Question on Inductor EMI and proper spacing on PCB

Thread Starter

BackyardBrains

Joined Aug 26, 2010
33
Background:
I added a circuit for TI's LM2674 which is a Power Converter High Efficiency 500mA Step-Down Voltage Regulator (http://www.ti.com/product/lm2674) This requires an inductor, which I am using Coilcrafts D03308 (http://www.coilcraft.com/do3308p.cfm). (see: "layout.jpg")

The device I am working on amplifies biosignals (~1 - 10 mV) with an AD623 and the op-amp, TLC2272 to an audio jack. It also amplifies the signal with a LM386 to an 8Ohm speaker. We are using the Voltage Regulator to split the power for the instrumentation amps to be on regulated voltage and the speaker amp would be on the 9V battery. We are trying this because if we turn up the gain on our LM386, we get feedback when these three ICs are on the same power source.

Issue:
With the new board I built up, there is an unwanted hum over the signal. Other than the addition of the Voltage regulator circuit and arranging the board, nothing has been changed.

My guess:
The inductor is creating and EMF that is creating this hum on our signal trace. I read that it was shielded but perhaps it's still too close? (see "issue01.jpg")

My question:
How much space do I need to give the EMF-inducing components so they don't interfere with the sensitive signals? Does anyone have experience with this LM2674 circuit in particular?
 

Attachments

crutschow

Joined Mar 14, 2008
34,280
Hum implies low frequency pickup from the mains power (appearing as a 60 or 120Hz signal). That wouldn't be caused by the switching supply which generates high frequency noise at the switching frequency and its harmonics.

So I would look for any ground loops (more than one ground path between circuits) in all the connections (such as sneak ground paths through any mains safety grounds). See what may have changed in the grounding between the circuit that worked and the circuit that has hum.

If you have feedback with three devices on the same power source then you would seem to have insufficient power line decoupling between the circuits. You might try to add extra decoupling capacitance on the circuits and perhaps a small series resistance between the power input and each circuit to further improve the decoupling.
 

AnalogKid

Joined Aug 1, 2013
10,986
I agree with alpha, a switching regulator seems like too much work for the job at hand. If you can't post a complete schematic, can you tell us what the output voltage and expected current draw is from the regulator? If it's nothing but an inamp, a linear regulator should handle it with fewer issues. And, how many layers in the board?

Feedback often comes down to decoupling, especially with IC power amps, high gain preamps, or both. Sprinkle 0.1uF ceramics and 10 uF electrolytics around like you're seeding a lawn.

ak
 

Thread Starter

BackyardBrains

Joined Aug 26, 2010
33
...can you tell us what the output voltage and expected current draw is from the regulator? If it's nothing but an inamp, a linear regulator should handle it with fewer issues. And, how many layers in the board?

Feedback often comes down to decoupling, especially with IC power amps, high gain preamps, or both. Sprinkle 0.1uF ceramics and 10 uF electrolytics around like you're seeding a lawn.

ak
Thank you AK (&alfa &cruts),
I've attached an image of the board and of the schematic. The regulator's output voltage is set to be 7.43 V and testing with a multimeter, it's output is ~7.38 V. I am not sure about the current that the AD623 and TLC2722 are pulling from it. Our board has two layers.

Could you please provide a more precise picture of what you mean by "sprinkling capacitors"? I understand the basic concept of a decoupling capacitor and it makes sense to add to our circuit here. Although, I am not clear on "seeding my lawn" here, do you mean I would need a pair of caps before each IC's power input? Can you explain why so many would be necessary as compared to one pair near the power source?

I am reading up linear regulators per the recommendations and am going to measure the current draw on this ICs to get a better picture here. Thank you all for your quick replies!
 

Attachments

Top