Odp: [sdiy] 1/8th inch questions

[Seb Francis]

Ok, so for a MIDI2CV output, you don't want the voltage to drop by say more than 0.2mV for a 100K load.  This pretty much rules out putting a 100R resistor after the opamp buffer ...

7.5V - (100K / (100K+100) * 7.5V) = 7.5 mV

But does it work to connect the inverting input of the buffer to the CV output, after the 100R resistor, as suggested by Roman?  It seems to me that this would ensure the voltage doesn't drop at all.  And if the opamp can go up to 15V, then for the max output of 7.5V, the current through the 100R resistor and output would be limited to 75mA.  This seems like a good design.

Seb



Gene Stopp wrote:

> It's a good question - the answer's not obvious and it does bring up some
> side issues. Many times a signal is used internally to the module as well as
> being brought out as an output. For example, in a VCO, the core could be
> triangle-based and then the triangle wave is used as the source of the
> triangle-to-sawtooth converter, the sine converter, and the pulse wave
> comparator. If the triangle wave is brought directly out without a series
> resistor, and then you insert a tip/ground cable into that output, and then
> start to patch it into another jack, when the tip of the jack touches the
> sleeve of the (common grounded of course) other jack, the whole oscillator
> will "die". If you are listening to it while you are patching it, you'll
> hear it glitch when you plug in that last cable. Series output resistors
> will prevent this.
>
> Also, no matter what the data sheets say, I just get nervous if I know that
> I'm temporarily shorting mutli-volt signal sources to ground repeatedly. Not
> nervous because of the energies involved, not nervous because I fear
> embarrasment in front of a huge crowd under hot stage lights, but rather
> nervous because I don't want to have to pull the *&%&%$#^% thing apart to
> change a chip every couple months. Homebuilt stuff is sometimes just not
> easy to take apart.
>
> Also with the main pitch CV, the direct (non-resistor) op-amp output is best
> to avoid scale errors due to loading, but this is subject to the same
> "patchcord events". What I do with those is provide two or three output
> jacks, and have a single op-amp buffer on each alone.
>
> - Gene
>
> -----Original Message-----
> From: Seb Francis [mailto:seb at is-uk.com]
> Sent: Tuesday, October 22, 2002 4:46 PM
> To: John Blacet
> Cc: synth diy
> Subject: Re: Odp: [sdiy] 1/8th inch questions
>
> So why use a resistor on the output at all if the buffer opamp can withstand
> indefinite short circuits?  Is it something to do with not glitching the
> power bus?  I guess the normal short circuit situation is simply plugging
> and unplugging jacks.
>
> Seb
>
> John Blacet wrote:
>
> > I believe that this refers to the practice of including the output
> > resistor (1K) in the feedback loop of the op amp. For example, if you
> > have a 100K feedback R and a 1K output R, the actual output is taken
> > from the junction of the two, not the output pin of the amp. Of course,
> > this changes the actual gain, so you have to figure the Rs for that. I'm
> > not sure if this is something you can do with a buffer configuration.
> >
> > In practice, we have opted to use a 100 ohm resistor for the output. The
> > 074X series for example, is rated for indefinite short circuits to the
> > supplies and ground. The 100 ohm is just extra current limiting, but
> > gives us 10X the "fanout" of the 1K typically used.
> >
> > --
> > Regards,
> > --/////--
> > John Blacet
> > Blacet Research
> > http://www.blacet.com