ADV7612 Initialization

Hi,

 We are also following the same sequences as mentioned in ADV7612-VER.2.9c.txt, But we don't face any issues.

 It seems you need to crosscheck your schematic b/w Reset/Power pin 1602.ADV7612_rec_sch_pcb_bom.zip

Thanks,

Poornima

Hi,

Thank you for your quick reply.

Could you please give me the recommended reset timing diagrams so that i can confirm with my schematic b/w reset.

Also, ADV7612 power up sequency timing diagrams will be very helpful.

Thanks,

Amani

Hi,

 Our Eval board schematics is under 7343.EVAL-ADV7612-7511_rev0p1_schems.pdf.  Please crosscheck with that.

 Please refer Page11 for POWER SUPPLY SEQUENCE in ADV7612 datasheet.

Thanks,

Poornima

Hi,

 Our Eval board schematics is under 7343.EVAL-ADV7612-7511_rev0p1_schems.pdf.  Please crosscheck with that.

 Please refer Page11 for POWER SUPPLY SEQUENCE in ADV7612 datasheet.

Thanks,

Poornima

Hi,

I checked the power sequence on my pcb and found out that it differs from what is written on page 11 of ADV7612 datasheet. Means that, 1.8V supplies power up first then 3.3V supplies.

Can the power sequence be the cause of noise on the video outputs of ADV7612?

Thanks,

Amani

Hi,

 Yes, the recommended power-up sequence of the ADV7612 is to power up the 3.3 V supplies first, followed by the 1.8 V supplies. Reset should be held low while the supplies are powered up.

Thanks,

Poornima

To add to Poornima's comment.  The power rails can come up at the same time as long as 1.8V rail does not exceed the 3.3V rail.  In general we recommend reset going high ~5ms after the power rails are stable. Followed by the first I2C transaction ~5ms after reset goes high.  A simple RC circuit can be used on the reset pin.

There are multiple sources of noisy images including layout and power rail poor power rail decoupling, 

Also make sure the exposed pad is connected to ground

If you can attach a schematic we can look it over for any obvious problems.

Hi,

Thank you Poornima and GuenterL for your quick answers.

I attached the schematic design, please could you look it over and let me know if there are any mistakes or problems that can be associated with getting noisy images.

ADV7612_schematic_.zip

Thanks,

Amani

1) Power rails look good.

2) EPAD connected to ground, good, often designers miss this.

3) XTALP pin needs to be sourced from the 1.8V clock.  See data sheet: 89 XTALP Miscellaneous analog Input Pin for 28.63636 MHz Crystal or an External 1.8 V, 28.63636 MHz Clock Oscillator Source to Clock the ADV7612. This might be the source of the noise.

4) You might enable LLC_DLL_EN bit and sweep LLC_DLL_PHASE.  Do this in case there are routing length differences between LLC and the data lines.

5) Check data, llc and sync pins for rise time/ ringing, adjust them using DR_STR, DR_STR_CLK and DR_STR_SYNC fields.  You have 33 Ohm series resistors which is good.

Hi,

Thank you for your reply.

For XTALP pin, this is a big problem since my product is already in market. I'll remake the pcb and correct this. So, after getting noisy images from ADV7612, i tried to stop the input tmds clock (make it LOW) for about 500ms and then release again and by doing so i'm getting clear images(images without noise). Is this the correct way to do? If it's correct means i can just update the software for the product which is already in market.

Thanks,

Amani

Of course the best way is to fix the board properly.  There are 3 simple ways to do it.

1. put low value resistor divider network on the output on IC6.  Maybe a 33 Ohm resistor from TP7 to ground will do.
2. Replace IC6 with something like a SN74AUP1G04 and connect the power ral to 1.8V.  This part has 3.6 tolerant inputs even with only a 1.8V rail.
3. Replace the whole crystal circuit with a crystal oscillator device like https://abracon.com/Oscillators/ASA.pdf

For units in the field, do what you have to do to fix it without a recall notice (very costly, up to you).  The crystal input over voltage may lead to long term reliability issues.

I am not sure what is causing the noise exactly.  Also on power up you might try powering down the  XTAL by setting XTAL_PDN bit to 1 then back to 0. 

Let me know how it goes.

Hi,

I tried to power down the XTAL by setting XTAL_PDN bit to 1 then back to 0, but nothing changed. I then tried to power down the chip by setting POWER_DOWN(0x0C) 0x62 then back to 0x42 and the image becomes clear without noise. Is powering down the chip, a correct procedure to solve the problem?

Thanks,

Amani

The XTAL_PDN power cycle was just a suggestion to try. I do not have the hardware here to duplicate the exact problem you are seeing.  

The only real problem I can see from the schematic is the crystal input levels.  If power cycling the device works then fine.  I still suggest fixing the crystal levels.