Author Topic: Clamping DC Biased Composite Video Signal Back to ground  (Read 2793 times)

0 Members and 1 Guest are viewing this topic.

Offline syntax333Topic starter

  • Regular Contributor
  • *
  • Posts: 158
  • Country: 00
Clamping DC Biased Composite Video Signal Back to ground
« on: March 26, 2019, 12:28:54 am »
Hi, as the title suggests I want to clamp the DC Biased Composite Video Signal back to its original position.

I have tried to clamp the signal by extracting sync information and drive a switch which connected to ground according to sync info.

The DC bias which I added is approximately 5V. After trying clamp the signal back to ground DC bias drops to 2V. I want the signal at its original position which is 0V.

Are there any suggestions? Thank you.
 

Offline xavier60

  • Super Contributor
  • ***
  • Posts: 3034
  • Country: au
Re: Clamping DC Biased Composite Video Signal Back to ground
« Reply #1 on: March 26, 2019, 01:15:29 am »
I have a circuit that will clamp the sync tip to 1V, you want to clamp to 0V?
Can you show us your circuitry?
HP 54645A dso, Fluke 87V dmm,  Agilent U8002A psu,  FY6600 function gen,  Brymen BM857S, HAKKO FM-204, New! HAKKO FX-971.
 

Offline syntax333Topic starter

  • Regular Contributor
  • *
  • Posts: 158
  • Country: 00
Re: Clamping DC Biased Composite Video Signal Back to ground
« Reply #2 on: March 26, 2019, 02:57:45 pm »
The circuit I used is shown below. I want to clamp to 0V.
 

Offline xavier60

  • Super Contributor
  • ***
  • Posts: 3034
  • Country: au
Re: Clamping DC Biased Composite Video Signal Back to ground
« Reply #3 on: March 26, 2019, 03:29:06 pm »
The CSOUT pulses are positive going. They should not be inverted to drive the 4066.
DC restored video signal doesn't tolerate much loading.

EDIT: I was wrong about the CSOUT polarity.
« Last Edit: March 26, 2019, 04:06:08 pm by xavier60 »
HP 54645A dso, Fluke 87V dmm,  Agilent U8002A psu,  FY6600 function gen,  Brymen BM857S, HAKKO FM-204, New! HAKKO FX-971.
 

Offline dzseki

  • Frequent Contributor
  • **
  • Posts: 524
  • Country: hu
Re: Clamping DC Biased Composite Video Signal Back to ground
« Reply #4 on: March 26, 2019, 03:32:35 pm »
You can use a DC coupled opamp (Au=1 or 2 depending on what output termination you want ), but with DC bias applied to the inverting pin, therefore subtracting that bias from the input signal.
HP 1720A scope with HP 1120A probe, EMG 12563 pulse generator, EMG 1257 function generator, EMG 1172B signal generator, MEV TR-1660C bench multimeter
 

Offline xavier60

  • Super Contributor
  • ***
  • Posts: 3034
  • Country: au
Re: Clamping DC Biased Composite Video Signal Back to ground
« Reply #5 on: March 26, 2019, 03:49:59 pm »
I think I'm wrong about the CSOUT pulses. So maybe there is a loading problem.
« Last Edit: March 26, 2019, 03:52:24 pm by xavier60 »
HP 54645A dso, Fluke 87V dmm,  Agilent U8002A psu,  FY6600 function gen,  Brymen BM857S, HAKKO FM-204, New! HAKKO FX-971.
 

Offline xavier60

  • Super Contributor
  • ***
  • Posts: 3034
  • Country: au
Re: Clamping DC Biased Composite Video Signal Back to ground
« Reply #6 on: March 26, 2019, 04:03:31 pm »
If there is 5V bias on the input, that would apply the wrong polarity to C3.
HP 54645A dso, Fluke 87V dmm,  Agilent U8002A psu,  FY6600 function gen,  Brymen BM857S, HAKKO FM-204, New! HAKKO FX-971.
 

Offline syntax333Topic starter

  • Regular Contributor
  • *
  • Posts: 158
  • Country: 00
Re: Clamping DC Biased Composite Video Signal Back to ground
« Reply #7 on: March 26, 2019, 04:15:17 pm »
Circuit works fine when I give a ref voltage instead of ground in order to shift a CVBS signal to ref Voltage therefore I don't think that CSOUT logic is wrong.

At the moment no load is connected to Video_out.

I will invert the polarity of C3 you are right.

Also I might apply negative voltage instead of ground to get rid of remaining 2V.
 

Offline xavier60

  • Super Contributor
  • ***
  • Posts: 3034
  • Country: au
Re: Clamping DC Biased Composite Video Signal Back to ground
« Reply #8 on: March 26, 2019, 04:19:31 pm »
If C3 has the wrong polarity applied to it, it will likely to be leaking current.
HP 54645A dso, Fluke 87V dmm,  Agilent U8002A psu,  FY6600 function gen,  Brymen BM857S, HAKKO FM-204, New! HAKKO FX-971.
 

Offline Benta

  • Super Contributor
  • ***
  • Posts: 6420
  • Country: de
Re: Clamping DC Biased Composite Video Signal Back to ground
« Reply #9 on: March 26, 2019, 05:20:39 pm »
The circuit is fine, but you need to watch the polarity of C3. You can get around this by using 2 x 220 uF electrolytic back-to-back, then you have a bipolar cap.
The output needs buffering. I can highly recommend the TI LM6171, it's the best video amp I've worked with.

