TDP142: output signal swing level

Would you please try the following experiments? Would you please send me the schematic and layout for review?

1. Please swap the scope probe between Lane 0/3 and Lane1/2 to make sure it is not probe/probe setting issue.

2. Replace the AC coupling capacitors on Lane 0/3 

3. X-ray DP142, does it have a solid thermal pad connection to GND? If necessary, please re-flow DP142.

Thanks

David

Hi~David:

Schematic as attachment and reply your suggestion as below --

1. I have swapped the probe, same symptom.

2. Actually, I measured 2 point for TDP142 output (before and after AC cap), same symptom.

3. I will try the X-ray.

Schematic--

Thanks

Dodge

Dodge

Do you have AC coupling caps on the DP142 input? 

How does AUX being connected between the source and the sink? 

Otherwise, the schematic looks ok.

Thanks

David

Hi~David:

Even we know the root cause already but I'm curious that input AC coupling cap impact so much for TDP142 output.

From waveform of lane 0 input signal (w/o input AC cap) that swing level is same as lane1/2 (w/ input AC cap).

But lane 0 output signal can be attenuated to ~2/5.

Would you please let us know why TDP142 so sensitive for input AC coupling cap?

Thanks!!

Dodge

Dodge

1. TDP142 is a linear re-driver. So as long as the input signal amplitude is within the linear region of TDP142, the output signal will have the same amplitude as the input without compression.

2. On the input side, when the AC coupling capacitor is present, the AC coupling capacitor changes the DC component (common-mode voltage), while passing along the AC component (voltage swing) of the signal. The internal termination of the DP142 and USB/DP combo PHY then provides the proper DC bias of the AC coupling capacitor. But without the AC coupling capacitors (DC coupled), then the internal termination will impact the voltage swing of the signal.

Thanks

David

Hi~David:

Got it, thanks.

Dodge

Hi~David:

Another questions 

1. TDP142 is a linear re-driver. So as long as the input signal amplitude is within the linear region of TDP142, the output signal will have the same amplitude as the input without compression.

Q1: What's signal amplitude? DC or AC component

Q2: What's the linear region range? Can we see that range listed in datasheet?  Is the parameter VIC (Input common mode voltage) = 0 ~ 2V  that listed in datasheet page 7 ?

Q3: If input over the linear region, why not to keep max input for output but go compression ?

Thanks!!

Dodge

Dodge

Please refer to Figure 2, DisplayPort Linearity Curve in the TDP142 datasheet.

Take for example EQ0 linearity curve, when input differential voltage is 1V, the output differential voltage is also 1V. But as you go higher on the curve, you can see the output differential voltage is smaller than the input differential voltage, this is where the compression happens.

Thanks

David

Hi~David:

But you can see my original waveform for Lane0, the input differential voltage is around 1V(p-p) but output is compressed already?

That different with the linearity curves.

CH1=Lane0, Input, (EQ=15, w/o input AC cap), V(p-p) ~ 1V

CH1=Lane0, Output, V(p-p) ~250mV

Thanks

Dodge

Dodge

Please see my previous response. The majority of reduction in signal amplitude comes from the missing AC caps, not because the linearity compression.

2. On the input side, when the AC coupling capacitor is present, the AC coupling capacitor changes the DC component (common-mode voltage), while passing along the AC component (voltage swing) of the signal. The internal termination of the DP142 and USB/DP combo PHY then provides the proper DC bias of the AC coupling capacitor. But without the AC coupling capacitors (DC coupled), then the internal termination will impact the voltage swing of the signal.

Thanks

David

Hi~David

Could you let us know if there is no input AC cap that will not meet or over which parameter of DC or AC Electrical Characteristics in datasheet to cause this amplitude reduction?

Thanks!!

Dodge

Dodge

The characterization data in the datasheet assumes the input is AC-coupled per the DisplayPort spec, it does not list the condition when the input is DC-coupled as it will be a spec violation.

Thanks

David

Thanks, David