AD2S1205 startup issue

Hello, I'm testing a new design with a AD2S1205 resolver decoder. I have my design assembled, and applied 5v to the Vdd. I have been using a 100 ohm resistor to pull down the reset signal by hand, and when I remove it I observe the current draw by the 5v rail to go from <10mA to >10mA, so I believe that to be working. In addition, I see the voltage at EXC and !EXC as well as the clock input go from 0v to ~2.5v. However, I can't seem to find any other indication that the AD2S1205 is operational. I assume that the complimentary EXC signals should be generating a 10khz signal, and that the IC should be trying to oscillate the 8.192mhz crystal, but neither seems to be the case.

I suspected that the IC was damaged, and replaced it with a new one, but the new one behaves the same way.

What behavior should I look for to indicate that the device works? What do I need to do to get the IC to excite the resolver coils?

Parents
  • 0
    •  Analog Employees 
    on Dec 21, 2019 6:25 PM 10 months ago

    Jackson,

    Can you tell me a little bit more about your application in terms of how you've connected the 1205 to the resolver?

    Please note that Analog Devices offices will be closed from 12/23/2019-01/06/2020 and thus responses may be delayed.   Our apologies in advance for any inconvenience.

    Happy Holidays

    Sean

  • My assumption is that it shouldn't matter that the IC is or isn't attached to a resolver, and that the IC should exhibit some kind of activity if it's functional, perhaps a signal from the EXC pins or otherwise. Regardless, I'll attach an image of the schematic:

  • 0
    •  Analog Employees 
    on Dec 23, 2019 1:21 PM 10 months ago in reply to JacksonG

    Jackson,

    I apologize but I misinterpreted your original message, I had incorrectly assumed the EXC/EXCN pins were loaded by the resolver but I missed that CLKIN was not toggling..   As the clock is not starting up (CLKIN) you can expect that EXC, EXCN will also not wiggle as their is no clock to advance the internal circuits.    

    The reason your clock is not oscillating is that the conditions for oscillation have not been established or at least you don't have conditions for sustained oscillation.   Unfortunately debugging this is going to be related to the oscillator you've chosen but let me try to give you some suggestions on how to get started debugging this.

    1) Verify the values of C1 and C2 are correct for the chosen Y1.   So much of oscillation is ensuring there is sufficient gain and phase around the oscillatory loop and it's important to follow the crystal manufacturer's recommendations.   

    2) Make sure that you've cleaned up all residual solder flux and contaminants from the surface of the board that might be causing leakage paths.

    3) Try placing a 1 Megaohm resistor across XTALOUT and CLKIN as a ballast resistor.  This will place the internal inverter in a ideal condition for oscillation on start-up.

    4) Try inserting a small value resistor (you'll need to play with this value experimentally) between XTALOUT and C1/Y1.  This may help regulate the current into the crystal to help with oscillation.

    Hope this helps.

    Sean

Reply
  • 0
    •  Analog Employees 
    on Dec 23, 2019 1:21 PM 10 months ago in reply to JacksonG

    Jackson,

    I apologize but I misinterpreted your original message, I had incorrectly assumed the EXC/EXCN pins were loaded by the resolver but I missed that CLKIN was not toggling..   As the clock is not starting up (CLKIN) you can expect that EXC, EXCN will also not wiggle as their is no clock to advance the internal circuits.    

    The reason your clock is not oscillating is that the conditions for oscillation have not been established or at least you don't have conditions for sustained oscillation.   Unfortunately debugging this is going to be related to the oscillator you've chosen but let me try to give you some suggestions on how to get started debugging this.

    1) Verify the values of C1 and C2 are correct for the chosen Y1.   So much of oscillation is ensuring there is sufficient gain and phase around the oscillatory loop and it's important to follow the crystal manufacturer's recommendations.   

    2) Make sure that you've cleaned up all residual solder flux and contaminants from the surface of the board that might be causing leakage paths.

    3) Try placing a 1 Megaohm resistor across XTALOUT and CLKIN as a ballast resistor.  This will place the internal inverter in a ideal condition for oscillation on start-up.

    4) Try inserting a small value resistor (you'll need to play with this value experimentally) between XTALOUT and C1/Y1.  This may help regulate the current into the crystal to help with oscillation.

    Hope this helps.

    Sean

Children
  • I'm using a 8.192mhz crystal with two 20pF capacitors on C1 and C2, which I believe the datasheet reccomended. I can try using a balast resistor across the crystal, but putting one in series with the xtal out pin may be difficult. 

    The datasheet implies that the crystal isn't required for the chip to function. "To achieve the specified dynamic performance, an external crystal is recommended at the CLKIN and XTALOUT pins". Is this the case? If so, shouldn't I be seeing activity at other signals on the chip, such as CPO or the two EXE pins?

  • 0
    •  Analog Employees 
    on Dec 27, 2019 3:17 PM 9 months ago in reply to JacksonG

    Jackson,

    The wording of the datasheet may be a bit misleading as an external clock is required.  Your options are a parallel resonant crystal as recommended or you could also use a oscillator and a clock generated from the microcontroller.   We recommend the crystal due to the stability and low jitter of the implementation which leads a more stable result.

    Just for clarification the values of C1 and C2 should follow the crystal manufacturers recommendation and not what is in our documentation.   This may be what you were referring to but I just wanted to make sure we're on the same page.  For my reference would you mind sharing the part number on the crystal you are using?

    Sean