I'd place R1 after C3 and change it to 75 ohm. Then you'll never have to think about DC biasing again.
« Last Edit: March 26, 2019, 05:22:24 pm by Benta »
 

Offline syntax333Topic starter

  • Regular Contributor
  • *
  • Posts: 158
  • Country: 00
Re: Clamping DC Biased Composite Video Signal Back to ground
« Reply #10 on: March 27, 2019, 07:08:03 pm »
circuit doesn't seem to work for biased CVBS signal. LM1881 does not trigger as it suppose to. Is this because bias?
 

Offline xavier60

  • Super Contributor
  • ***
  • Posts: 3034
  • Country: au
Re: Clamping DC Biased Composite Video Signal Back to ground
« Reply #11 on: March 27, 2019, 07:21:00 pm »
The bias shouldn't matter because C1 would block the DC. Can you show us the input video waveform? What is the p-p and polarity?
HP 54645A dso, Fluke 87V dmm,  Agilent U8002A psu,  FY6600 function gen,  Brymen BM857S, HAKKO FM-204, New! HAKKO FX-971.
 

Offline syntax333Topic starter

  • Regular Contributor
  • *
  • Posts: 158
  • Country: 00
Re: Clamping DC Biased Composite Video Signal Back to ground
« Reply #12 on: March 27, 2019, 07:56:51 pm »
Since my CVBS signal is biased 5V and it is 2Vpp overall voltage is approximately 5-7V. I just checked the datasheet of LM1881 which states that for Vcc=5V Input can 3Vpp. So  I AC coupled the input and buffered then send to lm1881. Now lm1881 triggers correctly.

However, as a buffer I use pnp follower configuration, output of my buffer seems noisy, this causes lm1881 to trigger wrong sometimes.

I want to use dedicated buffer (opamp) for cvbs.

I have only "MAX457CPA+"or "LMH6704MA" as a buffer. I have never used one before which one do you suggest to use?

Unfortunately, I don't have LM6171 at the moment.
 

Offline xavier60

  • Super Contributor
  • ***
  • Posts: 3034
  • Country: au
Re: Clamping DC Biased Composite Video Signal Back to ground
« Reply #13 on: March 27, 2019, 08:02:59 pm »
I don't know, I have only ever used  discrete circuitry for amplifying, sync separating and clamping.
« Last Edit: March 27, 2019, 08:04:50 pm by xavier60 »
HP 54645A dso, Fluke 87V dmm,  Agilent U8002A psu,  FY6600 function gen,  Brymen BM857S, HAKKO FM-204, New! HAKKO FX-971.
 

Offline syntax333Topic starter

  • Regular Contributor
  • *
  • Posts: 158
  • Country: 00
Re: Clamping DC Biased Composite Video Signal Back to ground
« Reply #14 on: March 27, 2019, 08:07:50 pm »
Your circuit seems interesting I have never tried to extract the sync information with discrete circuitry.
 

Offline xavier60

  • Super Contributor
  • ***
  • Posts: 3034
  • Country: au
Re: Clamping DC Biased Composite Video Signal Back to ground
« Reply #15 on: March 27, 2019, 08:13:43 pm »
Sync separation is easy so long as not too much low frequency content has been lost from the video signal.
The sync tips are clamped at 1V. It cant be set to 0V.
HP 54645A dso, Fluke 87V dmm,  Agilent U8002A psu,  FY6600 function gen,  Brymen BM857S, HAKKO FM-204, New! HAKKO FX-971.
 

Offline xavier60

  • Super Contributor
  • ***
  • Posts: 3034
  • Country: au
Re: Clamping DC Biased Composite Video Signal Back to ground
« Reply #16 on: March 27, 2019, 08:25:10 pm »
Maybe a small MOSFET could be used for clamping. Something like a 2n7002 or lower rated to keep the capacitance low.
HP 54645A dso, Fluke 87V dmm,  Agilent U8002A psu,  FY6600 function gen,  Brymen BM857S, HAKKO FM-204, New! HAKKO FX-971.
 

Offline IconicPCB

  • Super Contributor
  • ***
  • Posts: 1564
  • Country: au
Re: Clamping DC Biased Composite Video Signal Back to ground
« Reply #17 on: March 27, 2019, 10:18:39 pm »
A simple one transistor clamp.

One side of capacitor is clamped to ground the other ( input ) floats at an offset.
 

Offline xavier60

  • Super Contributor
  • ***
  • Posts: 3034
  • Country: au
Re: Clamping DC Biased Composite Video Signal Back to ground
« Reply #18 on: March 27, 2019, 10:27:46 pm »
I tried unidirectional clamping years ago. I found that it didn't cope well with scenes that contained large brightness difference  between top and bottom.

Extra: A good test of the performance of sync and clamp circuits is to put an undersized capacitor in series with the input video cable like 47uF. This causes large fluctuations in sync tip level with various video content.
« Last Edit: March 27, 2019, 10:35:48 pm by xavier60 »
HP 54645A dso, Fluke 87V dmm,  Agilent U8002A psu,  FY6600 function gen,  Brymen BM857S, HAKKO FM-204, New! HAKKO FX-971.
 

Offline syntax333Topic starter

  • Regular Contributor
  • *
  • Posts: 158
  • Country: 00
Re: Clamping DC Biased Composite Video Signal Back to ground
« Reply #19 on: March 28, 2019, 10:26:52 pm »
I manage to solve the problem. It was broken decoupling cap at sync separator LM1881. After I replaced it, LM1881 triggers as it suppose to.

Which results clamping the CVBS to ground.
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf