RC vs RL hpf

Most (or at least many) older HPF circuits were RL based.

Are there any benefits using RL over RC filters? I would imagine that even a 2nd order RC filter would be cheaper than using an inductor. What am I missing?  

Many older HPF circuits were CLC circuits with a third order response. (Neve)

Unfortunately you cannot simply cascade identical CR HPF stages (hoping to get identical poles) unless you put buffers in between them so each stage is driven by a low impedance and loaded by a high impedance. If you do not do this, the interaction between stages shifts the poles so they are no longer coincident. If this is not a problem for you then successive CR stages are fine. The first mixer I sold had a second order HPF just like this.

Cheers

Ian

Thanks Ian,

Can you please explain a bit further the "the poles so they are no longer coincident." part?

Do you mean that the cutoff frequencies will be close but not identical? And what would be the result? Maybe not 12dB/oct but 10dB/oct for example? Any other drawbacks?

thanks again!

warpie said:

Thanks Ian,

Can you please explain a bit further the "the poles so they are no longer coincident." part?

Do you mean that the cutoff frequencies will be close but not identical? And what would be the result? Maybe not 12dB/oct but 10dB/oct for example? Any other drawbacks?

thanks again!

Click to expand...

Yes, that;s right, the turnover frequencies will be different. Difficult to say how close - the analysis is complex- I would prefer to simulate it. The ultimate slope is still 12 dB per octave, it just gets there a little differently.

Cheers

Ian

Thank you Ian, one last basic question if you don't mind The Fc on a 2nd order filter is still -3dB, isn't it? I mean, it's not -6dB...

warpie said:

Thank you Ian, one last basic question if you don't mind The Fc on a 2nd order filter is still -3dB, isn't it? I mean, it's not -6dB...

Click to expand...

Good question. If the poles are coincident then at Fc the response will be 6dB down. However, I suspect a lot of people will spec the filter at the -3dB point.

Cheers

Ian

One way to reduce the interaction is to make the impedance of the second stage of a CR HPF 10 times greater than the first. SO, if the first is 100nF and 10K, make the second one 10nF and 100K. The Fc of both of these is about 160Hz. The simulated -6dB point is 165Hz so it is pretty close. The -3dB point of the first stage is at 173Hz.

if you make both stages from 100nF and 10K you get a -6dB point at 250Hz and the -3dB point of the first stage is 363Hz so you can see just how much interaction there is.

I suggest you try an LTspice simulation yourself to get a feel for what is going on.

Cheers

Ian

ruffrecords said:

Good question. If the poles are coincident then at Fc the response will be 6dB down. However, I suspect a lot of people will spec the filter at the -3dB point.

Cheers

Ian

Click to expand...

That is true, also the Q of the filter plays an important role, specially in active filters.

You lost me there. Are you saying that although the  filter might be 2nd or even 3rd order, a manufacturer will spec the Fc at -3dB?

On a 1073 for example,  does the HPF frequencies on the faceplate represent the -3dB or the -6dB points? (I realise that the 1073 is an LC HPF).

warpie said:

You lost me there. Are you saying that although the  filter might be 2nd or even 3rd order, a manufacturer will spec the Fc at -3dB?

On a 1073 for example,  does the HPF frequencies on the faceplate represent the -3dB or the -6dB points? (I realise that the 1073 is an LC HPF).

Click to expand...

I am saying I don't know but I can imagine it being that way. For example, there is no agreed way to specify a shelving EQ but all manufacturers say shelving at 10 Khz or something similar. So when a manufacturer say HPF at 80Hz it could mean either. Hopefully they all properly specify Fc.

In practice I think they do not. Attached is the frequency response of the 1073 taken from SOS magazine. The lowest mid boost is a 360 Hz so the vertical line to the left of the peak is most likely the 300Hz line. Almost directly below that peak you will see the highest frequency HPF curve. As this is a third order filter Fc will be where the response falls to 9dB which I reckon is a lot nearer to 200Hz than 300Hz.

Cheers

Ian

I see, you are right. It just seems to be a bit odd to spec at -3dB no matter what but it's definitely the case, at least on this example.
Thanks Ian!

warpie said:

I see, you are right. It just seems to be a bit odd to spec at -3dB no matter what but it's definitely the case, at least on this example.
Thanks Ian!

Click to expand...

I suspect you have to think in terms of what recording engineers think. They are interested in the point at which the EQ starts to take effect rather than in the technical definition of Fc.


Cheers

Ian