Post Go back to editing

AD8021 and DAC AD9102 noise and limitation

Category: Hardware
Product Number: AD8021 and AD9102

Hello,

I use the DAC 9102 and I need to be able to generate an output in -10/+10v or -5/+5v on frequencies of a few hertz up to several tens of MHz ideally. For this I added the AD8021 because it is one of the only ones to my knowledge that allows a -10/+10v output on high frequencies (the hundred MHz at -3dB). Below is my schematics based on evalboard AD9102-ARDZ-EBZ_Schematic.pdf:

The ADG1401 is there to control the switch to go from an amplification of 5 to 10.

Are there anything that shocks you about this schematics?

I have several "problems" with this setup, I have an SMB output and when I connect a cable I reinject noise on my board which alters my output signal because I interfere with my output signal from the DAC. 

I added a capa of 0.5pF on the feedback resistor R220 and a series resistance of 49.9 and a capa of 1uF just before the SMB connector, and I increased my resistance value for generating the gain (initially R186 and R219 was from 499 and pass to 4k99 and R220/R198/R126/R187 was from 4k99 and pass to 49k9),  
this improves the situation a little but it does not seem sufficient to me.
I also have to redo the general routing of the function but first I would like to be sure that I have not missed something...
Currently I also have a problem because I should have an amplification of 10 compared to my signal in DAC output and I have rather an x6 I don't understand why, I have to change the AD8021 to check that I haven't damaged it...
Thank you in advance for your response.

Best Regards.
 

Thread Notes

Parents
  • Hi,

    According to the datasheet the resistor values around the AD8021 are much higher than recommended. Follow the guidelines given in the APPLICATIONS section precisely.
    On the other hand, the ADG1401 adds large unwanted capacitances to the circuit. Cs, Cd in the off and on state are well beyond 60pF, so this can certainly affect the amplifier behavior and may make it unstable and sensitive (there is a capacitive load at the inverting input due to pin1 of the ADG1401). This together can also reduce the gain at higher frequencies. 


    So, I recommend to have feedback resistor value close to 1k and to remove the analog switch. Implement a fixed gain of 10 block and if required, halve the signal using the DAC input data only. It is a 14 bit dac, you only lose a single bit by halving the output signal.

    It could be also useful to make an LTSpice simulation to check the amplifier part of your circuit.

    Zoltan

  • Hi  

    As what Zoltan said,  you can refer to the recommended component values on the datasheet. [See table 6]

    Can you try to adjust resistor values, remove the switch and observe the results?

    Thanks,

    Gio

  • Hi,

    Thank you for your feedback.

    The reduction of the gain resistor value improves the problem, however the output of the AD9102 being limited in current, it forces you to use two followers amp or make an amp + amp diff setup. So I'm also looking at INA849 instrument amps that integrate this function.

    thnaks

  • Hi,

    The INA849 has a bandwidth of 13MHz at a gain of 10, so won't be enough, if you wish to have upper corner frequency of several tens of MHz, as you said.

    Best wishes, Zoltan

  • Hi,

    You are right, after testing the INA849 is extremely limiting. So I switched to a 3 AOP assembly with a first amplification stage with a common gain resistor (*2) and a second differential stage with a second gain to obtain *10. Thank you all for your feedback.

Reply Children
No Data