Post Go back to editing

AD831 Bias Current

Hi again,

I am having some problems adjusting the bias current of the AD831. According to the datasheet a resistor must be connected between the BIAS pin and Vp. And for minimum bias current, connect it directly without resistor.

The problem is that the BIAS pin is connected directly to Vn (confirmed with a multimeter). If a shortcircuit Vn to Vp... nothing good happens , and if I put a small resistor (100 ohm) the overall current is increased obviously (10V difference in a 100ohm resistor -> high current).

Anyone know how is the bias current generation implemented in the chip? (and why is shortcircuited to Vn?) Or any experience reducing the bias current?



  • Is this happening on multiple parts? 

    I'm wondering if the ESD is damaged somehow...


  • Hi jdobler, we are finally replacing the part, we'll see...

    They have asked me though to check if the design was right, before mounting the new part. Let me ask you a few questions:

    1. The LO input was dc-coupled (0V) with a center-tapped transformer (being the center tap connected to ground). This didn't work, the signal was strangely deformed. I took out the transformer and replaced with a couple of capacitors (ac-coupled), and it started working. Strangely the DC of the signals was around -2.5V.

    2. The RF input was dc-coupled (0V) with a center-tapped transformer, and working. Is there any problem in having one dc-coupled and the other ac-coupled? (and is one preferred to the other?)

    3. Because I don't use the output amplifier, the AN&AP inputs are unconnected (which I think it shouldn't be a problem), and all the pins from the output amplifier are unconnected too (I wonder if I should connect VFB to OUT, and COM to gnd, even if I don't use them). The datasheet doesn't give a clue.

    Thanks for your support. I think some information is missing in the datasheet in this regard

  • Hi Julio,

    I'm guessing you're running a dual supply, +/- 5V?

    1)  The LO input should be able to be biased to 0V when using dual supplies.  Have you tried driving it single ended like what is shown in Figure 9?

    2)  No you can DC couple the RF input, just make sure the common-mode is between -1V and +1V.

    3)  You should be able to leave the unused pins unconnected.

    Can you attach a schematic?  It would better help me understand exactly what you're doing.



  • Yes, I should've started uploading the schematics.

    The problem I found was with T27. I had no signal. So I replaced it with two 100nF capacitors (between pins 3-4 and 6-1), and it started working.

    On the other hand, they just communicated me the AD831 that arrived has no continuity between BIAS and VN (that's a good sign). Today or tomorrow I'll have it mounted, and I'll know if all my problems were only related to a damaged part.

    Thanks Joel!

  • Schematic looks OK.  Do you need differential drive?  The part works with a 50 Ohm shunt across the RF and LO inputs.  You could probably forgo the balun on the LO port due to the LO limiting amplifier chain.