[sdiy] Re: LVM, or not

[Gene Stopp]

You got it, dude. Another recent thread talked about how an integrator
works, as the op-amp puts out a rising or falling linear ramp in order to
provide a current sink or current source to the virtual ground. The mantra
is: an op-amp will put out whatever it can to make sure the (+) and (-)
inputs are at exactly the same voltage, at all times. Current can't go into
the op-amp, so if it enters the summing node it has to leave somehow.

I remember being a member of a group of production line repair techs in an
all-digital company, and then a new product came along with some op-amps
(around a modem chipset). They could not figure out how a device could have
a complex output with ground at both inputs!

- Gene



-----Original Message-----
From: owner-synth-diy at dropmix.xs4all.nl
[mailto:owner-synth-diy at dropmix.xs4all.nl]On Behalf Of Tim Parkhurst
Sent: Tuesday, November 16, 2004 3:49 PM
To: 'john mahoney'; Roman
Cc: synth-diy at dropmix.xs4all.nl
Subject: RE: [sdiy] Re: LVM, or not


> -----Original Message-----
> From: john mahoney [mailto:jmahoney at gate.net]
> Sent: Tuesday, November 16, 2004 2:18 PM
> To: Roman
> Cc: synth-diy at dropmix.xs4all.nl
> Subject: Re: [sdiy] Re: LVM, or not
> 
> > sure I'm with you! I just thought 5V modular wasn't so popular.
> 
> Cool! My comment was kind of stupid, anyway. :-/
> 
> There seems to be some interest. I'm just collecting people's thoughts
> on it, looking into where the problems might be (e.g., transistor
> ladder filter), etc. Hopefully I (we) can start designing LVM modules
> in Spring 2005. I've got plenty going on for the next few/several
> months, though.
> 
> 
Just in case anyone's interested: I think I've figured out how the LVM CV
summer should work. This may seem simple to some (and it is, once I had
thought it through), but I had to sit and think about this for a minute (any
gurus out there are more than welcome to critique or correct my assumptions
here). This is kind of lengthy, so be warned.

FIRST: My understanding of basic op-amp behavior...
Basically, an op amp summer works by keeping the output equal to the SUM OF
the inverting and non-inverting inputs. If we ground the non-inverting
input, apply a voltage through a resistor to the inverting input, and (this
is important) take the output and feed that back through another resistor to
the inverting input, THEN the output will adjust as necessary to keep the
voltage at the inverting input AT THE SAME LEVEL AS THE NON-INVERTING INPUT.
If, as in my example here, we've stuck the non-inverting input to ground,
then the op amp will try to keep the voltage at the inverting input at 0
volts. It's not ground in the strict sense, but it acts just like it, hence
the name 'virtual ground.' 

Am I correct on this so far?

So if we run 1.5V into the inverting input, the output will go to -1.5V in
an attempt to keep the voltage at the inverting input at 0V (remember we're
summing the 1.5V input and the -1.5V feedback through resistors to the
inverting input). We can add more input resistors and sum more voltages to
the inverting input. The output will still adjust as needed to keep the SUM
OF THE VOLTAGES at the inverting input at 0V (until it runs into the supply
rails, of course). 

Right?

SO, FOR A LVM INPUT SUMMER: We want the modulation based around 2.5V. A
signal less than 2.5V will produce negative modulation (lower frequency,
lower gain, etc), and a signal greater than 2.5V will produce positive
modulation. THEREFORE, all we need to do is change the non-inverting input
on our op amp summer and put it at 2.5V instead of ground! Again, this will
seem dirt stupid basic to many of the experts out there, but I was rather
pleased with myself for having FINALLY figured this out! Now what we get is
that any input (to our summing node at the inverting input) LESS THAN 2.5V
produces a positive voltage at the output. If we put 1.5V at the input,
we'll get +1V at the output (1.5V + 1V = 2.5V which means our the inverting
and non-inverting inputs are equal again). By the same token, any voltage
greater than 2.5V produces a negative voltage at the output. Now we have a
CV response centered around 2.5V instead of 0V. At this point, we either
apply this to an expo converter with an inverted response (our summer
inverted things, remember), or just stuff it through another inverting
op-amp and THEN to the expo converter.


SO, BASICALLY: Changing the 'center point' of our CV modulation from 0V to
2.5V is relatively simple. Just change the 'virtual ground' point, and
you're off and running. 

Well, if you've read this far, thanks! Like I say, this is really a basic
concept, but I wanted to explain it 'out loud' so that 1)I can get corrected
if I'm wrong and 2)I can share my explanation and hopefully help others
understand some basic electronics.

Okay, and now... back to work!!!

Tim (out loud!) Servo