• State Not Answered
  • Replies 2 replies
  • Subscribers 72 subscribers
  • Views 485 views
  • Users 0 members are here
Options
Related

AD7134 with ADA4254

squirest

Hi,

My application is to create an 8 channel, synchronously-sampled data acquisition system for signals from +-5mV to +-10V synchronised to an external 40kHz sample clock (and I'm looking at ways to increase that for various performance & spec reasons). I'm looking at the AD7606B, AD7609 and now the AD7134 ADCs. 

Some signals are direct transducer inputs whilst others are in the form of an amplitude modulated 5kHz sine. I like the functionality offered by the ADA4254 but I wonder if it makes sense as an AFE for the AD7132. Current nett bandwidth is only a 20kHz but that's something I'm trying to extend too. Can anyone offer advice?

Thanks, Steve



Typo
[edited by: squirest at 12:01 PM (GMT 0) on 13 May 2020]
Parents
  • 0
    •  Analog Employees 
    on May 14, 2020 8:07 AM

    hi steve 

    what are your performance targets in terms of noise and DC performance? ADA4254 has a little higher noise. but if thats ok ADA4254 can be used with the AD7134.

    thanks

  • 0 on May 14, 2020 8:05 PM in reply to Wa$iM

    Hi, Wa$sim

    We get a Johnson noise from the bridge transducer (200 to 600nV rms at our current 20kHz bandwidth) so I expect the lower-end bits to be swamped. I can mitigate this somewhat by 16*oversampling down to 40ksps ODR.

    To be honest, 24 bits of resolution more than we need - 18 bits would be sufficient (our current system is 16 bit, 40ksps). Having a high-speed, high resolution ADC is more for specmanship purposes.

    FYI, I am currently running a bench prototype with an ADA4254 EVM and AD7609 EVM looking at a 50mVpp 5kHz sine and the data looks, subjectively, at least as good as our current system. I'm investigating the AD7134 as the digital end may be easier to interface to a Cortex M7 via SPI and DMA the data out.

    Steve

Reply
  • 0 on May 14, 2020 8:05 PM in reply to Wa$iM

    Hi, Wa$sim

    We get a Johnson noise from the bridge transducer (200 to 600nV rms at our current 20kHz bandwidth) so I expect the lower-end bits to be swamped. I can mitigate this somewhat by 16*oversampling down to 40ksps ODR.

    To be honest, 24 bits of resolution more than we need - 18 bits would be sufficient (our current system is 16 bit, 40ksps). Having a high-speed, high resolution ADC is more for specmanship purposes.

    FYI, I am currently running a bench prototype with an ADA4254 EVM and AD7609 EVM looking at a 50mVpp 5kHz sine and the data looks, subjectively, at least as good as our current system. I'm investigating the AD7134 as the digital end may be easier to interface to a Cortex M7 via SPI and DMA the data out.

    Steve

Children
No Data