Part Number: OPA858-Q1
Tool/software:
Title: Op-Amp Behavior Differences After Repeated Use: Could It Be Soldering or PCB Material Issues?
Hello Team,
I am facing an issue with two identical buffer circuit boards, and I’m trying to understand why their performance differs. I suspect it might be related to either the handling/soldering process or PCB material. Here are the details:
Setup:
- Both boards are configured as simple buffers (non-inverting unity gain).
- Input and output are connected to an SMU for testing ( giving voltage as input and for taking output).
- both operated at supply voltages(+5V and 0)
- only opamp soldered on board and nothing else component soldered.
Observations:
- Board 1(where R25 R24 resistors are there ):
- The bias current and offset voltage values are as per the datasheet during initial testing.
- After 2-3 experiments, the quiescent current from the supply exceeds the datasheet limits, indicating possible damage.
- If the op-amp is replaced, it starts working fine again, with values matching the datasheet.
- Board 2:
- The bias current and offset voltage values are much higher than the datasheet values from the start.
- After 2-3 experiments, this op-amp also starts pushing a lot of quiescent current beyond the specified limits.
- Replacing the op-amp is giving the results as buffer but the values are not as per datasheet ; it stays the same.
Questions:
- Why does one board have values as per the datasheet and the other does not? Could this be due to soldering issues, handling during assembly, or PCB material differences(one board made of FR4 and one board made of Rogers)?
- I’m wondering if the PCB material (ROGERS vs. FR4) might be influencing the op-amp’s performance. Could material differences be causing inconsistencies in offset voltage or bias current?
- Why does the op-amp start drawing excessive quiescent current after a few experiments on both boards?
- Could this be a result of thermal or electrical overstress during testing?
- Is it possible that damage occurs in the op-amp after several cycles, or is the op-amp being exposed to conditions outside of its safe operating range, even if it works fine initially?
Could this behavior be due to
handling or soldering issues ?if soldering might be the issue can you please tell me what precautions to take care(Temperature etc). For example, is the op-amp being damaged during assembly or subsequent handling? while handling i am making sure of ESD protection.
so sir can you please help me where it is going wrong in simple buffer mode?
Greetings,
Please note that the specifications in the datasheet only apply when the stated test conditions are met.
Because the amplifier is only stable when configured for gains of 7 or greater, operation as a unity-gain buffer is not allowed.
Also, please note that the amplifier’s common-mode input range is quite narrow. If configured as a unity-gain buffer, the range of test inputs shown above would exceed the input common-mode limitations of the device.
1 Like
To clarify, I was using the buffer configuration to investigate the source of the inconsistencies between the two boards, as I initially observed unexpected behavior even at gains >7 V/V , the output is not as expected.
- On Board 1, the offset voltage and bias current values matched the datasheet during initial testing in buffer mode.
- On Board 2, the offset voltage and bias current values were outside the datasheet specifications even during initial testing in buffer mode.
- I understand that the unity-gain buffer configuration is not recommended due to stability and input common-mode range limitations. However, I am still puzzled as to why:
- The results differ between the two identical boards.
okay may be the reason for excessive quisent current is i am operating out of offset voltage , but why the results differ between the two boards.
can you please help me with this issue .
It has been stated that the two boards in question are made of different materials, so it is not quite correct to say that they are “identical”.
Does this mean that there is no local filter/bypass capacitance for the power supply provided?
High-speed amplifiers like these require special care and attention. They are capable of amplifying and oscillating at speeds into the GHz range, which is well beyond the bandwidth of most test equipment, especially equipment designed for precision DC measurements.
Different PCB materials can have different dielectric constants. For two boards with the same geometries, this can result in different parasitic capacitances, different trace impedances, etc. In the case of high-speed devices like this, effects of that kind can be quite significant. It is a thing worth considering as part of an explanation. Note however that the Rogers corporation makes a variety of PCB materials with different characteristics, so identifying which particular one is in use would be useful.
My advice would be to configure the amplifier exactly as indicated in the datasheet, measure it only within the indicated range of common-mode inputs, and evaluate the test setup from an RF perspective. The test leads one is likely to be using will look like full-wave antennas in the range of frequencies where the amplifier is likely to oscillate.
2 Likes
thanks for the reply team,
- Sorry for the confusion earlier. The results I provided were obtained with decoupling capacitors soldered at the supply. In this setup, only the decoupling capacitors at the supply, one terminal block to provide input from the supply, and the op-amp are present on the board. No other components are soldered.
- At the time of soldering is it possible that the component non linearities are changing ?Are there specific soldering precautions I should take to avoid damaging these components?