Post Go back to editing

Signal output from LT6011 DD Package

Hi everyone,

I have the following schematics:

The V_att is a 1kHz Sinewave input at 1V peak to peak. The R9 is essentially an open connection(no resistor there) and R6 is a short. V_filter is the sinewave I am interested in

The following is the simulation of the same circuit, but without the 3.3V to 5V gain from the second op amp.

Behavior:

+ In my actual circuit based on figure 1, the v_filter is attenuated by approximately 4 times and its average is shifted to about negative 2V

+ In my simulation using LTSpice, the v_filter is exactly the same as the V_att (input voltage V3) and there is no DC bias on the output.

+ Since this circuit is a 2nd-order low pass filter with cutoff frequency at approximately 8kHz, the output (V_filter) should be the same as that in the simulation (figure 2).

+ have replaced the LT6011 DD package chip 5 times and use different copies of my board but the signal output is still the same as that in figure 1.

Question:

+ What could possibly be the reason for this problem?

+ The DD package has a pad (pin 9) underneath the chip which I did have the copper layer for it, but I didn't connect it to anything. Is this the reason for the above issue?

  • Hi:  There is no reason I see for any DC offset.  The circuit is a simple unity-gain non inverting Sallen Key 2nd order filter.

    I'm guessing, though, that the layout has some antennae in it, and you are picking up RF.  The LT6011 is a micropower chip, so it will not be able to reject high frequencies and may rectify them into a DC offset.

    Do you have a way to better shield the circuit with ground planes?  I sometimes resort to battery operation inside a (grounded) cookie tin. 

    Also, does the circuit have bulk bypass on the supplies as well as .1uF bypass caps close to the supply pins?

  • It should not attenuate signal by 4 times either. What I did was removing the Lt6011 and using an LT1037 op amp on a separate breadboard and the problem was completely fixed. I wonder how easily damaged the LT6011 was when using heat gun soldering. I used 240C heatgun with minimal blowing speed.

    I will try the following things:

    1. replacing the Lt1037 with LT6011 8-SOIC version on breadboard and see if the problem fixed. 

    2. I am ordering a different board that use LT6011 8-SOIC footprint instead of QFN and see if the problem fixed.