Post Go back to editing

ADV7393 I2C

Good morning.

Is there a short check list for fault in I2C?

A single card with ADV7393 have I2C fault as follows: the ADV7393 does not issue I2C acknowledge. The I2C on the card worked before, and there are other cards that don't have this error.

already checked:

  • I2C singals to the ADV7393 - SDA, SCL - are OK on the oscilloscope, but the acknowledge is missing
  • power is OK
  • The ADV7393 is not in RESET
  • there's 27MHz clock to the ADV7393
  • the component is not hot
  • ALSB pin is low through 4.7K pull-down resistor, address 0x54

questions:

1. is there a short check list for this problem? I mean requirements to verify that must be OK to support any I2C activity?

2. other suggestion before ADV7393 replacement? it will take time before i get another ADV7393.

thanks.

Ofek Nir

Here's my list:

  1. make sure the power rails are good
  2. make sure the device is out of reset and the reset goes high after both power rails are stable.
  3. make sure SDA and SCL signals are going to the right pins.  I've swapped them before, it was very embarrassing when someone finally pointed it out to me.
  4. Verify the pull up resistors on SDA and SCL, 2.2k - 10k. 
  5. Verify the SDA and SCK signals wiggle and are the right amplitude.  0 - 3.3V
  6. Verify the I2C SDA master becomes an input or is tri-stated during the ACK bit so the 7393 can pull SDA low.
  7. Verify you are using the correct device address.  For 7393 with ALSB low it should be read = 0x55, write = 0x54.  Match against Table 16, Figure 48.
  8. Verify the start command is correct. See Figures 49, 50
  9. Check to see if you can access anything else tied to the I2C bus, like eeprom.  Tie something to the bus to make sure the bus is working as expected.

I think you've done most of this.  If you do exactly what's done in figures 49 & 50 the device should work just fine.

  • thanks very much, the list is very good.

    I found that althouh the I2C signals look OK at the scope, except the acknowledge, something happend on SDA line. most probably chip were destroyed.

    thank you for the support, from here I can continue to isolate the probelm.

    Ofek.

  • Hello,

    I've got the same problem with ADV7391. The device doesn't response on the I2C, but I've got a composite video signal from one of the DACs. The video has the sync impulse and the color sync signal.

    System conditions:

    The ADV7391 has an additional back-end to drive 2 composite and 2 YC signals at the same time. Therefor the RSET resistor is set to 4K2 to reduce the output currents to a load of 300R. 

    Here's my checklist for mounting the device:

    - The IC was soldered with 320 °C hot air.

    - The soldering was reviewed under a micorscope

    - Electrical Tests were made

         - check of netlist integrity (the probe was set on the top of IC-pin to check if it is really connected)

         - short circuit check between the nearby IC-pins

         - short circuit check to GND (because of the heat sink pad)

    - Start of operation test with 12Vdc and a current of 50mA. Composite and YC Outputs open.

    - Adding I2C interface with 3V high level, ALSB=low, device address=0x54 for write op. (No acknowledge)

    - checking the bit stream with the oscilloscope (Signals look good)

    I had the idea that the chip might be destroyed, too. But I had the same results after replacing it. I checked the supply voltages, they are alright. I also checked if I might rotateted the IC or if my schematic model is wrong, but found no mistake.

    Before the ADV7391 I already used the ADV7179. I had some problems with the BLANK pin, which is recommended in the application circuit when not used to tie to ground, but differs from the other descriptions in the datasheet. I switched to the ADV7391 because the blank impulse is about - 300mV. I would be very pleased if someone could tell me if that is correct, otherwise I can go back to ADV7179 and do a manual offset calibration.

    Thanks so far, regards alex.

  • good morning, Alex.

    Yot may try the checklist seggested by GuentelrL in the second letter of this chain. It is very good.

    Also you may try to compare resistance SDA-ground to the resistance SCL-ground. if the difference is high, probably at least one I2C component failed. That can happen even if the signals look perfectly ok on scope (exclude acknoledge).

    The video on the screen may be default of the component.

    good luck.

    Ofek.

  • Hello,

    the I2C SDA and SCL lines are all HIGH-Z. I already had checked the points of the checklist by GuentelrL. As I mentioned I had used the ADV7179 which is not far different and it worked well with the master in the system. I also put out the board and used an external supply and evaluation I2C-RS232 adapter, but it showed the same result. The only thing that is really important for me is if the blank impulse is about -300mV as it can be seen in the datasheet output plots. Otherwise I could go back my old system.

    Regards, Alex.

  • Hi Alex,

    Other then following the list above and since you've already replaced the part once, I have no other suggestions.  The part should respond just like any other normal I2C device.  Just make sure that during the ACK bit the master is really tri-stated.

    If I'm reading the tables correctly I believe upon reset DAC1 is active and should have CVBS on it.