I'm not even sure what forum this goes in so I picked the best one that applies I suppose... please move to appropriate forum if this is in the wrong place..

I have a general question reguarding generic AV products for consumer TV use. Is the industry standard for video outputs to be ac coupled and for the reciever to handle dc restore (if needed by the rx) the signal? Or is it industry standard to provide a DC bias to the video signal? I would figure that consumner TVs would expect an AC coupled singal by I just wanted to be sure.

Since I am talking about AC signals this of course doesnt apply to hdmi/dvi. This is for component/rgb/etc signals.

Consumer products will do whatever is the lowest cost. If the output is DC coupled I would not assume that blanking is at ground level. DC coupling the output avoids the need for large output coupling capacitors but requires that the device have both positive and negative power supply voltages if blanking is to be near ground potential. I have seen devices that had DC coupled outputs with the video waveform, sync and all, above ground which would be the case if there is no negative power supply rail in the device. The bottom line is that the receiving device needs to be able to handle video with a DC component that may vary significantly and if a known blanking level voltage is required the receiving device needs to create it.

Consumer products will do whatever is the lowest cost. If the output is DC coupled I would not assume that blanking is at ground level. DC coupling the output avoids the need for large output coupling capacitors but requires that the device have both positive and negative power supply voltages if blanking is to be near ground potential. I have seen devices that had DC coupled outputs with the video waveform, sync and all, above ground which would be the case if there is no negative power supply rail in the device. The bottom line is that the receiving device needs to be able to handle video with a DC component that may vary significantly and if a known blanking level voltage is required the receiving device needs to create it.

Well yeah, the absolute lowest cost is not to AC couple. If consumer products are really just made at the absolute lowest cost I would suspect that very few devices are ac coupled. I already have my design finished and I made with a single rail supply so blanking and sync are not a ground/below ground but have a DC offset. I was going to see if I could get away without the AC coupling caps as that would save money/area. It sounds like I should theoretically be safe just leaving the outputs DC coupled.

Your thoughts?

Well yeah, the absolute lowest cost is not to AC couple. If consumer products are really just made at the absolute lowest cost I would suspect that very few devices are ac coupled. I already have my design finished and I made with a single rail supply so blanking and sync are not a ground/below ground but have a DC offset. I was going to see if I could get away without the AC coupling caps as that would save money/area. It sounds like I should theoretically be safe just leaving the outputs DC coupled.

Your thoughts?

Here are mine;

There is still responsible engineering practice even at low price points. Having more than a few mv of DC on an analog video signal line output is poor practice IMO when an electrolytic cap is fractions of a cent in consumer electronics manufacturing. Where I have seen poor video engineering is not the major consumer manufactures but rather the garage companies that make small switchers and distribution amps for consumer use. I don't think some of these companies even own a scope - if it makes a picture, it works!

Here is the potential problem. If the receiving device is DC coupled and your sender DC offset is too high, it will upset the bias in the receiving stage. In video that usally means clipping in one direction or the other.

Now coupling caps have drawbacks too. Too big they are overly inductive. Too small and they are low pass filters. This is especially difficult with analog video expected to run 10hz to 5mhz - and that's just NTSC not HDTV which is 30mhz.

Maxim has a paper on an equalization technology they developed that allows small caps (approx 300uf) to be used in portable applications 75ohm line drivers such as camcorders. I don't have a link but it's probably on their site.

www.maxim-ic.com

Well yeah, the absolute lowest cost is not to AC couple. If consumer products are really just made at the absolute lowest cost I would suspect that very few devices are ac coupled. I already have my design finished and I made with a single rail supply so blanking and sync are not a ground/below ground but have a DC offset. I was going to see if I could get away without the AC coupling caps as that would save money/area. It sounds like I should theoretically be safe just leaving the outputs DC coupled.

Your thoughts?

Is there a reason you can not AC couple at the input of whatever you are building? Assuming the first stage has a high input resistance you can use a relatively small cap and you will then not care about the presence of DC on the source video. Another possibility is to DC couple but put an integrator in the circuit to adjust the DC level to whatever long term average you want to maintain.

FYI I designed analog and digital professional television broadcast equipment for a living back in the 80s and 90s but have never dealt with consumer stuff other than looking at what comes out of some of it.

Here are mine;

There is still responsible engineering practice even at low price points. Having more than a few mv of DC on an analog video signal line output is poor practice IMO when an electrolytic cap is fractions of a cent in consumer electronics manufacturing. Where I have seen poor video engineering is not the major consumer manufactures but rather the garage companies that make small switchers and distribution amps for consumer use. I don't think some of these companies even own a scope - if it makes a picture, it works!

Here is the potential problem. If the receiving device is DC coupled and your sender DC offset is too high, it will upset the bias in the receiving stage. In video that usally means clipping in one direction or the other.

Now coupling caps have drawbacks too. Too big they are overly inductive. Too small and they are low pass filters. This is especially difficult with analog video expected to run 10hz to 5mhz - and that's just NTSC not HDTV which is 30mhz.

Maxim has a paper on an equalization technology they developed that allows small caps (approx 300uf) to be used in portable applications 75ohm line drivers such as camcorders. I don't have a link but it's probably on their site.

www.maxim-ic.com

I would not limit the universe of poor equipment design to garage consumer companies. I have looked at schematics of what was supposed to be professional equipment that could not possibly meet their published specifications. This was not a major manufacturer but still was being sold into that market. How many buyers ever check to see if the equipment they buy actually meets specs? Back when I was designing video equipment a lot of time was spent making sure that it met the published specs. It seems consumer equipment companies avoid problems with meeting specs by not publishing any. I would agree that for some of the garage consumer companies having a picture on the output is all the testing they do. I doubt they have any video test equipment. I assume you intended to say that a too small coupling cap acts as a high pass filter.

