Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[Benmrx]

Is a High Pass filter the same as a Low Shelving EQ?  I just discovered what you can really do with both Low and High Pass Filters, where everything past your cutoff frequency is completely cut.  So if you use a Low Pass and a High pass in close proximity on the same track, you're left with just a narrow band.  I want to do this in the analog world, but don't see many EQ's that say Low Pass and Hi Pass Filter, instead it's Low Shelving and High Shelving.  I really should already know this, but is it just a different name for the same thing?

Thank you in advance

[Brian Roth]

A shelving EQ is call that because it has three "areas" interms of frequency response.  I'll use a low EQ shelf as an example, in the "cut" mode.

At high frequencies, it does essentially nothing to the frequency response.

At some "cut off" frequency, the response begins to droop.

At yet another, lower frequency, the "droop" begins to level off vs. continuing the droop.  Eventually, the frequency response levels off (ie, "shelves"), but at a lower level than at high frequencies.

A low cut (high pass) filter continues the roll-off indefinitely.

Bri

[Benmrx]

Brian

Cool, thanks for the info.  I'm wondering why you don't usually see variable Low-Pass and Hi-Pass filters on newer hardware EQs.  Usually you get something like a Hi-Pass filter set at 75Hz or something like that, but that's it.  

I had a great time using a Hi Pass Filter on a clean guitar sound, set around 200Hz, and a Low Pass set around 15Khz.  But this was using a McDSP plug, and I would love to have multiple hardware boxes, maybe just the Hi-Low Pass Filters and nothing else.  Any ideas?

The only affordable boxes I can see is the Electrix Filter King or something like that.  Or am I missing something here?

Thanks again for clearing up the difference between a Low Shelf and Hi Pass Filter

[Brian Roth]

IMHO, the lack of variable high- or low-cut filters in desks and preamps is due to cost.  It's not THAT expensive to implement, but the Bean Counters rule the world, and if they can shave a penny or a dollar, then that is the result.

In the desk I built in 1977, I included a variable (sweepable) low-cut filter at the request of the client.  Using a decent pot and components in this day, it would easily add $10 or more per channel, so it's an easy target for the Bean Counters.

Bri

[Benmrx]

Huh, it still kinda surprises me that it's pretty hard to find a piece of rack gear that has a variable low-pass and/or hi-pass filter.  Would this be something fairly easy to make?  I'm pretty decent with a soldering iron, but don't have much knowledge in buiding components.  Know of any kits or schematics?  Thanks again for you're help.

[Brian Roth]

I have "as builts" of the sections of the desk I built, but the filter design was rather specific to the system (low Z input, etc).

National Semiconductor has a nifty online tool for filter designs.

http://webench.national.com/appinfo/webench/filters/design_r equirements.cgi

Bri

[Benmrx]

That's a pretty cool page.  However, that goes beyond my knowledge....forgive me.  I wouldn't trust myself to put a design together.  And I'm sure I'm wrong here, but it seemed that online tool was more for a preset frequency low pass filter, where you decide in the design where the cut-off frequency is and it's engaged with a switch, and I would definately want it to be variable with a pot.
To bad those "as builts" won't work, I would definately need some kind of kit.  

[Brian Roth]

Yes, you would have to run two sets of calculations from the National "calculator" to find R values for the extremes.

As for my filter, here it is, for what it's worth:

http://www.brianroth.com/projects/m77/hi-pass-filter.pdf

The "wire" appearing in from the far left side is the signal input.  As you can see, the input impedance is quite low due to the input pad consisting of the 120 Ohm R in series with the 520 Ohm R.  The low values were chosen to ensure a reasonably low input impedance into the filter network.  Perhaps a better location for the pad (to match the levels when the filter is switched in and out) would be at the output.

With 27+ years worth of hindsite (I was 23 when I did the original design work):

http://www.brianroth.com/projects/m77/m77.html

I see other flaws now <g>.

In no particular order:

The opamp was a TDA1024 (now well-known as a NE5534), and I didn't have a full set of docs to work with.  That 20 pF cap between pins 2 and 3 was to stabilize the opamp.  It should be connected between pins 5 and 8, although the circuit didn't oscillate as drawn.

The 47 uFd output cap is shown as a tantalum.  WRONG choice...use a high quality aluminum, and bypass it with a film cap.

The input impedance becomes even lower when the pot is "tuned" to the higher frequencies (plus the opamp stage works into a low impedance as well).  If you can stand a cut-off below 350 Hz, then increase the values of the 1K0 Ohm resistors.  A better solution would be to rescale the values of the pot(s), series resistors, and 0.68 uF caps to present a higher load impedance through the network.

I don't use tantalum caps any more for power supply bypassing.  When they fail, they fail as a short circuit which brings down the  entire power supply rail.

Like I said, hindsite is 20-20.....hence my reluctance to tell you "this is IT!"  <g>

Bri