10mv Resolution

[tt]

Hi, All

My boss want a DDS signal generator having 10mv resolution output. Does that
kind of stuff exist on the world? D/A Vref error, INL DNL error, amp op
error and system noise would make 10mv not available. How could precise
instrument achieve that kind of resolution?


Regards

 

[PeteS]

Without wishing to seem pedantic, resolution and precision are not the
same thing, although they can overlap.

The resolution of a system may be millivolts, but it's overall error
much larger, where the error is precision.

Perhaps a better question for you is:

Do you need 10mV of differential resolution (and what precision) or do
you need 10mV of precision?

Cheers
PeteS

 

[John Larkin]

On Mon, 14 Mar 2005 16:42:33 +0800, "tt" <tt@ee.ff> wrote:

Hi, All

My boss want a DDS signal generator having 10mv resolution output. Does that
kind of stuff exist on the world? D/A Vref error, INL DNL error, amp op
error and system noise would make 10mv not available. How could precise
instrument achieve that kind of resolution?


Regards

10 mv out of what?

John

 

[Winfield Hill]

tt wrote...

My boss want a DDS signal generator having 10mv resolution output.
Does that kind of stuff exist on the world? D/A Vref error, INL DNL
error, amp op error and system noise would make 10mv not available.
How could precise instrument achieve that kind of resolution?

I've been using AD's nice 14-bit DDS chips, a pair of balanced
10mA fs current-source outputs with 14-bit resolution, which is
500mV fs, and 61uV or so lsb into 50 ohms.


--
Thanks,
- Win

 

[Adam Kumpf]

Hi tt,

10mV is not that difficult to achieve if you get it indirectly.

Imagine that you a 50ohm output impedance is desired. (as with most signal
generators) Now produce a 1Vpeak-to-peak output, pass it through a voltage
divider (say.. 10k on top, and 10ohm to ground) and at that midpoint, you
have a 1mV signal and the output impedance is parallel combination of the
two resistances to AC ground. (10 // (10k + 50) if the original signal
generator had 50 ohm output). Then put a 40 ohm in series with the center
tap and, there you have it.. a 50ohm impedance signal generator with a 1mV
signal output and a relatively easy to create 1V signal input.

These kinds of attenuators are used all of the time when testing high
gain circuits. I don't see any reason why it couldn't help you out here..
unless you want to go so low that you are greatly affected by thermal noise.
)

This does limit the upper range of your signal generator (you would need
1000V for a 1V output), so perhaps voltages this low may require a special
mode to be used. Depending on the frequency ranges you are looking at, a
relay could switch between the modes fairly easily.

I hope that helps a bit.

Best Regards,
Adam Kumpf

kumpf@mit.edu
http://web.mit.edu/kumpf/www/kumpf-projects.html


"tt" <tt@ee.ff> wrote in message news:d13iqr$hl2$1@newshost.mot.com...

Hi, All

My boss want a DDS signal generator having 10mv resolution output. Does
that
kind of stuff exist on the world? D/A Vref error, INL DNL error, amp op
error and system noise would make 10mv not available. How could precise
instrument achieve that kind of resolution?


Regards

 

[Winfield Hill]

Winfield Hill wrote...

tt wrote...

My boss want a DDS signal generator having 10mv resolution output.
Does that kind of stuff exist on the world? D/A Vref error, INL DNL
error, amp op error and system noise would make 10mv not available.
How could precise instrument achieve that kind of resolution?

I've been using AD's nice 14-bit DDS chips, a pair of balanced
10mA fs current-source outputs with 14-bit resolution, which is
500mV fs, and 61uV or so lsb into 50 ohms.

Right, 50 ohms ... These parts use 400MHz clocks.


--
Thanks,
- Win