INA126: INA126 output unstable at GAIN = 368

Hi Nader,

Please post a schematic including the circuitry around your input, output, and reference voltage.

Hi Zak,

The schematic is attached. Thanks

Zak, sorry, it got attached upside down. Do you want me to reattach it ? Thanks

Hi Nader,

take care, the NMH2415 switcher is producing lots of ripple and needs a proper ripple filter at its outputs. Your ripple filter is not adequate. You should add another 3µ3 cap to the other side of 100µH choke so that a pi filter is formed.

Another ripple filtering is needed at the + input of OP27. Connect a 100n cap from the + input to signal ground.

Kai

Hi Kai,

Thanks for the advice. Well taken. Just 2 points for clarification. 1. The output filter of the NMH2415 in my circuit is per Murata recommendation. 2. I measure output of NMH2415, it showed solid +/- 15V. Also, the measurement at pin 3 of the OP27 showed solid output. 3. As I stated in my original message, as I increased RG, the output of the INA126 becomes solid. Considering the above 3 points, do you still think I should add extra filtering. Thanks

Hi Nader,

Typically stability problems get better, not worse with increasing gain. This leads me to believe that there is a ripple somewhere in your system as Kai has mentioned that is getting gained up. If this ripple is on the source for the bridge then this will lead to a fluctuating common-mode which is going to introduce additional error due to the CMRR of the device. You can add the additional filtering as Kai has mentioned or test the bridge with a low noise lab supply to verify if this is the issue.

Hi Zak,

You are absolutely correct, stability improves with gain. Maybe I was not careful when I used the term instability. I think I have the issue of either noise or ripple on my INA126 output. I will do as advised and let you and Kai know about the outcome. Thanks.

Hi Nader,

in your schematic you have reversed the L and C of output ripple filter:

Kai

Hi Kai,

I did not get what you mean by reversing the L and C ! Do you mean the values ? Thanks

Hi Nader,

you have:

But you should have:

Kai

Hi Kai,

I don't quite understand how does it make a difference having cap closer to the PS ? We still have Low Pass filter with this formation. Thanks

Hi Nader,

if the position of L and C is reversed you no longer have a functioning output ripple filter. See the huge difference of damping at the ripple frequency of 80kHz:

Kai

You are correct. I see that. I will correct the circuit. I am hoping that this correction will take care of the output ripple and so I may not need to apply PI filter. We will see. Thanks