MOTM

>So, this mixer would have 6 or 7 inputs-- four AC or DC reversing and 2 or 3
>DC inputs with at least one without any attenuation.  So, you have 4
>reversing knobs and one or two non reversing knobs.   The mixer would have 2
>or 3 outputs.  One would be the high accuracy "DC-only" outputs (and maybe
>an inverted output).  The other output would be a combination of ALL inputs.
>You could put a master level control on this output too.   So, you could
>have all the features I hear people wanting.  High accuracy DC (3 in 1 out),
>reversing inputs for AC or DC, ability to mix AC and DC, and the DC does not
>get compromised with the "drift" or other bad things until the final output
>op amp that is selected for audio performance. You have the good 1013 DC
>output before you get to that stage.

Maybe I'm missing something here, but do you need more than one
opamp for gain make-up?  I thought the goal was the less the better.
Personally I'd like to see all the inputs treated equally.  If
people want AC coupling on a few of the inputs they can tack solder
a cap on the input jack.  Or use TRS jacks - half-inserted would
be AC coupled, all the way in would be DC coupled.

-Cary

----- Original Message -----

From: Cary Roberts <cary.roberts@...>

Maybe I'm missing something here, but do you need more than one
opamp for gain make-up?  I thought the goal was the less the better.
Personally I'd like to see all the inputs treated equally.  If
people want AC coupling on a few of the inputs they can tack solder
a cap on the input jack.  Or use TRS jacks - half-inserted would
be AC coupled, all the way in would be DC coupled.
----------

List - Sorry for the DIY content, but I know no other way to describe what I
am thinking.

Cary,
I think you are missing that NONE of the op amps that could be selected for
AC would be good for high accuracy DC (at least that is what I read - not
trying to put words in anyone's mouth).  In my suggested design no signal
would go through more than 2 op amp stages (minimum if you do any summing
anyhow).   Total op amps = 1-1013 and 2 1/2 OP-275s or similar.  The
difference with what I am suggesting is 2 different outputs.  Output 1 for
DC only where the high accuracy DC (2 or 3 inputs) hits only the 2 stages in
the 1013.  The other output where each AC/DC input hits first a OP275 (or
Paul's audio choice).  The final summer is a OP-275 with inputs from the
other 275s and the first 1/2 of the 1013.  All the polarity ends up right
and you can have up to 7 inputs (8, if you give up the panel space needed
for master on the output stage).  Remember one of the DC only inputs is
1/volt oct and no attenuation.

Larry H

P.S.  All you guys wanting the dual-function, auto-retracting, super
knob/controls keep to yourself. <snicker>

The easiest thing to do is to take Blacet's mixer and put it behind
a MOTM panel.  Larry Hendry did an amazing job with his.  Maybe
I'm the only oddball on the list that needs a mixer that can do
several AC inputs (50hz-5kz) as well as slow DC stuff like EGs.

With that said, Paul mentioned an opamp that was only good to
15Hz.  No matter how low of a DC offset it may have, it is totally
worthless in synth applications.  The slew rate on said opamp
would be horribly slow, and would mangle most normal control signals,
and most definitely would alter a saw waveform.  Analog Devices
makes some instrumentation amplifiers with wide bandwidth (some
out to 25Mhz) with extremely low DC offset and drift.  A couple
have <25microvolt offset with <.25microvolt per Centrigrade drift.
That would satisfy me.  The price?  $1780.

-Cary

"Surprised no one has brought up lin versus log pots! <<duck>>"

I used linear pots in the MX2040, but used a resistor trick to flatten
the curve a bit around the center detent position (off). This allows
finer control at lower voltage levels.

People (who are OUTSTANDING folks in every other way) who think mixer
design is easy and cheap just drive me (us?) crazy..... 

Regards.
-------------------------
John Blacet
Blacet Research Music Electronics
http://www.blacet.com
-------------------------
blacet@...
-------------------------
Are you on our mailing list?
http://www.blacet.com/mailform2.html

Op-amps for mixers...

OK I specify the TL072 in my multimix. Its OK for audio, and less than
perfect for CV. But still usable in most if not all situations. I would
be glad to hear of any problems that people have had regarding the use
of this unit. But you can use the OP275, great for audio and IMHO good
enough for pitch CV use. The other one to think about is the dual OP-27,
(OP272??)

But I would hasten to add why we really need to be 'hugely' accurate at
DC for an external mixer. This only really applies for VCO pitch. The
MOTM VCO has a 1V/oct input which are used for the pitch CV already.
Presumably we need to mix other CV sources to our VCO through the two of
the other potted inputs. The mixer performance that we need to control
is therefore offset drift. I don't think absolute offset is that
important unless you are feeding the mixer into a VCA then the VCO's
pitch control. In most cases this could be nulled out anyhow. So for
offset drift this will only be apparent if the op-amp does drift badly
with temperature, and the VCO's input pot is set to its maximum
senstivity. It is best then to use your mixer pots over their fullest
ranges and set the VCO input to the lowest setting to get the sound you
require.

Another issue for mixers, particulary at DC is the gain. This is worse
to set up accurately in a reversible attenuator system. In other words
accurate calibration of +1 and -1 settings.

Regards,

Tony Allgood  Penrith, Cumbria, UK

Modular synth circuits, TB303 clone and Filter Rack
www.techrepairs.freeserve.co.uk/projects.htm
My music: www.mp3.com/taklamakan

I second this approach.  I'd rather have a good audio mixer that does DC than a
hugely accurate DC mixer or a very expensive mixer that does both.  The voltage
summers on the inputs of the modules should do the accurate DC mixing. When I'm
mixing control voltages through a mixer, the offset and drift are usually not as
important because I'm, well, mixing up a bunch of control signals and probably don't
care as much about 1V/oct scaling, etc.  Besides, a little drift and non-linearity
can be very creative now and then...

