HDMI DS34RT5110

Greetings -

Sorry for the holiday delay in response.  The DS34RT5110 has a TMDS style driver.  It is a sink only driver, the Receiver that it talks to provides the pull up resistors (50 Ohm) which are to 3.3V.  5V is not expected.  With the normal load applied, the VOH should be VCC (3.3V) and the VOL of 3.3V - 0.5V for about 2.8V.  This will set up a differential of +/-500mV for a 1Vp-p differential swing.  There is also a VOD_CRL that sets the correct swing for a DC or AC coupled application.

If you are measing the signal, depending upon probe / scope combination, please validate the signal on a known reference such as a 3.3V LVCMOS signal to make sure the set up is correct.

The Datasheet indicates pins 16, 17, and 18 are NC pins - leave them open, make no connection.  For factory test, they are a gated serial port for test purposes.

The CAT 5 / 6 cable is outside of the HDMI specification, compliance testing is usually done only with the standardized cable and loads.

John Goldie
DPS APPS / SVA / www.ti.com

Hello,

I think I need to clarify a couple of things.    I am using the DS34RT5110 EVKC demo boards, with the driver board and receiver board in the configuration. DVD player ->  HDMI cable->DS34RT5110 driver board->two Category cables ->DS34RT5110 Receiver-> HDMI cable ->TV 

  The TMDS signal appears to have a differential swing larger than 1Vp-p approx 2.1Vp-p  which comes from V(data+)  - V(data-).  This output is from the female HDMI receptacle from the RX board: the HDMI standard says a device sink/source is to have a female receptacle, I then use a test adapter with HDMI male connector that measures the signal's characteristic.  The Vswing (2.1 V) is outside of the HDMI specification, when the configuration is tested at the output of the HDMI transmission path ( DVD player ->  HDMI cable->DS34RT5110 driver board->two Category cables ->DS34RT5110 Receiver-> TEST point) - this was tested as a sink. 

In order to establish the HDMI link either a DVD or Signal generator is needs (I used a DVD player) and the DS34RT5110 Receiver was tested as a SINK.  The Test fixture has 50 ohm resistors to VCC (3.3V) per the spec.

Even when I tested at the output of the DS34RT5110 driver board, the Vswing is 2.1 V, so the Category Cabling or configuration is not changing the TMDS data signaling. 

Can you please explain why the DS34RT5110 TMDS data signals are 2.1Vp-p?

As for the VOC_CRL being set correctly, the DS34RT5110 EVKC quick start guide, instructs using the 3.3 V as the DC power which I am.  Under the Adjustment and Control Description, Driver Board, the JP37 and JP20 are to control VOD_CRL, the function that is described as JP20 sets the external resistor to 24K ohm for VO=1000mVpp  and JP37 sets the external resistor to 12K ohm for VO=2000mVpp.    [VO is the difference between the differential data signals per the Figure 2 in the “DS34RT5110 DVI, Retimer with Input Equalization and Output De-Emphasis” literature].   I am using the 24K resistor, per the driver board’s ref designator JP29, so the TMDS Vpp should be 1000mVpp.  (The ref designator on the PCB does not match the schematic or the VOC_CRL description, see my comment below).

I would like to point that the JP37 and JP29 are the ref designators on the board and on the schematic and are not JP37 and JP20.  Could someone please confirm the BOM and the schematic together and resolve the discrepancies?  (C52 on the schematic is 33uF and is 68 uF on the BOM, C53 on the schematic is 68uF and is 0.1 uF on the BOM, there are many more discrepancies, could continue to list the other issues, but I think it would be most productive is the BOM and schematic discrepancies are resolved). There are similar issues with the

Thanks,

Louann

Louann

Please bear with me as I ask for some more clarification.

When you read the signal swing, how are you doing that, and are you certain that the test point was offering the proper termination? - The HDMI sink (in your case, the TV) should be presenting a 50Ohm load to AVcc - you state that you have a test fixture with 50Ohm resistors to VCC, but it is not clear to me if this is the fixture that you are using to measure the TMDS signal swing.  

I am checking into the JP37/JP29 - JP20 issue - please wait for another reply regarding that....

Mark Sauerwald

Hi Mark,

The TMDS driver board has a number of 551600199-034 REV B.   The RX board is -024 REV B. 

As for the Test Fixture I am using it to measure the TMDS signal (differential scope probe).  Please note the data was taken about a month ago (the setup is not available as of now... unfortunately). 

Thank you

Louann

Louann

The documentation that you have is an old revision.  There is a new revision which will be up on the web site soon, but if you want to get a copy now, drop me an e-mail at mark.sauerwald@ti.com, and I will reply with the new versions.

For the overvoltage issue, what it sounds like is that the driver was not properly terminated, this would happed if the HDMI cable between the driver board and the TV were not plugged in, or if the TV did not have the proper termination at it's receive end. 

Mark Sauerwald

FYI - here are a few settings to check, also remember to connect a load to the board, a TMDS driver uses the pull ups in the "load" for VOH level.

To be short, for the VOD level control:

Driver board:

Connect JP29 (R19=24kohm) to get 1Vpp

Connect JP37 (R23=12kohm) to get 2Vpp

Receiver Board:

Connect JP46 (R68=24kohm) to get 1Vpp

Connect JP39 (R67=12kohm) to get 2Vpp

John Goldie
DPS APPS / SVA

Hi Mark,

In order to test the DS34RT5110, I used a BluRay DVD as the source in the following configuration

BluRay -> HDMI cable-> DS34RT5110 Driver board -> 2 Category Cables-DS34RT5110 RX board -> HDMI Test Fixture (Test Fixture is from Wilder Technologies) and the SMA outputs from the clk (clk + and clk -) and data 0 (data0 + and data0 -) to 2 differential sma scope probes. 

The differential probes (that are DC coupled and pulled up to 3.3 V) are terminated correctly, per the test vendor's data sheet and application engineer.  In addition to the Vswing being larger than expected, even when I turned on the emphasis, I saw the wave form distorted, but not really amplified as I would have expected. I have several pots that I can share via email I can share.

I am going to set-up the scope measurements again, but I am going to use BiasTs instead to DC couple to 3.3V and generic scope probes and  I will see if I can get the Vswing within range.  One question: Should the measurements be taken at both the Driver board and the RX board?

Thanks,

Louann