THS4631DDAEVM: THS4631DDA as current to voltage converter overheats after few seconds

Part Number: THS4631DDAEVM

Hi to everyone,

I've just bought THS4631DDA (from radiospare) to convert current to voltage and then sum and amplify to ouptut. But when i power then (+12,5V / -12,5V), the device overheats after few seconds. You can hardly handle the package as it's really hot. I really don't understand why as input is grounded to test, so there is no power to dissipate except the device itself. Pad has been connected to ground directly, and with thermal silicon compound but nothing changes. Do i make something wrong ? I've tried with other pin to pin compatible op amp and there is no problem ...

Thanks

  • Hi Francois,

    Would it be possible for you to attach your board schematic to help understand the issue better? Also, how are you converting the input current to output voltage? Do you use a photo-diode and if yes, what is the part number of the photodiode?

    Best Regards,
    Rohit
  • Hi,

    here is the really basic schematic i've used to draw my PCB.

    There are four current to voltage converters stages, then two summer stages to output.

    Input current is 50 µA max

    Output voltage is +10V to -10V

    The circuit is designed to use HAMAMATSU improved tetra-lateral type two-dimensional PSD referenced S5990-01

    I've tested THS4631 on my PCB board, then, as it was overheating, just with soldered wired on power supply pins, non inverting pin to ground, and inverting pin to output pin (follower) but same problem.

    Hope you will find an obvious explanation :)

    Thanx

  • In reply to Francois MARIONNET:

    Hi Francois,

    It looks like the Hamamatsu photodiode cap (~150pF) connected at the inverting input of the THS4631DDA could be causing the op-amp to oscillate, which could be the primary reason for over-heating of the package. Looking at Figure 1 below, the THS4631 op-amp open loop gain intersects the noise gain (1/Vfb) of the device at greater than 40dB/decade slope for feedback capacitance (Cf) of 1pF and 10k feedback resistor (Rf). As a result, the input to output transfer function has 11dB peaking which indicates the device could be oscillating (Figure 2).

    So, the fix here is to increase the feedback cap (Cf) from 1pF to 5.6pF to get rid of this oscillation, as seen in Figure 2. I think this should take care of the overheating issue of the THS4631DDA package.

    Best Regards,

    Rohit

    Figure 1: Open-loop gain and noise gain curves

    Figure 2: Trans-impedance gain comparison for Cf = 1pF and Cf = 5.6pF