• Locked Locked
  • Replies 2 replies
  • Subscribers 46 subscribers
  • Views 650 views
  • Users 0 members are here
Support feedback
Options
Options
Related

This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

TLV5610IPW problem

Maik Staberock
Prodigy 20 points
Other Parts Discussed in Thread: TLV5610

Hello!

The company I work for is using the TLV5610 for a long time now and we never had problems so far. It is used in various control modules for our power supplies and electronic loads. But since a few months it started having problems on one of the boards. 

What happens? The DAC does not put out any voltage

Date code? For example, 71T. Means 2007

Some information:

- I have an older board with a DAC of 6AT date code, which runs fine

- nothing has changed (same logic ICs, same microcontroller, same software)

- on the older board it puts out voltage on the 8 channels, on a newer board from current production it does not

- the minimum timing values are absolutely not exceeded

- reference voltage is 4.096V and supply voltage is 5V for digital and analogue

What we did so far:

- compared the related components (logic ICs, MCU, DAC etc.) of the older and the newer board

- investigated the signals to the DAC (MOSI, SCLK) with an oscilloscope on both boards and compared - identical

- checked all supplies (Vref, AVdd, DVdd)

The main problem is that it does not happen for every board. Within a charge of 100 boards we have, let's say, 30% with this error and the rest is OK.

30% waste is, of course, inacceptable regarding the unit price of the DAC.

We can not find out why this is happening. If we desolder a chip that does not work correctly and put it into another PCB (different type), then it works.

So it can't actually be the DAC itself, one might think. But what then? The board layout hasn't changed, only the production dates of the components.

 

Any help or suggestions are gladly appreciated.

TIA.

Maik Staberock

Elektro-Automatik GmbH & Co. KG
41747 Viersen, Germany 

 

  • TI__Guru**** 181981 points

    Hi Maik,

    Sounds odd - since you are able to move a non working device from one board to another and have it work, you might want to take a close look at the actual soldering of the chips on a non-working unit.  Be sure everything has good connection to the PWB - some of the newer RohS solders and assembly processes might be causing your trouble.  It could also be that something is marginal in the timing.  I see you've checked that already, but we'd be happy to verify again if you can send along some screen shots of your serial timings and perhaps a snippet of your schematic showing the DAC connections.

  • in reply to Tom Hendrick
    Prodigy 20 points

    Hi Tom!

    No, that doesn't seem to be the problem, though we can not definitely say. The soldering looks OK, because we solder according to RoHS for a few years now and never had problems with this. I suppose we came on a charge of chips from a source we can not trace back completely. We now receive chips from a different distributor and with date code 0, means 2010, and replacing the 2007 ones with the 2010 ones instantly works. So the 2007 chips seem to have some kind of trouble.

    In this case one would reckon that the 2007 ones which were not working on the one board and then desoldered and taken to another type board would break somewhen, but no. They continue to work. Also, the other type boards, assembled with the 2007 chips and soldered the same way, never had a failure.

    We also have this problem with an Atmel microcontroller chip, but not at such a high rate. There, after soldering, in-system programming would not work. Replacing the particular IC with a new one - success. It definitely sounds odd...

    We have of course checked with the assembly company and their soldering profiles etc. and our own ones, too. No chance of definitely finding out. This seems to be the curse of lead-free soldering.