[SI-LIST] Re: ADC Gnd Noise
- From: "Peterson, James F \(EHCOE\)" <james.f.peterson@xxxxxxxxxxxxx>
- To: <si-list@xxxxxxxxxxxxx>
- Date: Thu, 12 Apr 2007 08:30:51 -0500
Hey, interesting problem. An ADC can really shine a light on noise
problems. In your case you could almost say the level of noise you're
having to deal with is what is showing up in the toggling of your 8
lower bits....
so 256 x 0.3mv=3D75mv, and not the 300mv you measure with your scope.=20
If you haven't already, you might try injecting at the sensor input a
clean, pure, signal at several known voltage levels (known more
precisely than your lsb bit weight of 0.3mv). Compare the input voltages
to the digitized values and see if this 75mv noise stays the same at
different levels of input. This doesn't fix the problem, just gets you
more data that might provide leads.
Best regards,
Jim Peterson
Honeywell
-----Original Message-----
From: si-list-bounce@xxxxxxxxxxxxx [mailto:si-list-bounce@xxxxxxxxxxxxx]
On Behalf Of istvan novak
Sent: Thursday, April 12, 2007 7:56 AM
To: kd_kamath@xxxxxxxxx
Cc: Gupta, Deepali; prajitsnair@xxxxxxxxx; si-list@xxxxxxxxxxxxx
Subject: [SI-LIST] Re: ADC Gnd Noise
Hi Deepak,
You mentioned earlier that the analog and digital grounds are
connected through a ferrite bead. Assuming that you connect
the analog and digital grounds at a single-well-chosen point,
you could try to replace the ferrite bead with a wire or braid short.
Regards,
Istvan Novak
SUN Microsystems
deepak kamath wrote:
>hi,
> Thanks for the suggestion. We have measured analog ground noise with
respect to digital ground and it is around 300mV(p-p) on the scope. Even
if we subtract 60mV noise due to probe, the noise will be around
180mV(p-p). When there is no signal to ADC, the 8 LSB bits from ADC are
toggleing. That is, ADC is seeing around 80mV(p-p) noise at its input.=20
> =20
> The clock of 7.414MHz is generated using PLL inside the FPGA and is
passed through a 3.3V to 5V transceiver having driving capacity of 25mA.
The clock output from the transceiver is routed to ADC without any
sheilding. We are observing overshoots and undershoots of about 2V(p-p)
on the ADC clock, but there is no shift in the edges of the clock.=20
> Yes, we have checked both grounds and we are observing the noise of
around 300mV(p-p). That is, AGND w.r.t AGND has 300mV(p-p), DGND w.r.t
DGND has 300mV(p-p) and AGND w.r.t DGND has 300mV(p-p).
> =20
> Thank you and regards,
> Deepak.
> =20
>"Gupta, Deepali" <Deepali.Gupta@xxxxxxxxx> wrote:
> Hi Deepak ,
>First thing which does glare out is the stack-up. For a 12 layer board
>there are only two GND planes which will not be sufficient. The inner
>layers have no GND planes for reference. Not sure on how the board is
>routed but the sensor lines going to the ADC might just be getting
noise
>due to SI issues related to transmission lines and not due to noise on
>the GND planes. So I think you should spend sometime time looking at
the
>routing of the actual sensor lines and the main clock lines and see
what
>kind of SI issues maybe occurring on the these transmission lines.
>
>Second .....how did you measure the GND noise. I have seen that using
>basic probes give an inherent noise of 60-70mV of noise. (meaning that
>you see it but it is not actually there)=20
>
>Regards,
>Deepali
>
>
>-----Original Message-----
>From: si-list-bounce@xxxxxxxxxxxxx
[mailto:si-list-bounce@xxxxxxxxxxxxx]
>On Behalf Of PRAJIT S NAIR
>Sent: Tuesday, April 10, 2007 11:08 PM
>To: kd_kamath@xxxxxxxxx
>Cc: si-list@xxxxxxxxxxxxx
>Subject: [SI-LIST] Re: ADC Gnd Noise
>
>Hello Deepak..
>I think it is difficult to comment on the board as you have only given
>information on PCB layers and power connection. Routing Information is
>not
>available.:
>
>Did u check the ADC clock jitter?
>Did u check the ground planes for analog, digital and clock signals.
>
>With Best Regards
>Prajit S Nair
>
>
>On 4/11/07, deepak kamath wrote:
> =20
>
>>Hi,
>>
>>I have designed a signal processing board to sample thermal sensor
>> =20
>>
>data,
> =20
>
>>process data and display images. The sensor has the sensitivity of
>> =20
>>
>20mV and
> =20
>
>>an ADC (Analog Devices, AD9240,14-bits) is used to sample the analog
>> =20
>>
>signal
> =20
>
>>from the sensor. ADC has 5 volt analog and digital supply with analog
>> =20
>>
>and
> =20
>
>>digital grounds. The sensitivity of the ADC is 0.3mV per bit and is
>>configured for positive single ended mode of operation with input
>> =20
>>
>dynamic
> =20
>
>>range of 5 volts. The board also has FPGA and DSP which are used to
>> =20
>>
>process
> =20
>
>>the sensor signal. The clocks used are 100MHz to drive FPGA & DSP with
>> =20
>>
>ADC
> =20
>
>>operating at 10MHz.
>>
>>The required power generation of 5 volts, 3.3 volts and
>> =20
>>
>1.2volts are done using switching regulators and are mounted on a
>separate
> =20
>
>>board called power supply board. From power supply board power is
>> =20
>>
>given to
> =20
>
>>signal processing board. The received powers in signal processign
>> =20
>>
>board are
> =20
>
>>porperly filtered and converted into power planes of 5 volts, 3.3
>> =20
>>
>volts
> =20
>
>>and 1.2 volts. Similarly analog and digital grounds are etched out in
>> =20
>>
>a
> =20
>
>>single ground plane and are connected via ferrite bead.
>>
>>The problem we are facing is a noise of 80mV(p-p) is
>>observed on analog ground, digital ground, analog 5 vots supply and
>> =20
>>
>digital
> =20
>
>>5 volts supply at the ADC pins. This noise is corrupting sensor
>> =20
>>
>signal.
> =20
>
>>ADC's LSB 6 bits are toggeling even when no signal is applied to ADC.
>> =20
>>
>The
> =20
>
>>signal processing board constitutes of 12 layers and following layer
>> =20
>>
>stack
> =20
>
>>is followed.
>>1. Top signal layer
>>2. Ground Plane layer
>>3. Power plane layer
>>4. Signal layer
>>5. Signal layer
>>6. Signal layer
>>7. Signal layer
>>8. Signal layer
>>9. Signal layer
>>10. Power layer
>>11. Ground layer
>>12. Bottom signal layer
>>
>>Please suggest me ways to reduce ground and power noise.
>>
>>Regards
>>Deepak
>>
>>---------------------------------
>>Don't be flakey. Get Yahoo! Mail for Mobile and
>>always stay connected to friends.
>>
>> =20
>>
------------------------------------------------------------------
To unsubscribe from si-list:
si-list-request@xxxxxxxxxxxxx with 'unsubscribe' in the Subject field
or to administer your membership from a web page, go to:
//www.freelists.org/webpage/si-list
For help:
si-list-request@xxxxxxxxxxxxx with 'help' in the Subject field
List technical documents are available at:
http://www.si-list.net
List archives are viewable at: =20
//www.freelists.org/archives/si-list
or at our remote archives:
http://groups.yahoo.com/group/si-list/messages
Old (prior to June 6, 2001) list archives are viewable at:
http://www.qsl.net/wb6tpu
=20
------------------------------------------------------------------
To unsubscribe from si-list:
si-list-request@xxxxxxxxxxxxx with 'unsubscribe' in the Subject field
or to administer your membership from a web page, go to:
//www.freelists.org/webpage/si-list
For help:
si-list-request@xxxxxxxxxxxxx with 'help' in the Subject field
List technical documents are available at:
http://www.si-list.net
List archives are viewable at:
//www.freelists.org/archives/si-list
or at our remote archives:
http://groups.yahoo.com/group/si-list/messages
Old (prior to June 6, 2001) list archives are viewable at:
http://www.qsl.net/wb6tpu
Other related posts:
