SN65LVCP408 output

Hi Shareef,

I am a little confused what you mean. you mean that if no input, let the input float, then the output is toggling, keep changing between low and high? The output is a certain frequency waveform or something random?

Regards

Michael

Hi Michael,

Thanks. The inputs to the crosspoint switch are not left floating but are connected to the Equalizer outputs (LMH0394). The Equalizer outputs are held high when there is no valid input signal.

When all the four output ports of crosspoint switch are enabled to carry their respective input signals (default register settings), it causes the signals on output ports 0 and 2  to toggle in random manner. If I tri-state the output ports of 1,2 and 3 I dont see this behaviour on output port 0. 

Regards

Shareef

Hi Shareef,

Is the LMH0394 output connected to LVCP408 input with DC-couple or AC-couple? I think if AC-couple, the noise in the input of LVCP408 may cause toggling when output of LMH0394 is DC.

If you don't use DC couple, could you post a sch for the connection of input stage.  

Hi Michael,

Yes, the LMH0394 outputs are AC coupled using 0.01uF capacitors to the input of LVCP408. This is done as per recommendation in LVCP408 datasheet. One more observation is that when the toggling happens in outputs of LMH device I could see that the signal level on pins (C\D\ and MUTE) also toggles which is very strange.

Could you please recommend on how to prevent the noise from inputs of crosspoint switch affecting the LMH0394 outputs?

I also feel it cannot be DC coupled as the dc bias is different for outputs of LMH0394 and inputs of LVCP408. Please let me know. Thank you!

Regards,

Shareef

Hi Shareef,

You mean that the output of LMH0394 is toggling? Your previous words didn't mention that, so I may miss that. You said the output of LMH0394 is tied to high when no input.  

So actually the output of LMH0394 and LVCP408 both are toggling? Could you use scope to observe if these outputs are related?  the waveform is similar ?

Or maybe I misunderstand your words. You mean when LVCP408 is toggling, the LMH0394 pin MUTE, C, D is toggling, too. I am a little suspicious of your power decouple. How much nF/uF/pF capacitor do you use.

If the toggling is really from the noise, i have to think more to come up with an idea.

Hi Michael,

We could make the following observations on a scope,

1. Outputs of LMH0394 and LVCP408 were not toggling but the MUTE and C,D pins of LMH0394 were toggling. Sorry for giving incorrect information on my earlier posts regarding this.

2. I removed the ac coupling capacitors from output of LMH0394 and see that MUTE and C,D pins still toggle at LMH0394 side without any input connected to its BNC. So, this seems to be an issue with the LMH0394 design now.

Let me know if you have any input to find root cause of this problem.

Regards,

Shareef

Shareef,

Based on your description, I'm assuming you have the MUTE and CD pins (pins 14 and 15) connected together.  Is there anything else connected to these pins in your design?

Are you using the LMH0394 in Pin Mode (SPI_EN = GND)?

For adaptive cable equalizers such as the LMH0394, the carrier detect (CD) being active in the absence of an input signal is commonly caused by noise coupling to the input.  The equalizer gain is at a maximum when the input signal is at a minimum, which occurs at maximum cable length as well as when there is no input signal.  Interfering signals and noise can be amplified under these maximum-gain conditions.

Can you check your design and PCB layout to see if there could be noise or interfering signals near the LMH0394 inputs?  For best performance, the equalizer inputs should be isolated from such interference.

Also, for the AC coupling capacitors on the LMH0394 output, 0.01 µF is too low for SDI data.  SDI data streams include SMPTE pathological patterns with a large DC component.  When AC coupling the equalizer output, we recommend using 4.7 µF capacitors to reduce the DC shift caused by these pathological patterns.

Gary Melchior

Hi Gary,

Yes, the MUTE and CD pins are tied together and nothing else is connected to them; LMH0394 device is being used in pin mode with SPI_EN connected to GND.

We have used 4.7uF capacitors at SDO outputs as mentioned in datasheet and only difference that I see is that AEC capacitor is not connected in our case (as it is optional). 

I tried removing the input capacitors connected to SDI input and see that signal at MUTE and CD pins don't toggle.

From the layout, I see that two more LMH0394 devices are placed in bottom layer and routed in bottom layer. But we are observing the issue (MUTE and CD pin toggling) on two devices which are placed at top side. 

Please note that the SDI traces and devices on top and bottom layer are spaced apart. Let me know if you foresee any possible causes.

Regards,

Shareef

Shareef,

Are the LMH0394 equalizer outputs (pins 10 & 11) toggling as well (as you indicated earlier), or is it only the CD (/MUTE) that is toggling?

Based on the toggling issue going away when you remove the SDI input capacitors, it sounds like this may be an issue with noise coupling somehow to the equalizer input.

Do you have active signals connected to the other nearby LMH0394 SDI inputs when this issue occurs?  Can you try disconnecting all nearby inputs, and also muting the outputs (drive 2.5V on the MUTE pin) for the nearby LMH0394 devices, to determine if this is a crosstalk issue?

Gary Melchior