OK guys this is a weird setup but here is the breakdown. I had a house built recently and had them run a high quality VGA cable to my projector through the ceiling. The place is finished now and I'm stuck with a VGA cable running from my A/V equipment location to my projector location. The projector accepts DVI, Component and HDMI inputs (its a new Optoma HD80 1080p). Given my current setup I can not run an HDMI cable any more through there because everything is finished and drywalled etc... So here is my question. I would like to connect my Xbox 360, HD satelite, etc with hdmi. I was thining of getting an HDMI to DVI converter than a DVI to HDMI converter on both ends of my cable such that an HDMI signal leaves my A/V equipment and goes HDMI-DVI-VGA then through the existing VGA cable in the walls/ceiling and at the projector end it goes back from VGA-DVI to HDMI which is plugged into the projector. Will this setup work?

Keep in mind I'm not trying to convert the HDMI signal to anything other than HDMI since I want an HDMI signal at the other end, all I want really is to run the HDMI signal through existing wires in the VGA cable. Is this possible?

If you make the HDMI signal through the VGA cable, then the signal has to be converted. If you convert the HDMI to VGA, then VGA to HDMI, there will be signal quality lossses.

No, it won't work. The HDMI/DVI adapter will pass only a digital signal, and the DVI/VGA adapter will pass only an analog signal, so the net result is that nothing gets through that connection. There's no other practical way to do this, either; your VGA cable does not have the right number of conductors and does not have the right impedance (it's 75 ohm unbalanced; you need 100 ohm balanced) for HDMI.

Kurt
Blue Jeans Cable (http://www.bluejeanscable.com/store/hdmi-cables/index.htm)

Thanks for the responses. How many of the HDMI cable pins are actually used for video? The VGA cable I believe has 15 pins. So as long as the HDMI uses less than 15 pins, it should be enough. However, the impedance is a concern. Obviously a digital signal over the wire should be no different than an analog signal through a wire. A wire is a wire afterall, it doesnt care what type of signal it carries, however the impedance might be an issue.

If you could find an accurate pinout for HDMI, keep in mind you would still have to hack together a pair of adapters. If you make it that far, you'll still have to make an offering to the God of Amateur Electronics that no errors in the connections will cause destruction of your equipment. And finally, then you'll be able to test whether your impedance mismatch and balanced/unbalanced setup will work over your length of cable.

Personally, that would put me off right there. I would investigate instead how to pass component video over the VGA cable. Good luck to you.

Component cable is not a problem, I have a setup for that. However, from what I have researched, I can not get a 1080p signal from my Xbox360 - HD-DVD, or PS3 blu-ray over component cables. I'd like to utilize this 1080pprojector to its full capacity. Is there a way to actually get 1080p over a component connection?

How many of the HDMI cable pins are actually used for video? The VGA cable I believe has 15 pins. So as long as the HDMI uses less than 15 pins, it should be enough. However, the impedance is a concern. Obviously a digital signal over the wire should be no different than an analog signal through a wire. A wire is a wire afterall, it doesnt care what type of signal it carries, however the impedance might be an issue.

Eighteen of the pins on the HDMI cable have video-related functions; there is only one which is unused (reserved for possible future use). You could probably get away without having the CEC pin hooked up, and the four pairs have grounds which are typically tied together at each end in devices, so one could probably short them together into one pin in the connector with no ill effect.

But this still isn't going to work. First, the VGA cable may or may not have fifteen active conductors; you'd have to test to see exactly what's present but it could be as few as nine (R,G,B,H,V, Red Ground, B Ground, G Ground, Sync ground).

Regardless of conductor number, the problem you have here is that the impedance is badly, badly off, and at 1080p that's an enormous problem. The bitrate at 1080p/60 is 1.485 Gbps. Since one pulse up/one down is basically one cycle, this equates for bandwidth purposes to a fundamental frequency of half that, or 742.5 MHz, and to get good sharp digital transitions it's very helpful to have a strong third harmonic, which is three times that, or 2.227 GHz.

If we assume the critical distance (where characteristic impedance of the transmission line starts to play a large role in signal delivery) to be 1/4 wavelength, and we assume a foam dielectric with 83% VP, here's what we get. The wavelength "in air" at 742.5 MHz is about 40 cm (16 inches), so that the quarter-wavelength is about four inches. Multiply that by the VP and you get a critical distance a bit less than three-and-a-half inches. (And the wavelength of the third harmonic is a third of that). You haven't indicated how long the VGA cable is, but I'm wagering that it is much, much longer than that.

Now, turning from theory to practice, I can also tell you that short impedance mismatch segments have an enormous impact at 1080p. When we prototyped our Belden HDMI cables, one of the first things I did with our test reel was to run sample lengths and see how far we could run various resolutions on common equipment. Because HDMI connectors are a bear to solder and we wanted to play with the length, we built HDMI-to-DB25 breakout boxes and then put DB25 connectors on the HDMI cable so that we could quickly cut and resolder to change length. The difference between the length I could get to work in that configuration and the length I could get to work when we later had proper production samples was on the order of 50%; and that, mind you, was after I had taken extreme care to maintain proper impedance by separating my pairs by only the bare minimum required, maintaining the integrity of the foil wrap within a couple mm of each solder joint, and so on. So my impedance mismatch zone was something less than two inches total, I didn't have to convert balanced lines to unbalanced as you would have to do, and it screwed 1080p up so much that it was impossible to get reliable distance projections for the production cable. If you take the same signal and run it through, say, 20 feet of 75 ohm unbalanced cable (which is what most of the material in a high-quality VGA cable is), it will fall apart very, very badly and it is unlikely that you will have recoverable signal at the other end.

On the subject of a digital signal being no different from an analog signal over wire--yes, that's certainly true, so far as it goes. But reliable digital signalling does depend on having good transition rise and fall times, and those, being influenced heavily by return loss, are highly dependent on proper impedance.

Kurt
Blue Jeans Cable (http://www.bluejeanscable.com/store/hdmi-cables/index.htm)

Kurt.

Thank you very much for that informative post. As I said the impedance difference is what I thought the real problem was. I think I'll have to just stick with using a VGA or Component signal going to my PJ. Thank you all for your help.