• State
  • Locked Locked
  • Replies 2 replies
  • Subscribers 48 subscribers
  • Views 510 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.

LOG114: "printed" board consideration -- Must I connect thermal pad to V- or not? Why LOG114 could output only "rail" values?

Lev Serebryakov
Prodigy 120 points
Other Parts Discussed in Thread: LOG114

First question:

I'm developing device, which uses LOG114 log amp. I'm hobbyst, so I need to solder this chip "by hands" with hot air resoldering station, without soldering mask and "professional" equipment (my home technology allows me to make accurate enough board for this package, but it is pure copper, without mask).

Should/must I connect thermal pad to V-? Datasheet says "Yes", but user guide of evaluation board contains top copper drawing and it doesn't look like here are connection between thermal pad and V-.

Second question:

Now I have "breakout" board, which converts QFN16 to DIP16-like package, and I have strange behavior of LOG114 on it: there are NO connections between any two pads/pins, but LOG114 doesn't work properly on it. Ref source gives stable 2.5V, but LOG out is near V+ or V- without any in-between values no matter which currents are connected to I1 and I2 (via resistors from VREF).

I use+/-5V bipolar supply from galvanic batteries now, for experiments.

  • 0
    TI__Expert 6930 points

    Hi Lev,

    Thank you for your questions regarding the LOG114...

    The thermal pad must be connected to V- as per the data sheet. 

    The LOG114 EVM was designed for single supply operation, therefore V- in the case of the LOG114 EVM is ground.  As far as the User's Guide is concerned the connection from thermal pad to the ground plane is there although it may not be very obvious.  Thank you for bringing this to our attention.  Schematically in Figure 14 the connection from GND to ground is representaive of the thermal pad connection.  Additionally in Figures 16 and 17 the four "dots" under the LOG114 connect the underside of the device (ie thermal pad) to the top layer and bottom ground plane.

    With regards to your second question I strongly suspect that the currents flowing into the inputs of the LOG114 are not what you expect...for example if one of the currents was very close to zero either due to improper connection or through a leakage path the output will end up at one rail or another.  The board should be very clean to prevent leakages.  One thing to try is to increase the currents into both I1 and I2 if you can...increase the currents to about 10 or 100 micro amps...in this case board leakage should not be as much of an influence...if you setup both inputs to be the same the output should be near ground in the case of dual supplies and everything else referenced to ground, which I presume is how your configuration is arranged.  If you are still having issues, please reply with a schematic...I am happy to review it.

  • 0
    TI__Guru* 93105 points

    Lev,

    It is difficult to manually install and solder the LOG114 QFN package on a PC board using manual techniques. When we manually install one on a test board we dab solder paste on each of the PC board pads. The device is positioned on the paste and placed under the hot air station's nozzle. If the nozzle can be moved it may help to circulate the hot air flow around the device contacts and PC board pads to spread the heat. The hot air is applied until the solder melts and contact is made between the device contacts and the PC board pads.

    The LOG114 backside thermal pad must be connected to the most negative supply level. The EVM is a single supply arrangement and the thermal pad is connected to ground in that case. For your dual supply connection the thermal pad must be connected to -5 V. The negative voltage biases the die substrate and keeps the internal transistors isolated.

    I suspect that you have a connection problem somewhere in your assembly. The LOG114 will perform per the data-sheet if it is followed. It is easy to degrade the LOG114 performance with a bad layout, or poor circuit practices.

    Regards, Thomas

    PA - Linear Applications Engineering