Post Go back to editing

(ADA4433-1) To connect ADA4433-1 to ADI Video decodor

Hi !

I have some questions about ADA4433-1.

******************************************************************************************

Q1)

Our customer are planing to make evaluation board to change "single-CVBS" to "Diff-CVBS" to input ADV7281A-M.

To connect to ADI video decoder,  I think AC-coupling is necessary.

At ADA4433-1 datasheet page.22 Figure 55, there is "Difference Amplifier in a DC-Coupled Configuration" reference circuit.

Do you have "Difference Amplifier in a AC-Coupled Configuration" reference circuit?

******************************************************************************************

Q2)

Our customer will use like this.

"CVBS source" => "Single-CVBS" => "ADA4433-1" => "Diff-CVBS" => "ADV7281A-M"

Do our customer have to do AC-Coupling between "CVBS source" and "ADA4433-1"?

******************************************************************************************

Q3)

Do our customer have to do AC-Coupling between "ADA4433-1 source" and "ADV7281A-M"?

******************************************************************************************

Q4)

At ADA4433-1 datasheet Page.15, this is written.

Does this mean whether AC-coupled or DC-coupled, we do not change the connection of -IN and we need to bias with half the amplitude of + IN?

******************************************************************************************

Q5)

At ADA4433-1 datasheet Page.15, this is written.

Our customer are planing to use these value for each components.

C1 = 22uF

R1 = 62kOhm

R2 = 27kOhm (1V bias

*Fc = 0.26Hz.

How should they decide these value for each components?

Do you think that there is no problem with these values?

How should the cutoff frequency of the high-pass filter be taken into consideration while considering the NTSC signal frequency? (Does sag do not occur?)
Since 1Vpp signal is supposed to be input, is it OK to understand that 1.0 V bias is OK?
When deciding each constant, is there a point to consider other than the cutoff frequency and the bias voltage?

******************************************************************************************

Best ragards

Kawa

  • Hi Kawa,

    I'm currently looking into your queries. I'll get back to you.

    May I ask what will be the end application for this?

    Regards,
    Goz

  • Hi

    Thank you for your reply

    Our customer will use ARV728x for rear camera

    But they don’t have equipment to evaluate diff cvbs

    So they are planning to make 400 units for single to differential cvbs equipment 

    Beat regards

    Kawa

  • Hi Goz

    Thank you for your support.

    When can you answer?

    Our customer want your answer till next Monday.

    Best regards

    Kawa

  • Hi Goz

    Our customer have to fix their PCB layout.

    So please give us your comments ASAP.

    Best regards

    Kawa

  • Hi Kawa,

    Apologies for replying late.

    Answering your first and third question you can refer on the Figure 22 of ADV7281 datasheet the AC coupling is necessary at the input of the device as a DC block. 

    AC coupling between the source and ADA4433-1 is essential, this ensures that any other DC bias by the CVBS source is blocked or removed before the amplifier. The offsets in the input amplifier should then be biased say for example in your application you have 1Vpp signal, the offset should be as what you have said in question number 4 half the amplitude of the signal in +IN. You can use the default bias recommended in the datasheet which is R3 = 7.5kΩ and R4 = 1.33kΩ. It should be observed that the dc voltage bias at the inputs are equal so that the offset on the differential output of the amplifier will be on the same level, otherwise it will drift apart. In a scenario where your input signal is 2Vpp, the offsets on each input should be 1V, we can compute for the values of the resistors based on what is suggested on the datasheet.

     

    This will give us R3 = 3.059kΩ.

    I think the values of your components are too large compared to the input impedance of the device. I'd recommend you use the prescribed value in the datasheet where its value is near negligible for loading. Compute the combination for the sweet spot between loading effect and the capacitance value to be used. The resistor values of the biasing circuit should be equal for proper matching.

    Let me know if you have clarifications.

    Regards,
    Goz

  • Hi Goz

    Thank you for your reply!

    Answering your first and third question you can refer on the Figure 22 of ADV7281 datasheet the AC coupling is necessary at the input of the device as a DC block. 

     

    => I understood that our customer should use AC-coupling between ADA4433-1 and ADV728x.

    AC coupling between the source and ADA4433-1 is essential, this ensures that any other DC bias by the CVBS source is blocked or removed before the amplifier.

    => I understood that our customer should use AC-coupling between CVBS source and ADA4433-1.

    It should be observed that the dc voltage bias at the inputs are equal so that the offset on the differential output of the amplifier will be on the same level, otherwise it will drift apart.

    => Sorry I couldn't understood what you said.

         " the dc voltage bias at the inputs are equal "

         Q1) What do you mean equal? Equal to something?

         " It should be observed that the dc voltage bias at the inputs are

           equal so that the offset on the differential output of the amplifier

           will be on the same level, otherwise it will drift apart. "

         Q2)Do you mean that the output offset will be equal to input bias voltage so we

               have to understand that maybe output offset level will drift little because

               the bias level is made by divide resister and the resister value will change by temperature?

    In a scenario where your input signal is 2Vpp, the offsets on each input should be 1V, we can compute for the values of the resistors based on what is suggested on the datasheet.

     

    This will give us R3 = 3.059kΩ. 

    => I understood.

    I think the values of your components are too large compared to the input impedance of the device. 

    => Q3)Which components are too large? C1?

    The resistor values of the biasing circuit should be equal for proper matching.

    => Q4)How do our customer decide resistor value of biasing which should be equal for proper matching?

               

         Q5)I suppose that AC coupling is necessary for ADA4433-1. 
                Can you tell me the recommended value of each value in the attached circuit diagram?

                Should our customer make like this?(R1 = R3, R2 = R4)

    Best regards

    Kawa

  • Hi Kawa,

    Q1) What do you mean equal? Equal to something?

    What I meant is that the offset levels on each input pins +IN and -IN should be on an equal dc level.

    Q2)Do you mean that the output offset will be equal to input bias voltage so we have to understand that maybe output offset level will drift little because the bias level is made by divide resister and the resister value will change by temperature?

    In this question you can refer to the figures below.

    The output common mode voltage (Vocm) at the output is set at the midpoint supply which is 1.65V. As seen if the negative input is biased half of the +IN voltage, the voltage will now then extend to only 2.15V and 1.15V which is more efficient than the first. 

    You are correct for the resistor divider, to avoid such drift please use a precise and low thermal coefficient resistor.

     Q3)Which components are too large? C1?

    The resistors for biasing, such values will introduce loading effects on the input impedance of the amplifier.

    Q4)How do our customer decide resistor value of biasing which should be equal for proper matching?

    Q5)I suppose that AC coupling is necessary for ADA4433-1. Can you tell me the recommended value of each value in the attached circuit diagram? Should our customer make like this?(R1 = R3, R2 = R4)

    Yes the resistor values should be R1 = R3, R2 = R4.

    Actually, if your CVBS source is what have you said is 1Vpp, and do not vary AC coupling is not necessary. You can just put a dc bias circuit the same on the Figure 53 on the datasheet.

    But if the CVBS source varies in magnitude AC coupling is necessary and we have to compute for the resistor values for biasing and the value of the capacitor for the proper bandwidth.

    I'm currently contacting the product expert for the proper consideration for input loading and bandwidth. I'll get back to you tomorrow.

    Regards,
    Goz

  • Hi Goz

    Thank you for your reply.

    I understood.

    I will wait your additional answer.

    Actually, if your CVBS source is what have you said is 1Vpp, and do not vary AC coupling is not necessary. You can just put a dc bias circuit the same on the Figure 53 on the datasheet.

    Do you mean that our customer don't need to do AC coupling between CVBS source and ADA4433-1 if GND is same?

    What will happen if CVBS source output is 1Vpp but 0V bias?

    Also please give me some advice for these question too.

    How should the cutoff frequency of the high-pass filter be taken into consideration while considering the NTSC signal frequency? (Does sag do not occur?)
    Since 1Vpp signal is supposed to be input, is it OK to understand that 1.0 V bias is OK?
    When deciding each constant, is there a point to consider other than the cutoff frequency and the bias voltage?

    Best regards

    Kawa

  • Hi Goz

    Our customer have to fix their circuit till tomorrow.

    Can you give some advice for ADI recommend circuit for single to differential CVBS exchange and value for each components till tomorrow?

    Can our customer use Aluminum electrolytic capacitor for AC-coupling?


    Best regards

    Kawa

  • Hi Kawa,

    Do you mean that our customer don't need to do AC coupling between CVBS source and ADA4433-1 if GND is same?

    What I meant here is if the output of the CVBS has known DC level is ground referenced like on the Figure 53 in the datasheet which is an output from the DAC, you can just use DC coupling. 

    What will happen if CVBS source output is 1Vpp but 0V bias?

    The input voltage range of ADA4433 is 0 - 2.1V, if your signal is 0V biased the h-sync will be clipped.

    Since 1Vpp signal is supposed to be input, is it OK to understand that 1.0 V bias is OK?

    Yes, 1V bias is fine but not because your input signal is 1Vpp, you are just raising the biasing of your CVBS signal that will ensure that the sync tip will not be clipped at the input of the amplifier due to the input voltage range.

    When deciding each constant, is there a point to consider other than the cutoff frequency and the bias voltage?

    You also need to consider sag. 

    How should the cutoff frequency of the high-pass filter be taken into consideration while considering the NTSC signal frequency? (Does sag do not occur?)

    The location of the cutoff frequency should be set properly. It doesn't need to be 0.26Hz up to 4Hz I guess is fine depending on the refresh rate. 

    For the values of your components, try lowering it to R1 = 6.2kOhm, R2 = 2.7kOhm, your capacitor value is fine. Just make sure that R1 = R3, R2 = R4, and the capacitors should be equal too.

    Let me know if you need any clarifications.

    Regards,
    Goz