Servo Noise Decoupling

[Pauzed]

Hi, I decided to have a go at servo noise on some cheap plastic SG90s.
TBH I haven't breadboarded this even, I just jump straight into KiCAD as I've just received some photo sensitive PCBs today.

I found these:

and

What caused this noise in these servo signals and how do I properly describe the solution?

So in KiCAD I made these:

servoboard_01_01-B_Cu.pdf (2.9 KB)

So...

I'm inexperienced and uneducated in electronics, this will be my first PCB too...

Are the decoupling capacitors connected correctly, especially on the signals?

Any more wisdom and experience will be greatly appreciated!

Cheers

[LarryD]

Remove C2.

[Pauzed]

Why? (Just not needed?)
Also wouldn't that apply to C1 and C3 too?

[Willem43]

I think @LarryD is talking about C2 in the first circuit diagram.

It is not necessary to decouple the servo signal, it will probably degrade the signal. I have never done it.

[LarryD]

@Willem43

I think @LarryD is talking about C2 in the first circuit diagram.

Yes

[Pauzed]

Ah, in the research.

I also just looked in the thread post too and the response also only indicates the power line.

This is why I posted, it looked / felt odd.

Brilliant, many thanks !!!

[groundFungus]

@Pauzed

Why? (Just not needed?)

A cap on the servo signal will mess with the shape of the PWM pulses. You use a cap to smooth out pulses, not what you want on a control pulse train.

[Pauzed]

True! I half remember on a DSP video now that you mention it, thankyou

[Pauzed]

FYI

servoboard_01_02-B_Cu.pdf (2.7 KB)

Cheers all!

[groundFungus]

It does not hurt to put a 0.1uF cap in parallel with the 10/100uF caps on the servo supply rails. The higher value caps (10/100) are good for supplying a bit of extra current when the load changes (low frequencies). The low value caps (0.1) are good for shunting higher frequency noise (ie. motor brush noise) to ground.