A short while ago, I did an analysis of an unbranded metal cased HDMI splitter, to discover that it has been designed in such a way to be non-compliant with HDCP requirements. At the time, I had ordered another model, in a plastic case, which didn’t arrive until last night – so let’s put that one under the knife.
The unit came in a plastic snap-lock bag, and comes with a charge-only style microUSB B cable with a very thin and flexible wire. The unit comes with protective film, and has a single cardboard guide with specs on the front side (claiming HDMI 1.4a compliance) and a connection diagram on the rear. The seller claims the unit is HDCP compliant.
The plastic on the unit is a very shiny finish, and the top is emblazoned with the HDMI logo. No real branding is existent.
One side of the unit houses a micro USB B connector for power, with a red LED to indicate power. The single HDMI input is on this side.
The other side has two outputs, with a green LED to indicate link on status on each of the ports.
There are holes on each side, which are clear to the internals. It appears to function as ventilation holes to keep the unit comfortable. The rear of the unit is unmarked, with a lid that has a shiny finish like the top.
What is it made of? I try to find out. The unit can be opened by prying the rear lid off, although the lid is held together by internal clips, which are prone to snapping.
The rear lid is marked with a moulding mark of SFX.
This allows us to then pry up the PCB out of the remaining plastic frame.
The PCB is marked H1X2_SFX_V10_20141120, which implies that it was made by a company called SFX, is version 1.0 and was designed 20th November 2014 (very recently). The PCB was made in week 4 of 2015, which is also very recent.
Already visible is a debug and SPI connector at the top, and test pads for the HDMI signals near each of the ports. Similar to the other unit, I suspect the ST Microelectronics microcontroller is used primarily for EDID repeating. No inductors are provided for filtering all the HDMI signals, which may mean a potential RFI issue due to sharp signal transitions.
The main chips are hidden under a heatsink though, which isn’t so much fun. However, the heatsink near the USB connector appears to be a standard linear regulator. The heatsinks are thermal-siliconed onto the chips, and with some very judicious prying and cleaning with a knife, the chips were exposed.
This splitter is unlike any other unit I’ve seen as it uses a single chip solution from Lontium. The LT86102EX is a 3rd-generation HDMI splitter using “ClearedEdge” technology, claiming support for up to 4kx2k formats up to 3.4Gbps data rate. It has adaptive equalization for long cable compensation, and has an integrated HDCP engine under hardware or software control which supports HDCP 1.4 repeating with onboard keys, or external keys. The product brief doesn’t really add much information to this. At face value, this is a good solution to meeting HDCP requirements, as the decrypted signals are kept internal to the chip, meaning the requirements of HDCP on input and output being maintained with little chance for compromise. However, this remains to be tested.
The other chip, as expected, is a stock standard 3.3v linear regulator. No surprises there.
However, the design of this unit is especially good for several reasons:
- It is compact, so it is easy to hide away behind a TV.
- It has heatsinks, so it’s likely that it will not have overheating issues.
- It can be powered directly from HDMI (although, in abuse of the HDMI 5v line specs), or from USB.
- The USB power could even come from the TV’s PVR port, thus allowing you to “run a video bus” and distribute the same vision and audio to a chain of TVs.
How a video bus system would work is that each TV powers its own repeater. One port of the repeater is used to drive its own TV, with the other port used to hook to the next splitter. Because the splitter itself is digital, and is lossless, there is no loss from distributing it in this fashion.
Source <=====> SPLITTER 1 <====> SPLITTER 2 <== ... ==> SPLITTER N V V V TV 1 TV 2 TV N
The rear of the PCB isn’t very interesting. There are two spaces, which may be used for optional EDID EEPROMs, but are not populated. There are quite a few diodes/transistors on the rear as well. The soldering is pretty good, although the HDMI connectors only have a very minimal amount of soldering on the connector hold-downs.
Testing the Unit
As usual, I gave the unit a bit of a test. Hooking it up without anything downstream does not result in my laptop detecting a display, so I suspect the EDID trickery depends on getting a successful EDID read-out from a downstream display and repeating it up the chain.
The splitter was verified to operate 1080p 60fps with no issues, and it appeared lossless to my eyes. I didn’t verify 3D functionality or any higher resolutions as I didn’t quite have the equipment to make that happen.
Then I decided to test whether the output of the splitter was indeed HDCP compliant. Again, using the Avermedia Live Gamer Portable to attempt recording the output of the splitter, it seemed that the HDCP was not being passed to the output ports either. While the chip appears to be able to perform HDCP repeating, it must have been configured not to do so, and is thus in violation of HDCP standards. It seems like operation in this mode might not be the intention of the chipset designer, however, may prove to be advantageous if there are issues with the HDCP implementation.
I do not condone piracy, and I don’t condone violation of copyrights, and neither does Avermedia, hence the warning they have on their site. Please use the unit for responsible legal purposes. This information has been presented purely as an analysis of the product’s behaviour and design. I cannot be held responsible if you should abuse this.
It is strange, but two out of two popular HDMI splitter units sourced from eBay are not compliant with HDCP standards and have outputs that lack the HDCP that is on the input. This is understandable for the metal-cased design due to the choice of hardware, but on this unit, the chipset supports HDCP repeating but it seems like it has been turned off deliberately.
The design of this unit is very convenient if you want to run a “video bus” around the place, the USB and HDMI powering of the splitter is very convenient, and the use of heatsinks likely increases reliability. It also claims a longer transmission range than the metal cased unit, and is much cheaper as well.