Author Topic: Comparing two references  (Read 2258 times)

0 Members and 1 Guest are viewing this topic.

Offline GrandchuckTopic starter

  • Frequent Contributor
  • **
  • Posts: 624
  • Country: us
Comparing two references
« on: January 22, 2019, 07:52:03 pm »
Analog Devices offers the EVAL-AD5791SDZ with a companion ADR445 voltage reference.  An LTZ1000 reference is an expensive option.  A comparison shows that the expense is warranted.  Both screen shots are from 1 hour runs using Ian Johnston’s data logging software.  The vertical range is the same for both screens.  The difference is significant.

Also shown are the circuit board and the schematic for the ADR445.

The data were collected differently.  The ADR was connected directly to the DMM and the LTZ served as the reference for the AD5791.  Another variation is that the LTZ was housed in a tight cabinet and the ADR was exposed.  However, both runs were completed with no heat or AC on and any drafts were minimal.  The ADR was powered with +/- 15 volts from a lab grade linear supply.

Two capacitors appear to be missing (C7 and C8) … could mean extra noise in the output?  An oscilloscope does show more noise when the ADR reference is on.
 
The following users thanked this post: IanJ, Andreas, splin, hwj-d

Offline hwj-d

  • Frequent Contributor
  • **
  • Posts: 676
  • Country: de
  • save the children - chase the cabal
Re: Comparing two references
« Reply #1 on: January 23, 2019, 09:58:15 am »
I'd be interested in the way the ltz reference was constructed, or what type it is. Do you have a picture from inside the cabinet?

Thanks

Edit:
Found your "AD5791 & LTZ1000 in a box" - thread. So, i assume, that's the original exposed on the EVAL-Board, or the additional ADS ltz1000 board right?
« Last Edit: January 23, 2019, 10:27:47 am by hwj-d »
 

Offline splin

  • Frequent Contributor
  • **
  • Posts: 999
  • Country: gb
Re: Comparing two references
« Reply #2 on: January 23, 2019, 01:43:16 pm »
The ADR445 LF noise spec is 2.25uVpp typical compared to 1.2 for the LTZ1000 and yet the '5791 measurements are approx 60uVpp and 5uV respectively.  Something is very odd especially as a resistive divider DAC should be an easy load for a reference, unlike SAR and sigma delta ADCs.
 

Offline GrandchuckTopic starter

  • Frequent Contributor
  • **
  • Posts: 624
  • Country: us
Re: Comparing two references
« Reply #3 on: January 23, 2019, 06:47:11 pm »
The daughterboard made by Analog Devices (see attached picture).  The schematic is also attached.
 
The following users thanked this post: hwj-d

Offline GrandchuckTopic starter

  • Frequent Contributor
  • **
  • Posts: 624
  • Country: us
Re: Comparing two references
« Reply #4 on: January 23, 2019, 06:48:27 pm »
Please see attached graph.  Thanks.
 

Offline Magnificent Bastard

  • Regular Contributor
  • *
  • Posts: 139
  • Country: aq
Re: Comparing two references
« Reply #5 on: January 23, 2019, 07:05:28 pm »
The ADR445 LF noise spec is 2.25uVpp typical compared to 1.2 for the LTZ1000 and yet the '5791 measurements are approx 60uVpp and 5uV respectively.  Something is very odd especially as a resistive divider DAC should be an easy load for a reference, unlike SAR and sigma delta ADCs.

I don't trust ADI's data-sheet spec on this part.  The original ADI reference-IC group is known to make "false assertions" about their reference products.   :box:
 

Offline Andreas

  • Super Contributor
  • ***
  • Posts: 3221
  • Country: de
Re: Comparing two references
« Reply #6 on: January 23, 2019, 08:00:19 pm »
Hello,

I think you cannot compare the 3458A measurements (~6 samples per minute) with 0.1 ... 10 Hz noise measurements.
Slow drifts are usually below 0.1 Hz and filtered out by the high pass.
And keep in mind that the noise @5V will be amplified by a factor of 2 on the output.

On a similar device (ADR4550B) I measured 2.7uVpp in average with a 0.1-10 Hz LNA (Datasheet value is 2.8uVpp)
Of course those measurements are made battery supplied in a metal cookies box to keep EMI away and covered with cloth against air drafts.
I also have usually a 100 nF capacitor on the output of the reference (If it is stable with capacitive loads).

with best regards

Andreas
 

Offline Svgeesus

  • Regular Contributor
  • *
  • Posts: 73
  • Country: us
Re: Comparing two references
« Reply #7 on: January 24, 2019, 02:14:38 pm »
On a related "comparing two references" note, I posted earlier about comparing a PX LTZ1000 reference with the LM399 references in Ian Johnston PDVS2 (and in my Keysight 34465A). I set the two to opposing and notionally equal voltages, so I could measure the difference:

https://www.eevblog.com/forum/metrology/usa-cal-club-round-2/msg1972697/#msg1972697
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf