This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

ADS5483 EVM with DC coupled analog inputs

Other Parts Discussed in Thread: ADS5483, THS4509

Hi,

I recently got the ADS5483 EVM. I would like to use the Analog Input Option 2 with DC coupled signal.

I followed the instruction for DC coupled operation at e2e.ti.com/.../1399346
- unsolder and resolder SJP4, SJP1 and SJP2.
- remove both transformers and jump the signal path across the primary and secondary side.
- solder 0 Ohms resistors on all the series elements of the filter.
- remove the C1 and C4.

The common mode input voltage is set to 2.5V from the ADC. The THS4509 is powered with single +5V.
SMA J2 is not installed.

Here is what I observed:
Looking at AMP OUT+ and AMP OUT- for 0 V apply to SMA J1 they are at 2.1V and 2.8V respectively. I would expect them to be at 2.5V. When I send a sinewave (0.5Vptp, offset 0V, 100 kHz) I see half of the sinewave cycle on both AMP OUT+ and AMP OUT- with the remaining part of the cycle at DC level 1.9V and 2.9V respectively.

What am I doing wrong?

Thanks for your help.

Mathieu.

  • Hi,

    A couple of things I noticed and wanted to ask about or point out:

    Yes, you would remove the transformers and shunt across with zero ohm resistors if you want DC coupled setup, such as you would need if you need to include a bandwidth all the way down to DC for your input signal.  Do you need a DC coupled system?  I presume so since that is how you are setting up the system.  You mention removing C1 and C4 so I presume you are also putting zero ohm elements down in those positions too.

    The VCM of the ADC is not 2.5V, it is 3.1V.  That is not well centered for the amp unless you are not pushing to full scale. The reason the power banana jacks for the amp do not connect the lower supply to the same ground as the rest of the EVM is that we have had customers offset the amp power supplies a bit - like 6V for the positive rail and 1V for the lower rail.  That is a bit ugly, needing these extra supplies, but that has been done to get better balance and headroom for a full scale signal.   I believe you are seeing the signal at 2.5V because the zero ohm resistor R18 is not installed.  So the VCM from the ADC is not connected back to the amp.

    After that - if you are still having trouble getting the THS4509 to work properly then I may have to bring in support from our high speed amp group or possibly defer the whole amp discussion to their forum.   I'll help here where I can, but may need to bring in the amp people one way or the other.  Let's correct the common mode hookup to get 3.1V VCM and then see where we are.

    Regards,

    Richard P.

  • Hi Richard,

    Thanks for the answer.

    Yes I want a DC coupled system. I did put zero ohm element in place of C1 and C4.

    You are right, I did not installed R18, and 2.5V that I see is the midrange of the amp power supply.

    I think that if I want to use the full scale of the ADC with 5V single supply for the THS4509, I have to add a pull up resistor between Rg and Rf as discussed page 27, fig. 85 of THS4509 datasheet. This is not very convenient to do on the evaluation module. Please correct me if I am wrong.

    I ended up soldering SMA J2 and using an external differential amplifier with larger single power supply such that:

    - the mid way point correspond to 3.1V

    - I can use the full scale of the ADC.


    Could you please comment on that solution?

    Thanks for your help.

    Mathieu.

  • Hi,

    I have taken the ADS5483 schematics and marked it up with the modifications for DC coupling the amplifier and sent it to someone who supports our high speed amplifiers for comment.  He also now has the link to this forum posting directly if he chooses to respond here, rather than responding through me.

    It is also reasonable to install J2 and bring in a differential signal from an external amplifier EVM.  The thing to be careful about there is to match the lengths of the path for the two sides of the differential signal to avoid phase mismatch.  When I have used an external amplifier EVM into an ADC EVM (not this ADC EVM, though) I have often been able to use two right angle SMA-to-SMA barrel connectors on each side of the differential. (four total).  This way there is no cable matching needed, and the two right angle connectors allow for easily matching the SMA spacing on one EVM up to the SMA spacing on the other EVM.

    Regards,

    Richard P.

  • Hi Richard, Mathieu,

     

    I will reply here so that I get notification on the whole conversation.  So far it seems that Richard has offered the correct advice. 

    I think one more thing needs to be addressed,  With a single ended input the input of a differential amplifier is actively terminated.  I am looking at Figure 9 in the EVM user's guide on page 18.  I will give you the resistor values needed for a 50 Ohm input impedance on the Amp in- pin. (leaving R9 and R17 at 348 Ohms)

    R12: 88.7

    R14: 118

    R54: 0

    R53 (leave empty)

    Between Pins 4 and 5 on the transformer footprint (T2) solder a 73.2 Ohm resistor to provide the 50 Ohm input impedance at Amp IN-. 

    So that you can calculate additional gain configurations I am attaching a spreadsheet that does all of the active termination calculations (I'll have do do it in a subsequent post)

     FDA single to diff resistor calculations with BW and Noise calculations to share.xlsx