DOME-rectifier...

[puretube]

...or Chico-follower :idea:

I was off in la la land thinking of the Dome filter strictly in terms of another project I have cooking, making an envelope follower from dome filter quadrature outputs. My impression of this circuit was that it provided 2 outputs 90 degrees apart. Sort of akin to a Hilbert transform. I thought I could full wave rectify the quadrature outputs, connect the peaks and have twice the resolution of a full wave rectified signal.

great idea from the TZF-thread!!!

now if you`d take 4 outputs 90 degrees apart....

maybe even 2-way rectify each of the 4 outputs and add the whole bunch:

you`d get rid of a lot of the ripple, and get a much faster response....
(higher lo-pass filter/shorter time-constant)

ok, - a little higher part count, but much better behaviour for certain purposes, me thinks...

Thanks to Tom (aka Chico)!

[troubledtom]

ton,? right?
   the stuff you have been posting is really cool !!!!!!!!!!
     peace,
         -tom
ps: pruetube rocks , i know some guys at EH, they like my stuff too
   www.troubledvariance.com

[Chico]

Puretube:

You are reading my mind on this one.   Truth be told, my main inspriation was to try to take a shot at cobbling up a modern day modular guitar rig in the spirit of the AMS100 from C. Anderton's device articles.  

I was thinking about envelopes and interesting ways to use them, and figured that would make a core component of the modular rig.   Thats when I stumbled upon Juergen Haibel's implementation of the dome filter.  I next downloaded the Bode's patent, comingled a little dsp theory on implementing digital compressors, and that's where I am left.  I figured that full wave rectifying the quadrature outputs (of the dome) would produce a mighty fine envelope.  Four outputs would be even cooler.  If my initial computations are correct, a guitar friendly quadrature output filter could be constructed using a very reasonable number of op amps.  Four outputs would cost more in complexity, but it may be worth the journey.


This is next on the board as soon as I finish the flanger diversion (which is also going into the modular rig).  Did anyone see the ADA inspired envelope controlled flanger coming from my initial comment?


Best regards

Tom

[puretube]

the 2 extra outputs are just 2 inverters away: 180 out of the 0 degree,and 270 out of the 90 degree...


Re: did I say "for certain purposes"? :wink:  :wink:

[Paul Perry (Frostwave)]

I don't know whether the dome filter idea is worth it for an envelope follower, but, if it is, you can save some parts, because you only need one rectifier.  Take the 4 different phases, and run them to the inputs of a maximum value circuit (that is, via 4 diodes to the input of the rectifier).

[wayne kirkwood]

I realize that this is a really old post but I've been experimenting with Dome filters and using them with Blackmer RMS rectifiers to sum sin²+cos² for an essentially ripple free envelope detector.
https://www.proaudiodesignforum.com/forum/php/viewtopic.php?t=1361

There's also a post on the Studer 90° Dome Filter here: https://www.proaudiodesignforum.com/forum/php/viewtopic.php?t=1055

Blackmer's proposed sin²+cos² detector from the patent: https://proaudiodesignforum.com/images/ ... 681618.pdf



This could easily be implemented with a pair of THAT4305.

[Rob Strand]

If you look at a three-phase rectifier you effectively get low ripple due to the phase shifting alone, without the sum of squared stuff.

With a 90 deg phase shifter you can get a similar effect, a two-phase version.   For a full-wave rectifier you get dips in between the peaks but adding in a 90deg phase shifter it fills in the dips like the three-phase case.

In common envelope detectors with RC filters you get more ripple at low frequencies than high frequencies.  Because the phase-shifters have limited bandwidth I once experimented with an all-pass filter around the low frequencies to reduce low frequency ripple.   Some correction is required for the frequency response.


FWIW,  I haven't experimented with the Phase-shifter in the Blackmer patent.  It never looked right to me.  No design criterion are given.  Also the IEEE references aren't correct.

[R.G.]

IBMs mainframes used three phase power not least because with three phase, a full wave rectifier (six diodes) gives ~ 5% ripple with no filter caps at all. This was serious in the bad old mainframe days in the early 70s because the logic in some of those things used a supply of 2-ish Vdc at 200A if my memory of the grizzled old mainframe guy's story is correct.

One approach is, as Rob says, make two phase (90 degree shift) or alternately, three phase with three shifters of 120 degrees. The shifters need to be broadband shifters so all frequencies get shifted, but that's about four opamps per shifter. A dozen opamps seems like a lot, but it's really only three quad packages. Another quad for the rectifiers and you have a signal rectifier without a lot of ripple and without long time constants.

I personally have an aversion to analog squaring/square root devices. Maybe I just never found a good one, but they were always fiddly and drifty.

The peak-detector-and-hold from Mark's friend seems like a clever solution too.