SV: Re: SV: Re: SV: [sdiy] Re: (sdiy) English lessons

[karl dalen]

--- Magnus Danielson <cfmd at bredband.net> skrev:

> From: karl dalen <dalenkarl at yahoo.se>
> Subject: SV: Re: SV: [sdiy] Re: (sdiy) English lessons
> Date: Fri, 16 Sep 2005 00:06:20 +0200 (CEST)
> Message-ID: <20050915220620.16408.qmail at web25510.mail.ukl.yahoo.com>
> 
> > Halleualia! I always forget something!
> > 
> > The saw and PW has to be in opposite phase for this to work!
> 
> Yes, the trick is really to replace the sawtooths reset with the PWMs reset.
> What happends is that when summing them there will be a CV-dependent offset
> of
> the waveform, but it is compensated in my design, so it should be DC stable
> but naturally there is always a little leakage.
> 
> If you have the inversed PW you get the frequency doubler design, since you
> insert a new reset placement, basically subtracting of the top half of the
> sawtooth. The phase-shift design came as an aftertought of the frequency
> doubler design.

Yes, but in most cases you dont need both designs, since the phase
shift and frequency doubbler are achived in one designs, with one minor
drawback that the amplitude are halved at frequency doubbling.

I mean, thats if one want to go the el cheapo
way otherwise use both designs.

The DC compensation thing was also implemented in the ARP Chroma (CBS)
i dont think its needed really in AC applications, even at fast modulation
speed, but thats just my matter of taste, other people might prefer the DC
compensated version. I remember something that sequential once said about
DC popping sound at fast speeds if uncompensated!

KD

> Both these designs is fairly easy to understand if one draws the waveforms,
> make notes about voltages and is able to sum properly. Simulation is trivial
> actually. If your simulation tool fails, toss it since by brain obviously
> beats
> it. Not a good benchmark, but anyways...
> 
> Cheers,
> Magnus
> 
> > KD
> > 
> > --- karl dalen <dalenkarl at yahoo.se> skrev:
> > 
> > > Folks!
> > > 
> > > You dont need any super specialized clipping things,
> > > nor any 4000gates for supersharp edges (even if its nice to have,
> > > well if you want absolute precision it can be neat to have , 
> > > but the rounding of the saw wave, and retrace time at high
> > > frequencies will "showel bang" it up anyway) no specialized
> > > CV inputs for phase! (who asked for that?) ugh, and ugh again!!
> > > 
> > > You need:
> > > 1: a saw wave, 0 to +5v or 10V in peak unipolar or bipolar.
> > > 2: a PW wave ,0 to +5v or 10V in peak unipolar or bipolar.
> > > 3. add them with 100PW/50saw blend in amplitude and you get saw phase
> > > shifting.
> > > 4: add them 100saw/100PW and you get phase shifting saw wave, 0-180
> degrees.
> > > 5: The sawphase wizzard has spoken,thank you very much! :-)
> > > 
> > > After words:
> > > In any additive wave thing in wich phased saw waves actuaclly are
> > > wave amplitude are the crucical inferno. So dont base your phase
> > > waves on the asumption of stable power supplies! ugh!.
> > > Its very simmliare to the basic saw to triangle converters!
> > > 
> > > > The problem is that the first op-amp acts as a comparator, so that may
> > > > confuse
> > > > the hell out of you if your model does not saturate. The assumed supply
> > > > voltage
> > > > is +/- 15V, which is "all over the schematic" ;O)
> > > > 
> > > > KD tried it ages ago, and then it worked well! ;O)
> > > 
> > > Right, ages and ages ago! Ooh, now im starting to feel very old and
> crippled!
> > > 
> > > Reg
> > > KD
> > > 
> > 
>