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.

Video amplifier selection for DM814x and AM387x with integrated reconstruction filter.

Other Parts Discussed in Thread: LMH6714, LMH6715, LMH6738, AM3874, OPA361, THS7315, THS7375

Hi all,

Could you please help me to select appropriate video amplifier for DM814x or AM387x ?

DM814x and AM387x has internal video amplifier but I need to use an external one. In this case full scale output is 0.7V and according to data sheet load resistor must 1500 ohm. So required gain stated as 3.7V/V in datasheet.

I could not able to find an approprite TI video amplifier. There are some amplifier designer for other Davinci processor but they designed for 0.5V level and has an internal load resistor integrated.

Best regards.

  • Are you looking for a video buffer?  We have some LMH current feedback amplifiers that are good for high resolution video.

    The LH6714 family offers good performance.  If you use a 1500 Ohm resistor on your CPU you can use the non inverting input of one of these amplifiers. 

    Does your gain calculation include loss due to the termination resistor?  I would presume you want to drive a 75 Ohm transmission line. 

    http://www.ti.com/lit/ds/symlink/lmh6714.pdf (single and quad)

    http://www.ti.com/lit/ds/symlink/lmh6715.pdf (dual)

    http://www.ti.com/lit/ds/symlink/lmh6738.pdf (triple)

  • Hi Loren,

    Thank you for your answer and time.

    I want to implement the block given in figure below. (by amplifier with integrated reconstruction filter)


    There is some issues that I have confused;

    Gain is defined as of 3.7V/V by datasheet. I have asked a TI expert why the gain is 3.7V/V and he told me that;

    DM8148 or AM3874 DAC will source current into a load resistor (R-load).  With the full-scale curent (470uA) and R-load (1500 ohms) specified in the DM, gain of 3.7V/V should provide a 1.3V video range after video termination (470uA x 1500 x 3.7  /  2).  

    1. I could not understand why video output level is adjusted 1.3V instead of 1V as in standard ?

    2. The amplifiers you adviced do not include integrated reconstruction filter. Is there any other alternatives or could you reference a document for implementing external reconstruction filter ?

    Best Regards.


  • I will try to help, but these parts are not really my specialty.

    1.  Video is 1Vpp, but that is 0.7V positive (video data) and 0.3V negative (sync pulses); see reference below.  I am not sure about the 1.3V, but usually video signals are amplified to 2x the desired voltage so that after the 75 Ohm matching resistor there is the correct signal at the 75 Ohm load.  

    2.  No, the amplifiers I recommended do not have reconstruction filters.  I am new to TI, so I am not expert on the entire product line yet. 

    http://www.ni.com/white-paper/3020/en

    Video = +0.714 to -.286 V = 1.0Vpp

  • The reason for 1.3Vpp is for 100% saturated CVBS signal support. CVBS signals combine the Luma (Y) and the Chroma (C) together. When 100% color saturation signals are supported, then when you go thru the math in the standards, you will see that the potential is for the 2 signals may combine to be as high as 1.3Vpp (technically about 1.2335Vpp for PAL and 1.221Vpp for NTSC, but many people like to round off numbers). This does include the 0.3Vpp for the Sync.

    Keep in mind that these numbers are for the video signal as it should appear to a display. Due to double termination, the amplifier must support 2X these values at the amplifier output - or about "2.6Vpp".

    One solution for CVBS only with a simple integrated filter is to use the OPA361 for a single channel, the THS7315 for 3-channels, or the THS7375 for 4-channels. These parts have much higher gain than what you are specifically looking for, but there is a very simple solution to make the "system" gain anything you like. Please see the THS7375 datasheet - pages 27 and 28 for the details as it is a little cumbersome to post the info here. But the solution simply takes one more resistor to accomplish the gain you are looking for. This is applicable to any of the parts mentioned - just account for the slight gain differences of each part and the termination resistance the DAC requires.

    Cheers,

    Randy

  • Hi Loren,

    Thank you for your responses and time.

    Randsy S. has clarified the situation below.

    Best Regards. 

  • Hi Randy,

    Thank you for sharing your experiences and time.

    The guidance on THS7375 document fits my system requirents. 

    But I am new to video circuits, and there is one thing that is not clear yet for me;

    How the video signal amplitude precision is important ?

    The circuit I design can be used for both NTSC and PAL. And also as you stated gain is raugh.

    So DC level can be changed up to 6%. (1.221 to 1.3) By adding the other error budged (DG,DP, 3%) DC level can vary 10%.

    While Luma is directly related with DC level, how the receiver system restores DC level.

    Do I need to care for precise DC level output (gain) or video receiver uses AGC, or SW solution for DC level restoration.

    Best regards. 

  • If you go by SMPTE170 standard (NTSC), this states the recommended tolerance shall be +/-1IRE (7mV). But, it is "recommended" only and not a hard spec. Other documents also state roughly this 1IRE point. However, in consumer systems, this is not held to such a tight tolerance (1%). It depends on what specifications your system needs, but I have seen this tolerance be up to 5%.

    The reason this is not usually an issue is that modern day Displays use decoder which they usually have AGC circuits built-in. The AGC takes care of any gain issues.

    These decoders also require an input coupling capacitor. This allows the decoder to set it's own DC reference irregardless of the signal itself. So it is not critical to set the DC operating point of your signal other than meet EIA/CEA specs of having the backporch be at 0V+/-1V at the display input.

    My recommendation is to have the gain set to produce the SYNC amplitude of 0.3Vpp +/-1%.

    Cheers,

    Randy

  • Hi Randy,

    Thank you very much for sharing your experiences with us.

    You have clarified all issues that I could not able to be sure.

    Best Regards.

  • 8147: I tried an OPA361, Rset at 4K7, TV bypass mode so TV_VFB0 drives op amp with no resistor to ground or series.
    OPA361 has 500 ohms internal that the dac drives. With 75 out r and 75 load r you get ~1.2 p-p video.
    Hope this helps - Patrick