• State Resolved
  • Locked Locked
  • Replies 4 replies
  • Answers 1 answer
  • Subscribers 46 subscribers
  • Views 792 views
  • Users 0 members are here
Support feedback
Options
Options
Similar topics

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.

DAC3152 output level low

Johann Gerber
Prodigy 60 points
Other Parts Discussed in Thread: DAC3152, TRF370333

Hi there,

We are using the DAC3152 to drive a modulator (ADL5375) via a LPF. The interface between the is designed using the SLAA399 spreadsheet example to give

a DC offset at the DAC output and the modulator of 0.5 V with a single ended p-p voltage of 0.5 V.

For some reason we only see ~50 mV p-p as we generate a SSB signal at 4 MHz. Clock rate is 160 MHz.

I measured the bias resistor value (setting the maximum current) and it was populated correctly (~960R) which agrees with the designed 20 mA.

Any other ideas?

Thanks,

Johann

  • 0
    TI__Genius 12965 points

    Johann,

    There are a couple of issues I can see from your design.  The circuit as it is will not work properly.  The DAC to Modulator design requires meeting several criteria.  The common mode voltage of the DAC and Modulator needs to be satisfied.  The load will determine the compliance voltage, and the level shift network along with the filter termination will also cause an attenuation.

    The current network you have which shows 3.3V pull up with 162 ohms , 0 ohm step down, and then 71.5 ohms to ground will create a 0.5V common mode on the DAC pin as well as the modulator input pins.  This is a serious problem as the compliance voltage of the DAC is between 2.8V and 3.8V.  Typically this means the expected common mode voltage for the DAC is at 3.3V with a +/-0.5V swinig.  With a common mode at 0.5V for the DAC, the DAC output would not work well.  You can verify the common mode by using a volt meter.

    For this particular design you should be looking at section 4.11 of the SLAA399 app note.  This outlines how to set up this kind of network.  To get 3.3V at the DAC and 0.5V at the IQ modulator you will need to have a pull up voltage more than 3.3V, typically we use 5V and GND, and then the R1=66, R2=177, R3=32.  This would setup the appropriate common mode voltages, but the step down resistor R2 & R3 would also create a -16dB  attenuation.

    You mention a spreadsheet example.  Can you show me the result from that tool?  I have included a screen shot of my calculation as well as a TINA simulation showing your network and the one that is calculated.  I have a basic model of the current sink DAC (not specific to DAC3152), interface network, as well as a LPF in my model.

    E2E_DAC3152_ADL5375.docx

    Ken.

  • 0 in reply to Ken C
    Prodigy 60 points

    Hi Ken,

    Thanks for the feedback. I missed that the common mode voltage on the DAC had to be ~ 3.3V.

    Below is a screen shot with a common mode of 3.3V. I adjusted the DAC output pin load to achieve 100R diff impedance on the filter and modulator.

    Something that is not so clear to me, is the different tabs / names used on the spreadsheet: "unmatched impedance" / "Matched impedance" / "Matched Impedance pseudoDC". The latter is the image that I included and gives the voltages without attenuation.

    Any error in using it as such?

    Thanks,

    J

  • 0 in reply to Johann Gerber
    TI__Genius 12965 points

    Johann,

    Typically we use the matched impedance version since it makes filter design easier as the filter sees the same impedance on in and output of the filter.

    Your xls look incorrect - I wonder if it is corrupted.  Just looking at the resistor drop from R2 to R3//(R4/2) already shows me that the attenuation cannot be 0.  Also the load on the DAC is too high and would exceed the compliance voltage of the DAC output/  Theres definitely something wrong with the xls you have.

    I used the same inputs on my calculator and it gave different results, but it does lineup with a simple TINA simulation showing the correct DC bias points - although it wuold still violate the compliance voltage as the DAC load is quite large and the 20mA will cause the swing to exceed the limits.  I have attached the xls I used.  Try it out.

    Ken.

    Your calculations:

    DAC_3p3V_passive_interface_calculator_KC_external.xls

  • 0 in reply to Ken C
    Prodigy 60 points

    Thanks Ken,

    The attenuation looked wrong, because the cap accross R2 was not shown (pseudo DC). Unfortunately we do use a DC output, so the

    AC short circuit will not help.

    I will revisit the complete DAC / modulator selection. In the mean time I will order some TRF370333's to recover the low gain and test the remainder of the circuit.

    They are pin compatable of the NC's can be connected to GND. If not, I will have the hack the board to fit them.

    Thanks again,

    J