• State
  • Locked Locked
  • Replies 2 replies
  • Subscribers 49 subscribers
  • Views 792 views
  • Users 0 members are here
Support feedback
Options
Options
Similar topics

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.

OPA633 runs hot

Martin Button
Prodigy 20 points
Other Parts Discussed in Thread: OPA27

I am trying to use the OPA633 buffer to replace Intersil HA3-5002-5Z
which is now obsolete. Our product has been in production some time, and we do
not want to re-design it yet. The OPA633 is the only device that I can find that
is almost pin compatible with the old part. The application simply uses the OPA633 to buffer the output of a standard OPA27
op-amp to give a higher current drive. It is included into the feedback loop of
the op-amp in a way that is very similar to the application examples fig.3 and
fig.4 on the device data sheet. The power rails are +/-15V. My problem is that
the OPA633 runs very hot at all times, even when there is no load on the output
and the o/p is just static at zero volts DC. Even when loaded I only expect to
load the output to maybe 50mA. The package temperature quicky rises to over 40
deg C in free air. Am I doing something wrong? I was not expecting to need a
heat sink (difficult with a DIP8 package). I have tried different ICs from
different batches, so I know it is a design issue and not just a defective
chip. Can anyone suggest a cause?

  • 0
    TI__Guru* 93105 points

    Hi Martin,

    First thing, connect a high frequency oscilloscope via a low capacitance 10x probe to the OPA633 output and make sure that you circuit isn't oscillating. Probe the OPA27 output as well. If it is oscillating that can really increase the current.

    Assuming you don't find oscillation take a look at the OPA633 specifications. The typical operating current is 21 mA, with a maximum of 30 mA. If you calculate the quiescent power based on the typical current and +/-15 V supplies, the power being dissipated is Pd = IQ * (+Vs+|-Vs|) = 0.021 * 30 = 0.63 W. That is a fair amount of power to be dissipated by an 8-pin DIP package.

    The thermal junction-to-ambient thermal resistance, theta j-a, is listed as 90 deg C/W. If we calculate the die junction temperature (Tj) using an ambient temperature of 25 C, Tj = Ta + theta j-a (Pd) = 25 C + (90 deg C/W)(0.63 W) = 81.7 deg C. Roughly 57 deg C/W of the 90 deg C/W of the theta j-a, is theta j-c, junction-to-case thermal resistance. Calculating the case temperature Tc = Ta + Theta j-c, Tc = 25 C + (57 deg C/W)(0.63 W) = 61 deg. that is about 142 deg F, which should feel hot. The temperature will be closer to 76 deg C, or 170 deg F when the OPA633 is operating at the maximum current. If the OPA633 is driving current to a load, the dissipation will be higher and package temperature will be higher still.

    So it does appear the OPA633 runs hot normally. The device is designed to operate safely with a Tj up to 200 C according to the Absolute Maximum Ratings table. The Tj equation can be rearranged to find what the maximum power dissipation can be for a set of circumstances.

    It is always better for a semiconductor to run cool if possible, but do know a heat sink shouldn't be required for the OPA633 providing the thermal limits are observed.

    Regards, Thomas

    PA - Linear Applications Engineering

  • 0 in reply to Thomas Kuehl
    Prodigy 20 points

    Hello Thomas,

    Thanks for the quick reply - I fully agree with all you say. I had already noted that the typical operating current is quite high and that's power being dissipated inside the chip. I guess I was just not quite beliving the data - I've never used a chip that ran so hot just as supplied and not under load. The 200 C maximum Tj is 50 C hotter than a 'standard' chip so this one is outside my usual experience. I think you have summed it up when you said it does appear that it normally runs hot. I will carefully evaluate our running conditions and probably try to fit a heat sink. I remain concerned about long-term reliability of the component if the stress level is so high, even if it is within the official maximum ratings.

    Regards, Martin