HDMI: High Emissions, Poor Grounds Don’t Mix

Whilst I agree with the philosophy of maintaining good grounding design, HDMI does not utilise ground as the main signal return for any of the high bandwidth signals, since they are all balanced.

Most professional AV equipment utilises balanced connections which provide greater immunity to interference and do not use ground as a signal path.

i have had these issues.i spoke with several companiesand no resolve. When i was at Electronic House Expo in Orlando, i spoke with one of many vendors.Unfortunatley i didn’t get his name.He Recommended using Cat6 Shielded Cable with metal connectors,and solder the shield At each end to the connector. This Worked.

You are quite correct on one point. This is a highly complex subject and there are no simple answers. However the notion of “Good grounding techniques” is not that simple either. As describes in the comment above, HDMI uses balanced transmission, specifuically TMDS. The worst thing to do is to tie the ground on a long cable to the chassis of both the source and destination device, i.e. a BluRay player to a projector. This invites a “ground loop” back through the AC power system. Even though we are dealing with a “digital” signal, if the ground loop voltage exceeds the 50% point of the signal amplitude, it will corrupt it. An ideal situation is to only tie the HDMI shield at the source device. However this is not standard cable manufacturing SOP and as well may violate EMI/RFI compliance. Yes it is not simple at all!

Andy, that is not necessarily correct, although you are correct that it is a complex topic. These days, especially for balanced signals such as HDMI it is best to connect the shield to the chassis at BOTH ends, for EMI reasons. This is exactly what the manufacturers do. Many pieces of AV equipment are NOT connected to mains “earth”, to avoid the ground loops you mention. Even if both BluRay player and projector ARE both connected to mains earth, sensible wiring will avoid any issues (HDMI has a certain immunity to ground potential).

The origional article is questioning grounding in general. My point is what my seem a proper grounding scheme to an installation technician may not be electrically ideal. Yes TMDS like any digital signal has good noise immunity. If however the noise floor including AC hum exceeds that 50% amplitude, it will corrupt the signal. In most household situations, there should be no significant ground loop potential. However in a commercial venue with three phase power, there is a good chance of significant ground noise. I am not advocating breaking the shield at one end of the cable. However if you had a serious ground loop, doing this probably would eliminate the problem.loop

I design TV broadcast facilities for a living. I deal with this issue almost daily.

Andy, I’m wondering if you have actually tried breaking the shield at one end of a HDMI cable? I suspect that if there was a grounding problem, this could make things even worse for immunity.

By removing the shield, you are allowing the chassis of the equipment at each end to float to whatever the local ground voltage is (assuming the equipment is grounded). As you imply, in a commercial setting, this voltage could be significant. At least by leaving the shield in place, you are trying to equalise the ground voltage at each end. Of course, this could allow significant current to flow in the shield, but there is plently of copper in the shield, and this shouldn’t cause a big problem. Sometimes large parallel earth conductors (PEC) are used to prevent the ground current from flowing in the shield.

Well breaking the shield on all my HDMI connections is what I did after weeks of effort trying to solve a ground loop hum.  It did seem to work for me and lowered my noise db’s enough for me to finally enjoy my equipment.
It was a drastic measure but the HDMI cables were setting up a ground loop and I had no other way to break it.

OK, thanks for clarifying. By “hum”, I guess you are referring to an “unbalanced” analog audio connection somewhere in your system, which would be much more susceptible to small levels of noise than HDMI (or a balanced audio connection). Good to hear that you managed to fix the problem.

Two further points I’d like to make at this point:

There is one signal in an HDMI cable which is unbalanced (and therefore less immune to noise). This is the DDC/CEC signal. It has a much lower bandwidth than the TMDS signals described above, but this doesn’t make it less immune to noise.

The other point concerns disconnecting the ‘Ground’ signal in the cable or connector. I assume people are referring here to the cable screen which is connected to the outer case of the connectors. I cannot understand how anyone can manage disconnecting this path without physycally destroying the moulded connectror(s) and/or the cable. Such disruption would almost certainly affect the bandwidth capabilities of the cable given the close manufacturing tolerancies required for both cable and connectors.

I am unaware of any re-wireable HDMI connectors currently on the market, nor any one who has the eyesight and dexterity to re-terminate one in the field without specialist tooling and jigs.

I think some comments above refer to lifting ‘ground’ connections when using other types of signal. In my opinion this is bad (and sometimes dangerous) practice. There are always ways to overcome signal noise induced from ground loops. Sometime these require careful design of both electrical and AV installations and may also require additional equipment.

jeggers, I was wondering the same thing about how to disconnect the shield in a HDMI connector. I would never attempt this myself. 

The old idea of disconnecting the shield to avoid ground loops seems to be dissapearing these days, in part due to the compromised EMI performance that would result. With digital signals being used more, “hum” being injected isn’t such a big deal anyway.