Good points guys. I'll just add some coupling caps on my outputs and just take the increased costs/area.

And to answer a previous posters question, I was referring to the output stage not the input stage. I designed my input stage to be able to properly handle both DC or AC coupled inputs.

And despite (for now) me being a "garage company", I own a scope and will be doing lots of testing beyond just seeing video on the screen :)

Thanks for your input guys and I'm glad there seems to be many on this forum who is familiar with the design side of AV.

I assume you intended to say that a too small coupling cap acts as a high pass filter.


Ahhh, yeah :o

Ok I have a random question about composite sync/sync on green signals.

So I figure the display resolution that a monitor displays from a rgbs or component signal is purely based on the frequency/pulse with of the sync signal.

For example:

15.75Khz = 320x240
31.5Khz = 640x480

etc.

If generating video, from say, a DAC, what are the exact specs/timing information on these signals? And how do you control the refresh rate? Can you even control the refresh rate with only a sync on green/composite sync signal?

Also, what is the front porch/back porch/pulse with of each resolution? Is there some kind of table for this information?

Thanks

Ok I have a random question about composite sync/sync on green signals.

So I figure the display resolution that a monitor displays from a rgbs or component signal is purely based on the frequency/pulse with of the sync signal.

For example:

15.75Khz = 320x240
31.5Khz = 640x480

etc.

If generating video, from say, a DAC, what are the exact specs/timing information on these signals? And how do you control the refresh rate? Can you even control the refresh rate with only a sync on green/composite sync signal?

Also, what is the front porch/back porch/pulse with of each resolution? Is there some kind of table for this information?

Thanks

Broadcast TV standards are well published mostly within SMPTE. I beleive computer video standards are as well but I never really researched them.

Keith Jack formally of Brooktree published a great book on digital video a few years back. I think it's still in print.

Ok I have a random question about composite sync/sync on green signals.

So I figure the display resolution that a monitor displays from a rgbs or component signal is purely based on the frequency/pulse with of the sync signal.

For example:

15.75Khz = 320x240
31.5Khz = 640x480

etc.

If generating video, from say, a DAC, what are the exact specs/timing information on these signals? And how do you control the refresh rate? Can you even control the refresh rate with only a sync on green/composite sync signal?

Also, what is the front porch/back porch/pulse with of each resolution? Is there some kind of table for this information?

Thanks

Are you talking about computer monitors or TV displays? If you are talking about an analog signal the horizontal resolution is determined by the upper frequency limit of the signal and can not be determined from the vertical and horizontal scan rates. The total number of lines can be determined by the ratio of horizontal to vertical scan rates. For example for NTSC television the exact H scan rate is the subcarrier frequency (3,579,545) divided by the number of subcarrier cycles per line which is 455/2. This makes the H rate 15,734.263... There are approximately 29.97 frames per second (30/1.001 to be exact). 15,734.263.../29.97 = 525 lines per frame. This is the total number of lines which includes the vertical retrace time. The number of active displayed lines will be 525 minus the total number of lines in the vertical blanking intervals for the two fields (21 per field). So we have 525 - 2 x 21 or 483 lines of active picture. For compressed digital video in most cases we only get 480 lines ( three are lost ) because 480 is an even multiple of the MPEG DCT block height of 8 lines.

You can get copies of all of the SMPTE video standards (not computer standards) at their web site. SMPTE standards are not free you must pay for them.

Are you talking about computer monitors or TV displays? If you are talking about an analog signal the horizontal resolution is determined by the upper frequency limit of the signal and can not be determined from the vertical and horizontal scan rates. The total number of lines can be determined by the ratio of horizontal to vertical scan rates. For example for NTSC television the exact H scan rate is the subcarrier frequency (3,579,545) divided by the number of subcarrier cycles per line which is 455/2. This makes the H rate 15,734.263... There are approximately 29.97 frames per second (30/1.001 to be exact). 15,734.263.../29.97 = 525 lines per frame. This is the total number of lines which includes the vertical retrace time. The number of active displayed lines will be 525 minus the total number of lines in the vertical blanking intervals for the two fields (21 per field). So we have 525 - 2 x 21 or 483 lines of active picture. For compressed digital video in most cases we only get 480 lines ( three are lost ) because 480 is an even multiple of the MPEG DCT block height of 8 lines.

You can get copies of all of the SMPTE video standards (not computer standards) at their web site. SMPTE standards are not free you must pay for them.

Ok I think you're a little confused on what I am asking. Things like subcarrier frequencies and such don't matter because I am talking about synthesized digital video such as the output from a DAC. And specifically I am talking about sync signals or sync on green.

For example for analog RGBHV output, there are two separate sync signals and they determine the output resolution and the refresh rate. The same applies for composite sync and sync on green signals (such as used with component video).

Even though there is no subcarrier the component outputs use the H and V scanning frequencies for color television. The 15,750 H scanning frequency stopped being used when black and white TV went away a very long time ago. The SMPTE documents will tell you what the waveforms should look like for video scanning standards. As for computer standards: if you want to know what comes out of a computer VGA output at various resolutions why not use the scope you say you have to look and see.