Eric


--- Tony Allgood <oakley@...> wrote:
> Op-amps for mixers...
> 
> OK I specify the TL072 in my multimix. Its OK for audio, and less than
> perfect for CV. But still usable in most if not all situations. I would
> be glad to hear of any problems that people have had regarding the use
> of this unit. But you can use the OP275, great for audio and IMHO good
> enough for pitch CV use. The other one to think about is the dual OP-27,
> (OP272??)
> 
> But I would hasten to add why we really need to be 'hugely' accurate at
> DC for an external mixer. This only really applies for VCO pitch. The
> MOTM VCO has a 1V/oct input which are used for the pitch CV already.
> Presumably we need to mix other CV sources to our VCO through the two of
> the other potted inputs. The mixer performance that we need to control
> is therefore offset drift. I don't think absolute offset is that
> important unless you are feeding the mixer into a VCA then the VCO's
> pitch control. In most cases this could be nulled out anyhow. So for
> offset drift this will only be apparent if the op-amp does drift badly
> with temperature, and the VCO's input pot is set to its maximum
> senstivity. It is best then to use your mixer pots over their fullest
> ranges and set the VCO input to the lowest setting to get the sound you
> require.
> 
> Another issue for mixers, particulary at DC is the gain. This is worse
> to set up accurately in a reversible attenuator system. In other words
> accurate calibration of +1 and -1 settings.
> 
> Regards,
> 
> Tony Allgood  Penrith, Cumbria, UK
> 
> Modular synth circuits, TB303 clone and Filter Rack
> www.techrepairs.freeserve.co.uk/projects.htm
> My music: www.mp3.com/taklamakan
> 
> 
> 
> 


__________________________________________________
Do You Yahoo!?
Get Yahoo! Mail \ufffd Free email you can access from anywhere!
http://mail.yahoo.com/

> Op-amps for mixers...
> But I would hasten to add why we really need to be 'hugely' accurate at
> DC for an external mixer. This only really applies for VCO pitch. The
> MOTM VCO has a 1V/oct input which are used for the pitch CV already.
> Presumably we need to mix other CV sources to our VCO through the two of
> the other potted inputs. The mixer performance that we need to control
> is therefore offset drift. I don't think absolute offset is that
> important unless you are feeding the mixer into a VCA then the VCO's
> pitch control. In most cases this could be nulled out anyhow. So for
> offset drift this will only be apparent if the op-amp does drift badly
> with temperature, and the VCO's input pot is set to its maximum
> senstivity. It is best then to use your mixer pots over their fullest
> ranges and set the VCO input to the lowest setting to get the sound you
> require.

This is exactly where I'll probably compromise. The modules that really are
looking for accurate DC
*already* have a 3-in mixer. The question the begs itself: how many
instances do you ever need more than
3 simultameous control voltages into a VCO?

I am assuming that 99% of the time, a mixer module is for audio. But
certainly, you don't want to
introduce drift if you can avoid it just by better (read more $$$) op amp
selection.

I'll cruise data sheets this weekend and make my mind up.

Surprised no one has brought up lin versus log pots! <<duck>>

Paul S.

--- Tony Allgood <oakley@...> wrote:
> But I would hasten to add why we really need to be 'hugely'
> accurate at  DC for an external mixer. This only really applies
> for VCO pitch. The MOTM VCO has a 1V/oct input which
> are used for the pitch CV already.

----- Original Message -----
> From: alt-mode <alt_mode@...>
> I second this approach.  I'd rather have a good audio mixer that does
> DC than a hugely accurate DC mixer or a very expensive mixer that
> does both.  The voltage summers on the inputs of the modules should
> do the accurate DC mixing. When I'm mixing control voltages through
> a mixer, the offset and drift are usually not as important because I'm,
> well, mixing up a bunch of control signals and probably don't care as
> much about 1V/oct scaling, etc.  Besides, a little drift and non-linearity
> can be very creative now and then...
> Eric

I agree gentlemen. "Usefulness" and "do you need this or that?" should be
the basis for design anyhow.  But, I thought the discussion about how one
might go about it was interesting and worth having.  I think we will all be
happy with the compromise approach Paul will choose.

> Tony Allgood wrote:
> Another issue for mixers, particulary at DC is the gain. This is worse
> to set up accurately in a reversible attenuator system. In other words
> accurate calibration of +1 and -1 settings.

Yep.  So, you would have a mixture on inputs -- some or most reversing and
maybe one or two that are not reversing.  Kind of the same approach Paul
uses with FM inputs.

Larry H

Tony (and all),
         I built Tony's mixer with the plain old TL072s and was quite
pleased with  the results.  I was especially happy with the way reversing
was accomplished with only one op-amp.  These actually zero out better in
the middle than anything else I have.  As Tony said, it is not hugely
accurate for DC.  1% resistors are a must.   Personally, I would take my
stack of 47K resistors used for the construction and measure all their
values.  I would stick the highest value of the bunch in R72 and the lower
ones of the bunch in R69, 70 and 71.  This would serve to cancel out the
effect of the 1K resistors on the outputs of the inputs attenuators being in
series with the summing resistors and bring the summing stage closer to
unity gain.  Dave B's normalizing plans make this simple and small circuit
and very useful utility mixer/multiplier/reversing attenuator.
Larry H

----- Original Message -----
From: Tony Allgood <oakley@...>
OK I specify the TL072 in my multimix. Its OK for audio,
and less than perfect for CV. But still usable in most if not
all situations. I would be glad to hear of any problems that
people have had regarding the use of this unit.