we have fails with ADV7511W in automotive application
can anyone help?
Could you please elaborate your issue?
Do you have one of our evaluation for ADV7511W?
How do you configure the board either through script or software?
Which specific script did you use it?
Did you do the test using eval board also?
Thank you Jeyasudha.N
I will answer your questions
Q1) Could you please elaborate your issue?
A1) Sometimes when we have the following fail scenario:
Q2) Do you have one of our evaluation for ADV7511W?
A2) No (we will see now how to get one)
Q3) How do you configure the board either through script or software?
A3) By software in start up
Q4) Which specific script did you use it?
A4) I will check later and return to you
Q5) Did you do the test using eval board also?
The status of the HPD pin can be read through register bit 0x42. Did you check this?
Sorry for the delay
About HPD register bit 0x42
when we had the problem we download all registers but now we don't find it and also we don't have a fail board since have a limited boards so we replaced the HDMI fail chip
About script we use at start up (I owe you an answer from before)
We always recommend customer to purchase an evaluation board so that we can rule out any issues with your hardware or software.
Did you refer the ADV7511W design support files ADV7511W Design Support Files ?
You can refer the Hardware and Programming manual (Table 2) where it describes the register. Why I want to check this particular register is to know the state of the HDMI sink.
Finally I found the ADV7511 (did not find ADV7511W programming guide) programming guide and realized the important of HPD bit (note I was asked to handle this fail although I was not the designer of this ADV7511W circuit)
I will find failed HDMI board, then verify HPD bit and come back
I checked the HPD pin by verify 0x42 register
With board including the failed ADV7511W and connecting to HDMI monitor I found 0x42 register value is "D0" that is HPD pin is in "0"
With the board including fine ADV7511W I found 0x42 register value is "F0" when HDMI monitor connect and "D0" when HDMI monitor was disconnect
Any insights according our finding about HPD?
We are checking with part specialist and will get back to you.
Sorry for the long delay.
When disconnecting / connecting the display power cable (12V banana plug), the ADV7511W stop working
When there is no Rx connected to the ADV7511 or if the Rx is unplugged or in this case power removed the ADV7511 will see this as a Hot plug event and reset its registers.
When the display is reconnected or re powered the ADV7511 will need to be reprogrammed in order to get video output again.
It sounds like you may have a damaged AD7511 if when you attach the monitor you do not see a HPD event.
Have you tried enabling the HPD interrupt and see if you see an event on attach / detach of the monitor ?
The related registers to the HPD interrupt are given in the table below.
Yes, the ADV7511 is damage and we have many damaged ADV7511, this is the reason for all this discussion
I was asked to verify HPD bit in order to verify fail reason
I verify HPD bit in 0x42 at damage ADV7511 and found out 0x42=0 that is the ADV7511 "think" I did not connect the monitor although the monitor is connected
If I will replace the ADV7511 by another new ADV7511 it will work (I know it because we did it many times before before) that is we have no problem to initial the ADV7511
Now, tell me please if there is a need to "enabling the HPD interrupt and see if .. see an event on attach/ detach of the monitor" as you suggested?
I want to add that we keep not attaching the monitor while chip is working because it damage the chip (sometimes) so we first attach the monitor and then operate the chip (insert power) and the opposite, we first take out the power and then detach the monitor
Have you ESD protection on the TMDS lines ?
It could be ESD damage when you attach the cable is the issue. If parts are constantly getting damaged this is what I'd suspect first.
The HPD interrupt is the event the system sw should operate off to know when the monitor is attached/detached so it knows when to configure the ADV7511.
Thanks for the replay
We use the SESD series P/N SESD0802Q4UG-0020-090 Ultra low Capacitor Diode to protect all TMDS lines, Is it not enough?
Is there ant way to verify if it is ESD fail or else?
The data sheet says they are unidirectional; do you have two on each line?
Can we get a partial schematic??
Very interesting, indeed we connect only one unidirectional diode on each line (we found such HDMI applications) and now we see your recommendation P/N EZAEG2A bidirectional ESD suppressor and we will consider implement it at the next board, thank you very much
If you have multiple devices that you believe are damaged you can send them back to ADI for failure analysis.
They will be able to tell then if the failure/damage is due to Electrical Over Stress on the pins i.e ESD or something else.
In order to do this you should contact your sales channel and ask them for help in sending parts to the FA (Failure Analysis) process.
I had a quick look at the SESD part you are using, it looks fine to me, its touted for HDMI 2.0, HDMI 1.4 applications.
For reference we have used Semtech RClamp7534P or RClamp0524P on our eval boards.
When you say you don't connect the monitor when the device is working, what do you mean ? Usually our HDMI Txs are in a unprogrammed state and held in reset until the cable is connected. Have you overridden this feature ?
Thank you for the fast answer
Yes we have more failed IC's and we already sent them, they are on the way
About connecting the monitor:
1st - connect the monitor HDMI cable
2nd - turn on the system
Is the system powered up during 1st stage ? I.e is voltage, gnd connections available or is the board floating when you connect the HDMI cable.
It may be there is no good path for the ESD diodes to conduct to if a charged cable is plugged in and the board is more or less floating.
As soon as we receive the FA returned units I think FA will be in touch to explain the process.
Just want to update that lately I was busy in work with other urgent issues and also some personal matter
Hope coming back to this issue soon
As you can see from my discussion with Harryh on March-8 and March-9 we found out that we need to replace "unidirectional diode SESD0802Q4UG-0020-090" with "bidirectional ESD suppressor EZAEG2A" to receive a better ESD protection
we will do it at the next board and then we will see if we have any fail
Than you for your support
Retrieving data ...