, but I probably will not be able to find the resistors here and I don't like the idea getting these from eBay.
Do you guys plan to group buy the low TC resistors from VPG or some distributor?

For the last days, I was in contact with Edwin (the resistor expert) about the resistors for this project, although most answers can be found in the several LTZ1000 threads, I found them very useful, a little summary (with permission of Edwin):
  • Q : People advise me to use 12k5 or 13k iso 12k for the heater in case of a LTZ1000A (which I have), can you advise?
  • A1: In the case of the LTZ circuits, using resistors such as the Vishay Z202 and VHP203 will not get you any noticeable improvement in the Vref's performance.  That circuit is quite insensitive to low resistor TCR and low long term drifting.  Where the resistor performance becomes much more important is in the circuit that raises the LTZ voltage to about 10V at the output.  The best performance of the LTZ is achieved with the Linear Tech circuit as given in the data and app notes.
  • A2: The LTZ1000 (non-A) does allow for a slightly lower overhead temperature internally, about 5°C above the highest environmental temperature expected in operation, whatever the source may be, internal instrument temperature may be higher than room temperature.  Use of 12K, 12.5K, or 13K with the 1000 is often acceptable.  What you don’t want is for the LTZ to run out of compensation room for temperature because it will no longer be able to compensate for temperature and you will see a significant change in output voltage.
  • A3: For the LTZ1000A version, it requires at least 10°C overhead in temperature, this usually required at least 13K or higher.  According to LT, beginning about two years ago, the new production of the ‘A” version was recommended to use a 15K resistor period, this is the value (HP, Agilent,) Keysight, and Fluke are using in their units (the 3458A, ECT.). If I was building an LTZ module I would use the 15K as recommended.
  • Q : Do these heater divider resistors need to be thermally coupled (tracking TC), how would I do that?
  • A : The heater divider resistors is a bit sticky, if you encapsulate the module in say foam and not just a ‘box’, the resistors will achieve equilibrium without any special effort.  If the module is not encapsulated, then wrapping a layer of copper tape around the two resistors will make a significant difference in tracking.  You must remember that whatever the case is, it is going to take some time for everything to warm up and equalize, until then you will see a higher than ‘normal’ drift.  If after about 30-60 minutes you are still seeing some wobble, you have an air draft problem most likely.
  • Q: What is the best way to mount these type 802 resistors in TiNs 0,125" and 0,150" PCB sockets and can I use normal SnPb solder?
  • A : The dimensions of the 802 is 0.250” D x 0.375” L (6.35mm x 9.53mm) with 22 AWG leads, leaded solder is just fine.
  • Q : Do I have to take some special precautions to protect your  resistors against humidity?
  • A : No, the resistors are not humidity sensitive to any significant degree.
  • Q : What would be the cost and lead time for your resistors and could a group discount be applicable?
  • A: For 1-9 pieces, a group buy would lower these prices and reduce shipping cost (1st Class International is $13.50, International Priority is $33.95 for Europe):
    1x  120R0, ±0.1%, type 802       $6.85
    1x  1K, ±0.1%, type 802          $6.34
    1x  12K or 15K, ±0.1%, type 802  $7.35
    2x  70K, ±0.1%, type 802         $8.64

    Current ship dates are approximately 4-5 weeks, perhaps a little less depending on circumstances.

 :-+
« Last Edit: November 19, 2016, 12:30:10 pm by SvanGool »
# Don't hurry, the past will wait. #
 
The following users thanked this post: Echo88, dr.diesel, VK5RC, klaus11, Muxr, villas, 2N3055

Offline mimmus78

  • Supporter
  • ****
  • Posts: 667
  • Country: it
Re: [FT] LTZ1000A fairy-tale, or the story of little jumper.
« Reply #141 on: November 20, 2016, 07:20:15 pm »
UPDATE 25/nov/2016 22:30 Italy time:

All PCB in new batch are already sold but if you accept the one with this small defect I still have someone unsold that I'm not going to use.
If you want it just send me a PM.

Updated pcb shipping status --> HERE
« Last Edit: November 25, 2016, 09:51:28 pm by mimmus78 »
 
The following users thanked this post: dr.diesel, villas

Offline mimmus78

  • Supporter
  • ****
  • Posts: 667
  • Country: it
Re: [FT] LTZ1000A fairy-tale, or the story of little jumper.
« Reply #142 on: November 25, 2016, 09:54:00 pm »
RS components today shipped me the last two precision resistor I was missing for starting build this KX reference.
Is there any easy and effective way to check TC of the resistors without having a thermal chamber?
 

