Bi-pulse convertor (ON-topic)

[harry]

The term I've heard is "phase reversal"  (not inversion).
Linear Technology has some op-amps that are advertised
as "No Phase Reversal"

for any newbies... Phase reversal can happen with some op-amps
when they are hard overdriven, or operated with no feedback
(like a comparator).  When these are overloaded. (lets say with a
positive input...) you expect them to go to the positive power supply
rail and stay there!

They don't... they first go positive, then run for the negative rail and stay
there (latched) or "stick" there until long after the inputs have returned to
normal.  The inputs say the output should be Positive, but the chip refuses
to cooperate.

This usually happens when the "Input Common Mode Range" is exceeded.
Sometimes data sheets warn of this... sometimes it is up to you to know
if you are safe.

The 741 (1458) op amps NEVER do this. They can be overdriven into hard
clipping very politely.

The OP-07 does not... but the "improved" OP-77 DOES.  I learned this lesson
at $1000 per unit. (OUCH!).

Whay can you do... build an active limit circuit that takes effect before you
reach that point.  A zener diode, bridge rectifier, and two resistors can make a
very symmetrical
clamp that is fast to boot. There are also tricks using two transistors (1 NPN -
1 PNP)
but its a little harder to guarantee symmetry.

Or use an op-amp that does not "DO" that.

H^) harry

Patient: "Doctor... It hurts when I do this...."
Doctor "Then don't do that......"

danial stocks wrote:

> >
>
> >
> > > >Nope, but let's make one... this is done by using three op-amps and
> > > >four resistors...
> > > >
> > > >                           |\
> > > >                       +---|+\    ___
> > > >  CV ------------------|-+ |  >--|___|-+
> > > >          ___     ___  | *-|-/    1k   |
> > > >      +--|___|-*-|___|-* | |/          |
> > > >      |   10k  |  10k  | |             *---- Out
> > > >      |        |  |\   | | |\          |
> > > >      |        +--|-\  | +-|+\    ___  |
> > > >Saw -*           |  >-+   |  >--|___|-+
> > > >      |        <|-|+/    +-|-/    1k
> > > >      |           |/     | |/
> > > >      +------------------+
> > > >
> > > >That ougth to do it. Throw in buffers as required.
> > > looks good but I'd add in inverse parallel diode clippers after the 1k
> >res,
> > > then mix in an op amp summing stage... this circuit is making the
> >assumption
> > > that the + clip on the op amp occours at the same level as the - clip..
> >and
> > > this is not always the case [LM324.. etc ..]
> >
> >Right, it assumes balanced clip-levels. There are many ways to handle
> >this, but in this particular schematic I went for the quick-and-durty
> >level just to show how easy it can be. It is not a precission
> >curcuit for sure. It is just sufficient to explain how it could work
> >and it can also be made to work.
> >
> >If you assume that the +clip and -clip is fairly stable, then
> >individually trim the levels from each op-amp/comparator and then have
> >a common offset trimmer should improve the situation alot. Possibly up
> >to the level where nobody would preceive much audioble difference.
> 3 trimmers or 4 diodes? I'd take the diodes myself.. less stuufing around
> calibrating.. but a sep lvl for the + & - pulse setting could be interesting
> as a user controls...
>
> >
> > > also probs with phase inversion?
> >
> >Phase inversion? Please elaborate!
> >
> never mind about that .. I was thinking back to over driven TL0.. phase
> inversions and looking at this with no feedback path in the op amps.. high
> voltage between the inputs.. prolly nothing to worry over, just out of bed
> so I dont fully got my head on just now..
>
> > > OTOH a comparators should work well here, but the o/p res should
> > > be high to the op amp in mix,...
> >
> >The two output op-amps could very well be comparators.
> >
> > > also, feeding with a triangle wave will give interesting other
> >waveform..
> > >
> > > _           _
> > > | |___   ___| |___   ___
> > >       |_|         |_|
> >
> >Oh dear...
> >
> >Have anyone considered what happends if you feed diffrent waveforms
> >into this little device? Or maybe just diffrent frequencies...
>
> dont reckon different freq's would change too much  in the waveform unless
> you get to freq limitations of the op amps [pull out all you vintage pre 741
> op amps?] but this could be the basis of an interesting new fx thingo...
> like a distortion, maybe then feed the o/p thru a filter, nothing too
> resonant, just round off the corners a little or something.. maybe you put a
> envelope follower to generate a CV off the incoming audio signal so that the
> volume modulates the pulse width.. ok thats it Im gonna try it out...
> Cheers,
> Dan
>
> _________________________________________________________________________
> Get Your Private, Free E-mail from MSN Hotmail at http://www.hotmail.com.