Forums

Hi Min,

When you say there is no input signal applied to the LOG114 is the photodiode in place, or not? Is the dc output level at VO5 what you would expect for the said input condition? Is there a load connected to VO5 when you measure the high -5V supply current?

I suggest you take a very high input impedance DMM and measure the dc voltage at each pin and record it. We can then go through the voltages and see if they are appropriate. Also, connect a high-impedance input, wide bandwidth DSO to each accessible amplifier stage output pin; V_LOGOUT, VO4 and VO5. Check for high frequency oscillations.

Regards, Thomas

PA - Linear Applications Engineering

Hi Thomas,

Thank you very much for the quick response. What I said there is no input to LOG114 means the photodide is NOT in place. Since I am testing the boards first on electronic bench before passing the boards to be integrated in the optical system, my default testing setup is to leave the LOG114 input floating ( I have ten of them on each board).

Followed your instruction, I did the voltage measurement on all the pins. Please see the attached file.8171.LOG114 measurements.xlsx

I observed when there is NO input, the voltage at pin11(IN4-) is not zero and the three stage buffer doesn't function properly. When there is an input, the part (three stages) works properly. Could this be the cause of the higher current draw on -5V rail? I did not observe any high frequency oscillation on any pins.

Thanks,

Min.

Hi Min,

Okay, adding the series resistor to the LOG114 I2 input and biasing it from the 0.1 V voltage was a good step. That set up the I2 input for a current of 250 nA and you already had the I1 input set for 100 nA. When we place those values in the LOG114 gain equation, V_LOGOUT at pin 9 should be 0.150 V which is the level you measured. Then when you apply that level to the A4 amplifier operating in a gain of -1.07 V/V its output should be -0.161 V. You measured -0.2 V which I am assuming might be a measurement difference from the ideal level, but this may be an indication of a problem in one of the two user-configured operational amplifier circuits. I would expect the voltage to be very close to -0.161 V. A5 is just a +1 V/V buffer so the voltages at both input and the output pin should be at the same level as A4's output (-0.2 V) which was the case.

When you left I2 floating the denominator in the gain equation approaches zero and V_LOGOUT should try to become huge and as your measurements indicate the output is up against the A3 positive output rail. A4 attempts to amplify that large voltage by the -1.7 V/V gain and its output is locked up against the negative output rail. +IN4 is tied to ground (0 V) and you would expect -IN4 to be very close to that level too. Likely, because the A4 output is saturated the loop has collapsed and the voltage between the inputs is not representative of normal operation.

When the -4.94 V coming from A4's output it is applied to the A5 non-inverting input A5 should reproduce the -4.94 V input level within its capabilities; however, the output is at -0.78 V and that is odd. There may be a problem with A5. If it is damaged in some way that might account for the high negative supply current.

I suggest you take a different board and set it up for the I2, 250 nA condition and see if you get the same result as you had with this board. If the levels are more in line with the expected A4 levels, then observe its negative supply current and see if it is within the specified range.

Regards, Thomas

PA - Linear Applications Engineering

Hi Thomas,

I think the variation at A4 output is caused by the accuracy of the external 0.1V I used. I observed when there was no input at the LOG114, then A4 and A5 didn't behave properly but as soon as I provided input, A4 and A5 worked fine.

I did more tests todays. I have ten LOG114s on each board. I injected the input signals to 5 LOG114s, I observed the board current draw dropped by 40mA(@12V). This gives roughly 14mA(on -5V rail) current draw reduction on each LOG114. 

It looks like the high current draw problem is caused by the method the board is being tested (i.e, leaving LOG114 input floating). The proper way is to provide inputs to LOG114s.

Thank you very much for help me solve the problem.

Regards,

Min.

Hi Min,

Thank you for letting me know about your success finding the LOG114 set-up problem. I am happy I was able to assist you!

Regards, Thomas

PA - Linear Applications Engineering