Online Andreas

  • Super Contributor
  • ***
  • Posts: 2392
  • Country: de
Re: [FT] LTZ1000A fairy-tale, or the story of little jumper.
« Reply #143 on: November 25, 2016, 10:30:38 pm »
Hello,

I do it with a peltier (car) cooling box + heater foils on a heat spreader for a ~30 deg C temperature range:
https://www.eevblog.com/forum/metrology/t-c-measurements-on-precision-resistors/msg462298/#msg462298

Frank uses a freezer cooling pad in a isolated box.
https://www.eevblog.com/forum/metrology/t-c-measurements-on-precision-resistors/msg466319/#msg466319

If you want repeatable measurements you should keep the wires of the resistor at the same temperature.
I use aluminium stripes + silicone pads

With best regards

Andreas
 
The following users thanked this post: mimmus78

Online Dr. Frank

  • Super Contributor
  • ***
  • Posts: 1697
  • Country: de
Re: [FT] LTZ1000A fairy-tale, or the story of little jumper.
« Reply #144 on: November 25, 2016, 11:02:08 pm »
Well, I'm using a dynamic temperature change method, so the measurement are always done during ramping up/down of the temperature.
If that change is slow enough, nearly everything is in equilibrium, and the measurement will be precise enough.

My final setup is here: https://www.eevblog.com/forum/metrology/t-c-measurements-on-precision-resistors/msg617211/#msg617211

The aluminium block, which carries the resistor and the thermometer are additionally mounted inside an outer aluminium box so that also all the wires are on the the same temperature as the alu block and the outer case (thermal short circuit), This is essential, as most of heat transfer happens via the resistors wires.

I only have difficulties to observe very slow events correctly, like the described hysteresis and creeping effects on some of the econistor resistors.

It would  be very useful to let the resistors relax on a constant temperature after a change for many hours.
I'd expect that the formerly hysteretical resistors would have a much lower T.C. than being falsified by the hysteresis effects.
General Resistors told me, that they let their DUTs rest for 24h on the edge temperatures, and they had totally different results than me on the same specimen.
 
In that sense, Andreas solution is much more versatile, and the temperature can be controlled in a much better manner, and the whole temperature profile (T-loop) can be achieved automatically.

B.t.w., I will soon publish T.C. adjustment experiments on my new LTZ#5 reference, where these hysteresis effects, supposedly caused by exactly these resistors, are strongly dependent on the temperature change rate, and so does the evident T.C.

For a quick selection and characterization of the PWW resistors and the LTZ circuit itself, my dynamic measurement approach was completely sufficient.. the LTZ reference after a few attempts, now has nearly zero T.C.

Frank
« Last Edit: November 25, 2016, 11:16:54 pm by Dr. Frank »
 
The following users thanked this post: mimmus78

Offline mimmus78

  • Supporter
  • ****
  • Posts: 667
  • Country: it
Re: [FT] LTZ1000A fairy-tale, or the story of little jumper.
« Reply #145 on: December 06, 2016, 02:53:01 pm »
Something pop out todays ... unfortunately I'm still dealing with resistors TC characterisation where I found something strange (to me) and I hope I can publish some results in a few days.

It's not the time to build the KX yet ...[emoji41]
 

Online TiN

  • Super Contributor
  • ***
  • Posts: 4045
  • Country: tw
  • xDevs.com/live - 24/7 lab feed
    • xDevs.com
Re: [FT] LTZ1000A fairy-tale, or the story of little jumper.
« Reply #146 on: December 06, 2016, 04:14:05 pm »
Candies!  :popcorn:

I'm doing TCR testing too, but except it being painfully slow, nothing strange so far. Need build separate metal slug thermal unit for faster resistor testings.
YouTube | Chat room | Live-cam | Have documentation to share? Upload here! No size limit, firmware dumps, photos.
 

Offline mimmus78

  • Supporter
  • ****
  • Posts: 667
  • Country: it
Re: [FT] LTZ1000A fairy-tale, or the story of little jumper.
« Reply #147 on: December 08, 2016, 02:40:07 am »
Well really is not strange ...
 
As Dr. Frank wrote somewhere after some thermal stress you need to leave the resistor rest for many hours to make them recovery some thermal caused hysteresis ... only in this case you can get the real TC of the resistor. I found some resistors have more than 20% of TC improvement after waiting 16 hours. I also found this hysteresis to improve with thermal cycles ... than more you therm cycle the resistor than more it gets better.

Unfortunately I don't have so much data to have a clear idea of what is happening ... and collect this will take forever.

Hope I can post some chart in some days.
 

Online TiN

  • Super Contributor
  • ***
  • Posts: 4045
  • Country: tw
  • xDevs.com/live - 24/7 lab feed
    • xDevs.com
Re: [FT] LTZ1000A fairy-tale, or the story of little jumper.
« Reply #148 on: December 08, 2016, 04:17:39 am »
Quote
collect this will take forever

Just two weeks :)



Data DSVs: Ramp up, Ramp down.
YouTube | Chat room | Live-cam | Have documentation to share? Upload here! No size limit, firmware dumps, photos.
 

Online Andreas

  • Super Contributor
  • ***
  • Posts: 2392
  • Country: de
Re: [FT] LTZ1000A fairy-tale, or the story of little jumper.
« Reply #149 on: December 08, 2016, 07:15:51 am »
What resistors did you use?
Hermetically or with plastic housing.

Guess with this long time there is some humidity influence.
And some creeping effect if plastic housing.

With best regards

Andreas
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf