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.

TVP5151 power optimization

Other Parts Discussed in Thread: TVP5151

Hi,

We are using the TVP5151 for at video project.

It works fine, but we would like to optimize power consumption.

Since we are sampling and Interlaced signal (PAL), and MPEG encoding this video, we are only using one of the Fields (Field 1).

So is there any way to make the TVP5151 skip the other field?

I mean, it would be very nice, if it was possible to power down the ADC when sampling outside Field 1, because we are basically ignoring the data anyways.

Best regards

Jakob

  • Unfortunately no.

    BR,

    Steve

  • Okay thanks.

    We are currently measuring 175mW when decoding a standard PAL signal from a camera, which seems high, we where hoping for about 137mW.

    Is there any way to reduce power consumption?

    I'm thinking possible disabling Teletext, Macrovision engines ect.?

  • That seems a little high. The max should be about 150mW with nominal in the 115mW range.,

    Can you try disabling the IOs to see if we can determine where the power is going? If disabling the IOs reduces the power significantly then I would suggest looking at the capacitive load on the IOs.

    [Update] Sorry, the numbers above are for the TVP5150AM!. The TVP5151 should be about 137mW nominal, 185mW max. Can you measure the individual supply rail currents and compare to the datasheet ranges?

    You cannot turn off parts of the device independently, but the saving would be negligible anyhow.

    BR,

    Steve

  • Hi Steve,

    Thank you very much, we would really be happy, if we could get the power lowered to about 115mW.

    Okay, we will try to disable (high impedance) the YOUT port and SCLK output pin, writing 0x00 to register 0x03.

    I actually think we tried doing so earlier, which only reduced power consumption about 20mW.

    Are there any other outputs which should be disabled (high impedance)?

  • It is gonna be really hard to measure the supply rails, because they are used to power our video DSP ect.

    Would it be possible to power down the decoder at the end of each Field 1 and up again before Field 1 starts in next frame?

    I have played a little with the power down function, and it seems that the power down mode (bit 0 in register 0x02), it not all off, so I was hoping it was possible to use this register for powering down outside Field 1.

  • No, you cannot power down mid flight. The decoder is continuously tracking the video signal.

    What package are you using? If you are using the QFP then you could lift the leg of the device. Obviously this is not possible with the BGA version.

    If the supplies are not separate how are you determining how much power the TVP is consuming ?

    BR,

    Steve

  • Oh okay, we are using the BGA version.

    I'm just measuring the power fed into my board, and can see the raise in power when the TVP5151 is turned on.

    There is actually two TVP5151's, and they behave equally power wise.

    But what determines power consumption, other than temperature ?

    Is it dependent on input signal levels ect.?

  • Yes, the input signal can affect the power.

    How are you generating the required supply voltages for the TVP? If you are using linear regulators then don't forget that any voltage difference is wasted power burned entirely in the regulator.

    For example, if you are feeding 3.3V into a 1.8V regulator at 30 mA then 45mW is burned in the regulator.

    (3.3 - 1.8) * 30

    BR,

    Steve

  • Ok, maybe the input signal is wrong then I will do some more testing.

    We use dc-dc converters (step down), and they are about 90% efficient.

    I can actually see that the power is probably only about 20mW too high.

    Because we measure 170mW on the power input side, but if I multiply by the dc-dc converter efficiency , I get 170 * 0.9 = 153mW.

    If I disable the YOUT and SCLK, I get about 155mW, and if I multiply with 0.9, I get 155*0.9 = 140mW, which is very close to 137mW.

    So it is probably my YOUT and SCLK drawing the extra 20mW ?

    I read that you could download "patch code" to the tvp5151, but not how to make the code, or anything else?

    What is possible regarding patch code?

  • There isn't really a 'wrong' input signal but the amount of internal switching does depend on the video. The numbers in the datasheet are for a nominal color bar input but I would not expect the real world numbers to be significantly higher.

    OK, good on the DC-DC :-) Just wanted to check. Is the DC-DC really that efficient in the load region you are using it? 90% seems a little optimistic.

    IO switching will certainly consume a fair amount of power and is quoted independently in the datasheet. The capacitive loading on these signals will have an effect on the power requirements so reducing capacitance should help reduce the power if possible.

    Patch code is only for factory supplied bug fixes and performance updates/changes. There are no user accessible code patches.

    You can obtain patches from here...

    http://software-dl.ti.com/dsps/dsps_public_sw/dsps_swops_houston/ANALOG_VIDEO/Analog_Video_Decoder_Versions.htm

    BR,

    Steve

  • Would it be possible to have made a special patch code, which only transmits Field 1 onto the SCLK/YOUT port ?

    Hi Steve,

    Thank you very much for all your help, I have one more question.

    So tvp5151 would save as much power as possible from the end of Field 1 and all the way to the beginning.

    I'm looking into just sending the a 0x00 to register 0x03 via i2c and then again a 0x09 to register 0x03 when in Field 1.

    But It would be easier and a much more simple design, if the tvp5151 was capable of doing something like that automatically.

    Best regards

    Jakob

  • It is not possible to achieve what you are looking for unfortunately.

    BR,

    Steve