low pass filter for DSD

[abraxalito]

Looks fine except you've omitted the 50R load resistor.

 

[merlin el mago]

Now?

 

[abraxalito]

Its still missing - if your source has internal impedance of 50ohm you don't need R2. You should though check that whatever source device is on the Amanero doesn't mind seeing a 50R load.

 

[merlin el mago]

Now?

 

[abraxalito]

Yup - what's the power supply for the output device on the Amanero? If its 3.3V then the peak output current will be 66mA - few logic families can deliver that so the output's going to sag.

 

[merlin el mago]

Yes Amanero PSU is Salas BiB 3.3VDC 400-500mA I don't remember exactly the mA.

 

[abraxalito]

Perhaps better for the output logic device if you AC couple the 50R to 0V with a large electrolytic - 2200uF/6.3V should do fine. We don't need termination all the way down to DC.

 

[merlin el mago]

I use 10uF 200V MKP Auricap to DC blocking, is enough? or you mind in parallel with 50R load resistor?

 

[abraxalito]

No, I mean in series with the load resistor. I don't know if 10uF is enough, simulation will tell.

 

[merlin el mago]

Now?

 

[merlin el mago]

No, I mean in series with the load resistor. I don't know if 10uF is enough, simulation will tell.

I'm sorry I don't know how to use simulator.

 

[MarcelvdG]

No, I mean in series with the load resistor. I don't know if 10uF is enough, simulation will tell.

With 10 uF somewhere in between a 50 ohm source and a 50 ohm load, the AC coupling time constant is 10 uF * (50 ohm + 50 ohm) = 1 ms, giving you a first-order roll-off below 159 Hz. I'm neglecting the impact of the other filter components now, but I think that's OK as they presumably only kick in above 20 kHz.

 

[merlin el mago]

Marcel the 50 ohms source impedance of filter is OK but mine loads have 100K.

 

[MarcelvdG]

With a circuit like this you get a cut-off around 159 Hz. I gathered abraxalito wants the DC blocking capacitor somewhere between source and termination resistor to keep the source's DC output voltage from causing DC current to flow.

 

[merlin el mago]

abraxalito?

 

[abraxalito]

You called?

 

[merlin el mago]

Yes, thanks. What's the final filter?

 

[abraxalito]

Its as I said before - 2200uF in series with the 50ohm but no need to have it in the signal path as in @MarcelvdG's drawing, rather put it between the 50ohm resistor and 0V.

If you prefer 10uF rather than 2200uF it might work but you're the design authority for that change, not me

 

[merlin el mago]

As attached schematic?

What's the purpouse of 2.200uF cap? Dosen't affect the output caracheristics impedance of filter?

 

[abraxalito]

Yes your schematic looks correct. The 2200uF is in circuit so that there's not a DC current flowing through the 50R resistor which is just wasteful. We still need to terminate the filter at AC (20Hz and up), where the 2200uF looks very low impedance so barely affects the 50R loading.