On Mon, 17 Aug 2015 17:04:25 +0200
Andreas Jahn Andreas_-_Jahn@t-online.de wrote:

Yes, when i tried to understand the circuit, more and more question
marks popped up, until I decided to use a different approach that
does not involve black magic and misleading descriptions.

I am kind of surprised that the charge injection of the MAX4053
does not kill a lot of the performance. At least i don't see any
way how it would cancel somewhere.

		Attila Kinali

--
It is upon moral qualities that a society is ultimately founded. All
the prosperity and technological sophistication in the world is of no
use without that foundation.
-- Miss Matheson, The Diamond Age, Neil Stephenson

On Mon, 17 Aug 2015 15:38:37 +0000
"Poul-Henning Kamp" phk@phk.freebsd.dk wrote:

Hmm. The ADS1262 does indeed seem quite easily available.

I thought about using the LT1021. I don't need the long term
stability of the LTZ1000 and the LT1021 is available for less than 10USD.

But from the discussion here, think i have to go back and see whether
i really need such an high resolution or whether 20bit wouldnt be already
enough. Then i could get away with a 20bit DAC like the DAC1220.

Thanks

		Attila Kinali

--
It is upon moral qualities that a society is ultimately founded. All
the prosperity and technological sophistication in the world is of no
use without that foundation.
-- Miss Matheson, The Diamond Age, Neil Stephenson

I would think the noise numbers would kill most of the bits past 24. DAC
noise is bad enough but do you have a 32 bit reference. To be real the
reference voltage into the DAC need stable to 1 part in 4,.294,967,296 or
just 0.004 ppm
How can you use bits 25 to 32?

Just something to consider in your design.
If you are using this as an attenuator then what is the noise level in the
input signal.

On Mon, Aug 17, 2015 at 9:04 AM, Attila Kinali attila@kinali.ch wrote:

--

John Phillips

Hello,

I would rather use a reference with lower tempco than the LT1021.
(typical 7ppm/K)
E.g. (selected) AD586LQ or MAX6350.
Or at least LT1236AILS8 with around 3ppm/K.
The self heating alone gives around 3 deg C temperature difference.

With best regards

Andreas

Am 17.08.2015 um 18:19 schrieb Attila Kinali:

Look at LT6655 datasheet, page 14, about long term drift of MS8 versus
LS8 package. If others refferences are as bad regarding moisture then i
think this may be a problem at this level of accuracy.....

Daniel

Em 17/08/2015 13:29, Andreas Jahn escreveu:

Hello,

its not only the plastic (epoxy) package.
Its also the epoxy PCB which creates package stress to the die.
And I do not believe that any "slots" make it significantly better.
Thats why I use the AD586LQ; and soldering only one pin to the PCB.
(the others with thin VERO-wire).

With best regards

Andreas

Am 17.08.2015 um 18:46 schrieb Daniel Mendes:

At this level even barometric pressure changes con be a problem.

On Mon, Aug 17, 2015 at 10:58 AM, Andreas Jahn Andreas_-_Jahn@t-online.de
wrote:

--

John Phillips

Hi:

I haven't been following this thread but see it's about making a DAC with a lot of bits.
While researching a Quantec GPS receiver I discovered a patent that for a 4 bit DAC to be used to drive a precision
oscillator.
http://www.prc68.com/I/Q5200.shtml#Patent
It depends on using the carry out from an NCO that has a 48 bit input word.

PS HP/Agilent has some instruments with very high quality programmable DC sources such as the 4352 VCO tester.  These
have very low noise but I forget how many bits.

Mail_Attachment --
Have Fun,

Brooke Clarke
http://www.PRC68.com
http://www.end2partygovernment.com/2012Issues.html
http://www.prc68.com/I/DietNutrition.html
Andreas Jahn wrote: