Author Topic: Linear AN86-related DAC-Question  (Read 3502 times)

0 Members and 1 Guest are viewing this topic.

Offline Echo88Topic starter

  • Frequent Contributor
  • **
  • Posts: 821
  • Country: de
Linear AN86-related DAC-Question
« on: November 08, 2016, 09:03:40 pm »
I was under the impression that in a DAC, once it was set for a value, the corresponding switch will never change its state on its on. But now that i see strange voltage hops on my selfbuilt AN86-dual-DAC-circuit im doubting that. See the attached pictures, it seems to me that these are jumps corresponding to a change in the LSB of the MSB-DAC.

Of course in the end the AN86-circuit is closedloop and the error can be disregarded, but it just interests me. Maybe someone can contribute on this topic.

My circuit consists of a DC1684A-B 16-Bit-Dual-DAC-board, summed by 0,1% metal-film smd-resistors and a good OP i had laying around (OPA2277 i think). Circuit and DAC-Board of course thermally stabilized for hours and data-control connector disconnected after the 5,0V value was set.
 

Offline Andreas

  • Super Contributor
  • ***
  • Posts: 3221
  • Country: de
Re: Linear AN86-related DAC-Question
« Reply #1 on: November 08, 2016, 10:15:43 pm »
Hello,

at least the upper trace looks like popcorn noise (of the voltage reference).
Which voltage reference does the board use?

With best regards

Andreas
 

Offline Cerebus

  • Super Contributor
  • ***
  • Posts: 10576
  • Country: gb
Re: Linear AN86-related DAC-Question
« Reply #2 on: November 08, 2016, 10:18:51 pm »
Hello,

at least the upper trace looks like popcorn noise (of the voltage reference).
Which voltage reference does the board use?

With best regards

Andreas

That top trace is nearly 15 hours, those dips are several minutes long.
Anybody got a syringe I can use to squeeze the magic smoke back into this?
 

Offline Echo88Topic starter

  • Frequent Contributor
  • **
  • Posts: 821
  • Country: de
Re: Linear AN86-related DAC-Question
« Reply #3 on: November 08, 2016, 10:23:41 pm »
It uses the standard populated LTC6655CHMS8-5.
 

Offline ManateeMafia

  • Frequent Contributor
  • **
  • Posts: 730
  • Country: us
Re: Linear AN86-related DAC-Question
« Reply #4 on: November 08, 2016, 10:27:39 pm »
What are you using for the +/- 15V supply?

Also, do you have a way of logging temperature? It looks like the meter is tracking ambient temperature.
 

Offline Andreas

  • Super Contributor
  • ***
  • Posts: 3221
  • Country: de
Re: Linear AN86-related DAC-Question
« Reply #5 on: November 08, 2016, 10:29:10 pm »
Hello,

not uncommon for certain references.
I have this at least on one sample of MAX6350.
(plot has 6 minutes total time).

With best regards

Andreas
 

Offline Andreas

  • Super Contributor
  • ***
  • Posts: 3221
  • Country: de
Re: Linear AN86-related DAC-Question
« Reply #6 on: November 08, 2016, 10:32:37 pm »
It uses the standard populated LTC6655CHMS8-5.

Thermally isolated?
The 6655 heats quite a lot.

What are you using for the +/- 15V supply?

I hope the 6655 has a lower power supply voltage than 15 V.

With best regards

Andreas
 

Offline Echo88Topic starter

  • Frequent Contributor
  • **
  • Posts: 821
  • Country: de
Re: Linear AN86-related DAC-Question
« Reply #7 on: November 08, 2016, 10:55:36 pm »
The LTC6655 isnt thermally isolated and is supplied with 12V. The +-15V are provided by my Agilent 6629A Supply. But if it would be popcorn noise, then the second attached trace wouldnt make sense, as it stays low for about 45mins.

http://cds.linear.com/docs/en/demo-board-manual/DC1684AABFB.PDF
 

Offline Andreas

  • Super Contributor
  • ***
  • Posts: 3221
  • Country: de
Re: Linear AN86-related DAC-Question
« Reply #8 on: November 09, 2016, 07:12:59 am »
could also be mechanical stress from PCB.
I have measured 60-70uV change by applying mechanical stress to a PCB with
 LTC6655BHMS8-5 mounted in a U-slotted area.

with best regards

Andreas
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf