Hello: I'm using the MSP430F149 ADC unit to sample a sinusoidal signal from a waveform generator. Whenever I run the program, a high frequency sinusoidal noise is added to the ADC input signal. I noticed that the noise sinusoid has a frequency of 4.8MHz which is the frequency of the external crystal oscillator I am using as system clock. If I turn this external oscillator off and use the internal system clock instead, the 4.8MHz signal dissapears. Still, I get some low frequency noise but it is almost negligible. How should I connect the external oscillator? Could this be the cause of the noise? Does anyone know the reasons behind this phenomenon and how to solve it? Any ideas will be really appreciated.
ADC noise caused by external crystal oscillator
Started by ●May 18, 2005
Reply by ●May 18, 20052005-05-18
Hello, This sounds like coupling. Either the trace from the oscillator to the MSP radiates or couples into the ADC input or you have 4.8MHz riding on the supply. Is there a full common ground plane? If not that would be my first concern. It is hard to try a diagnosis without seeing the unit. If you post a layout that might help. Regards, Joerg.
Reply by ●May 19, 20052005-05-19
Thanks for the fast response. We are using the SoftBaugh B149 Breadboard Evaluation Board. The crystal oscillator is connected between pins x2I and x2O and the ADC input is connected to port 6 bit 0 (pin 6.0). The pins are next to each other. The circuit ground is common to all components. The schematic for the board can be found at http://www.softbaugh.com/ProductPage.cfm?strPartNo49 Any help you can give me regarding this matter would be really appreciated. Thanks --- In msp430@msp4..., "Joerg Schulze-Clewing" <joergsch@a...> wrote: > Hello, > > This sounds like coupling. Either the trace from the oscillator to the > MSP radiates or couples into the ADC input or you have 4.8MHz riding > on the supply. Is there a full common ground plane? If not that would > be my first concern. > > It is hard to try a diagnosis without seeing the unit. If you post a > layout that might help. > > Regards, > > Joerg.
Reply by ●May 19, 20052005-05-19
Hello,
> The schematic for the board can be found at
> http://www.softbaugh.com/ProductPage.cfm?strPartNo49
This board doesn't have a 4.8MHz oscillator. Do you run that
oscillator on the board that it plugs into? Considering that B149
doesn't appear to have a real ground plane and the processor is quite
far from the connector you may still have a ground loop or
insufficient ground conductivity.
One way to find out would be to (carefully) connect a piece of copper
tape from the ground plane on your board to a ground very close to the
MSP430. Short distances, no wires.
Regards,
Joerg.
Reply by ●May 19, 20052005-05-19
Maybe what you are seeing is an artifact caused by sample/conversion timing and not an induced 4.8MHz from nearby PCB traces. In order to see the 4.8MHz noise you probably are using a derivative of main clock to feed ADC circuit instead of internal ADC12OSC. You also need to be triggering the sample/conversion by software and not as an autorepeat mode (to be able to see such a high frequency you must sample at different points (phase) each time and this would not happen in the auto mode). What is the relative amplitude of input signal to the measured 4.8MHz noise? There are an internal crosstalk that could pick-up the main clock. It is normally around 1 or 2 divisions high (for adjacent channels) and easily filtered in software. Did you try to put a series resistor between ADC input and signal generator I did not see the schematics yet)? It could use the input capacitance to filter the high frequency noise in case it is coupled in the cable from signal generator. Hope it helps. -Augusto -----Original Message----- From: msp430@msp4... [HYPERLINK "mailto:msp430@msp4..."mailto:msp430@msp4...] On Behalf Of elcoqui00 Sent: Thursday, May 19, 2005 4:01 PM To: msp430@msp4... Subject: [msp430] Re: ADC noise caused by external crystal oscillator Thanks for the fast response. We are using the SoftBaugh B149 Breadboard Evaluation Board. The crystal oscillator is connected between pins x2I and x2O and the ADC input is connected to port 6 bit 0 (pin 6.0). The pins are next to each other. The circuit ground is common to all components. The schematic for the board can be found at HYPERLINK "http://www.softbaugh.com/ProductPage.cfm?strPartNo49"http://www.softbaug h.com/ProductPage.cfm?strPartNo49 Any help you can give me regarding this matter would be really appreciated. Thanks --- In msp430@msp4..., "Joerg Schulze-Clewing" <joergsch@a...> wrote: > Hello, > > This sounds like coupling. Either the trace from the oscillator to the > MSP radiates or couples into the ADC input or you have 4.8MHz riding > on the supply. Is there a full common ground plane? If not that would > be my first concern. > > It is hard to try a diagnosis without seeing the unit. If you post a > layout that might help. > > Regards, > > Joerg. . "http://docs.yahoo.com/info/terms/ -- Internal Virus Database is out-of-date. Checked by AVG Anti-Virus. Version: 7.0.308 / Virus Database: 266.11.7 - Release Date: 9/5/2005
Reply by ●May 19, 20052005-05-19
Hello: Yes, we have the 4.8MHz oscillator connected to the board where the MSP evaluation board plugs into. --- In msp430@msp4..., "Joerg Schulze-Clewing" <joergsch@a...> wrote: > Hello, > > > The schematic for the board can be found at > > http://www.softbaugh.com/ProductPage.cfm?strPartNo49 > > This board doesn't have a 4.8MHz oscillator. Do you run that > oscillator on the board that it plugs into? Considering that B149 > doesn't appear to have a real ground plane and the processor is quite > far from the connector you may still have a ground loop or > insufficient ground conductivity. > > One way to find out would be to (carefully) connect a piece of copper > tape from the ground plane on your board to a ground very close to the > MSP430. Short distances, no wires. > > Regards, > > Joerg.
Reply by ●May 19, 20052005-05-19
wHEN CONNECTING A CRYSTAL OSCILLATOR, AS OPPOSED TO A CRYSTAL, YOU only
CONNECT TO THE x2i PIN. yOU do not connect TO x2o. Check the manuals. As
they say RTFM before messing with the circuit.
Al
elcoqui00 wrote:
>Thanks for the fast response. We are using the
SoftBaugh B149
>Breadboard Evaluation Board. The crystal oscillator is connected
>between pins x2I and x2O and the ADC input is connected to port 6 bit
>0 (pin 6.0). The pins are next to each other. The circuit ground is
>common to all components.
>
>The schematic for the board can be found at
>http://www.softbaugh.com/ProductPage.cfm?strPartNo49
>
>Any help you can give me regarding this matter would be really
>appreciated.
>
>Thanks
>
>--- In msp430@msp4..., "Joerg Schulze-Clewing"
<joergsch@a...>
>wrote:
>
>
>>Hello,
>>
>>This sounds like coupling. Either the trace from the oscillator to
>>
>>
>the
>
>
>>MSP radiates or couples into the ADC input or you have 4.8MHz riding
>>on the supply. Is there a full common ground plane? If not that would
>>be my first concern.
>>
>>It is hard to try a diagnosis without seeing the unit. If you post a
>>layout that might help.
>>
>>Regards,
>>
>>Joerg.
>>
>>
>
>
>
>
>.
>
>
>Yahoo! Groups Links
>
>
>
>
>
>
>
>
>
>
Reply by ●May 19, 20052005-05-19
RTFM, yes. Thanks for the laugh, I had a bit of a chuckle.
Manuals aside, applying basic electronics knowledge should give them a
hint, heh? pssssssssssttt!
*Peter*
Onestone wrote:
>wHEN CONNECTING A CRYSTAL OSCILLATOR, AS OPPOSED TO
A CRYSTAL, YOU only
>CONNECT TO THE x2i PIN. yOU do not connect TO x2o. Check the manuals. As
>they say RTFM before messing with the circuit.
>