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.

TL071: TL071CP slew rate doesn't meet spec from data sheet

Part Number: TL071

We use TL071CP op amps for our ECE lab exercise.  Students measure the slew rate by setting up a voltage follower and inputting a 5Vpp square wave starting at 1kHz and then increasing it until the output rise and fall times are noticeably different - usually in the 100's of kHz.

In the past the parts that I have used from TI meet the spec pretty close to 13V/us.  But the latest shipment of parts bought directly from the TI store show that the Slew rate is around 0.5V/us.  I've checked several of the parts from the latest shipment and they all fail SR spec.  There is a different package for the latest shipment.  I attached an image:  The part on the image at the top with just the dot is from an older purchase and the SR is fine.   The part at the bottom with the half moon has a slew rate of 0.5 V/us.

I've contact TI customer support and spent a day with a person who kept telling me that they were possibly counterfeit, even though I kept telling the support person that they were purchased directly from TI.  Once i finally found the invoice number, the TI support person believed me and then told me that another group will now be responsible for helping me.  That group contacted me and said to investigate this further, they would have to do a FA analysis, which I understand , but they asked me to fill out a ridiculously long form.  So while I am waiting for their response to some questions , I thought I would ask here:

1.  Is it common that the measured slew rate would be low but other aspects of the TL071CP op amp appear to have normal behavior?  i.e. gain for inverting, non-inverting and differential configurations are all correct.  The Bode magnitude plots also seem to agree with the specifications for GBP. My input signal magnitudes are much lower (100s of mV pp sinusoidal signals)

2.  What would cause an internal failure in slew rate inside the op amp.  The internally compensated capacitor not charging correctly?

  • Hello Dale, welcome to the e2e forums.

    If all tested devices from new lot act the same then I would not suspect any damage. It seems that only the 5Vpp slew test is having an issue. So I'll focus on this test.

    The 5Vpp, what are the low and high level voltages?

    What are the two supply voltages? Pins 4 and 7 

    Are pins 1 and 5 open? If unsure, then bend those pins up before placing in test bread board or socket.

    By follower, do you mean input on pin 3 and pins 6 and 2 connected together?

    What type a probe is used on the oscilloscope?

    Lastly can you swap old and new in exact same setup and see this difference?

  • Hi and thanks for the quick response.  Attached is the procedure we use in the lab.  Sorry it is a 10Vpp square wave input from a function generator that has a common ground with the op amp (+5V magnitude and -5V magnitude).

    The supply voltages are +/- 15V.  The null offset pins 1 and 5 are not connected.  Non-inverting input with the inverting pin shorted to the output pin.  We use a standard 1x probes (nothing fancy) that is dc coupled and measure Vin and Vout

    And the last question is yes;  I set up a breadboard test and can swap out the op amp parts.  I tested 5 half moon parts and they all gave low slew rate results.  I swapped out and tested 3 older TL071 parts that have just the dot on the package with the exact same setup and they all tested fine.

    TL071slew rate measurement.pdf

    Thanks!

  • Dale,

    Because these devices are used for learning purposes, can I assume that learning why there is a difference may be more valuable than having parts that work the work they should? If not, the rest of the message won't make much sense. Do you also have a source meter or a current source? 

    If these are TI parts, there is a 0.00000001% chance we put the wrong die inside. X ray or de-capsulation could be a quick way to know, but few have that option on hand. 

    Alternate way to check, is characterizing the natural internal diodes (and natural series resistance) that are pin to substrate. Power off test. So pin 4 would be ground and and inject -10mA (clamp -2V) into each pin (one at a time) and measure the pin to pin 4 voltage (try to get a decent Kelvin connection for the voltage reading). This is temperature dependent so try to read quickly. Voltage may drift lower with time. Make a table for both lots. Compare and post the results too, please.

    Another method is to characterize slew rate versus input difference. In this case ground one input with a small resistance [like 1k] (I discourage connecting an input directly to ground with dual supplies). Power +/-15V is fine here. Connect a small square ware to other input with a variable amplitude.  Slew rate should vary with square wave amplitude for all amplitudes of +/-100mV or less.  I'm sure you seen this in the slew rate FAQ.  The older TL071 will have roughly 1/8 of the full slew at 100mV.  For bigger input amplitude, older TL071 slew will rise until about +/-800mV then slew rate will settle to the full fixed value. I wonder what the newer TL071 will do.

  • Dale,

    I actually have have an easier test that is in the data sheet.

    See section 9.3 Unity Gain Buffer , check that the result is same as figure 9.4

    https://www.ti.com/lit/gpn/tl071

  • Yes. Learning why there is a difference is the goal.  I will try some of these approaches soon! Thanks!

    I've also performed a full power bandwidth analysis and compared the results with the data sheet slew rate value.  In that test I basically input a sine wave to the same unity gain op amp circuit given above and using the known slew rate determine the frequency when the sine wave becomes distorted using the formula:

    Slew Rate  = 2*pi*full power bandwidth frequency * peak sinusoidal voltage

    I've done it on the older parts.  Attached is a plot for the older parts that had a slew rate of ~12.5V/us and based on a 5Vpeak sine wave (10V pk-pk) gives a full power bandwidth frequency of ~ 400kHz.  Attached is a oscilloscope capture at 500 kHz which is approximately the frequency where it started to show lots of distortion and agrees with the simple calculation.  It's a little more qualitative, but it gives a general idea of the frequency range that the output waveform will be distorted due to slew rate effect. I haven't tried this for the newer TL071CP parts yet.

    Thanks again for the suggestions!

  • Dale,

    Keep me updated. I should receive the return unit when available.