[sdiy] Re: MORE VCO reset time

[Harry Bissell Jr]

Hello All...

Just to clear up some rumors... It does NOT use
a BBD but it looks like it could clock a BBD if you
wanted to.  I don't know the linearity (yet) so I
can't say it would be better than some other
methods...


> > (you will need a lot more to get me NOT to talk)
> 
>    What a tease...you don't even give us a price on
> that option!  :P

As a child (almost histroically long ago - another
geological age perhaps)... I would "not talk" for 
one dollar an hour. The price has gone up.  :^P


> > Several questions...
> > 
> > 1)  What is a good ramp reset time ?   How fast
> have you seen ramp reset
> > be ?
> 
>    A _good_ reset time would be in the neighborhood
> of 1uS.  This is 1/10 
> of a wave cycle at 100kHz, which I would feel is
> quite acceptable for 
> pretty much all applications.  Of course faster is
> better, especially if 
> you plan on doing waveshaping, like I do with my
> VCO, then the 
> reset time is extremely crucial.

I have a reset time of .2us - that is if the
calibration of my test equipment can be trusted... and
it is very old.  with careful triggering I could see
the reset...after a half hour of playing around to get
it on the screen.  I'll try and get a measurement with
one
of the digital storage scopes at work.

>  
> > 2) How many octaves do you expect (in tune)
> without resorting to HF
> > compensation
> 
>    Approximately 15 or more octaves.  I never did a
> complete count of 
> what my own VCO is capable of, but it is staggering.
>  There IS a definate 
> relationship with the reset time/cap size/frequency
> range of a VCO.  The 
> smaller the cap, the smaller the current driving the
> VCO.  I find that 
> this is a distinct _disadvantage_ in the range of
> the VCO.  While you can 
> certainly get fast reset times usng a small cap, the
> low end range of the 
> VCO suffers since it's harder to drive those really
> small currents 
> without running into leakage issues.
>    I opt instead to put in as LARGE a cap as
> possible, because you can 
> run larger currents in the mirror allowing you to
> run the VCO into 
> extremely low frequencies yet still have excellent
> high end response.

Right now I have a .01uF cap (a mylar - yuck !!!).  I
have gone as small as 100pF and got to 5MHz.

One problem I have noticed is that self-heating in the
current source can cause temperature drift... evident
when you return to lower currents as a frequency drift
that eventually settles.  Probably not an issue in
the audio range.

> 
> > (I have not tried hf comp yet...)
> 
>    I don't use it personally.  But probably someone
> with more golden ears 
> than me would prefer it.  I suspect that really the
> HF compensation is 
> there because of waveform distortion at higher
> frequencies.  With a super 
> fast reset time, it shouldn't be necessary.

That is what I thought.  Guess we have to wait for the
linearity result... and that will probably include
the error of the expo itself. I'm using a 2SC798 dual
as the expo transistor.  I opted for the PNP expo
rather than an NPN and a current mirror. I thought
that would be easier to temp compensate...


H^) harry