dustbins, differential amps, cables

[media at mail1.nai.net]

>Each ladder filter, and most OTAs, have a positive and a negative input.
>And you also have to divide down your typical signal levels, so these
>inputs are almost perfect summing nodes. So instead of grounding one
>of them, you could build a second attenuator and connect it to the
>"ring" of your input jack.

OK :)  That makes perfect sense.  I am new to the world of diode ladders
and OTA's.  My personal response to the CA3080 data sheet was "wtf???"

>Does this frighten you?

Nothing frightens me since I've seen the insides of a DAT machine :)  I was
a bit concerned before I understood this "dustbin" concept.  As long as he
isn't grounding the chassis with a chip I'm fine with it.  I take it to
mean he is using an extra "ground" created by an op-amp to serve as a sort
of "electronic sewer" for circuits that tend to spam the ground.

>(Almost) each jack on my Modular has its
>own GND connection to the module's local star GND. And I don't
>use isolated jacks, so they are connected to each other across the
>aluminium front as well - and still the individual GND wires make a
><difference ! (Don't ask me why.) Ok, it's overkill.

No, it isn't, really.  Aluminum chassis, for some reason unknown to me,
often make noisey grounds.  How about using isolated jacks??   This might
solve some of your problems, or at least minimize the ways it can fail.
The simplest system that works is best.

>I'd use only three GNDs, if I had to do it again, as well.

So what are your three grounds again??

>The "dustbin" idea is that you save one of these GNDs. Say you have
>a LED in the middle of a PCB behinf the front panel, you can either
>use a separate GND to the power supply, or use an onboard opamp.

OK :)  I think I understand this dustbin idea.  This sounds like it might
be useful for the metering on a mic pre or some sort of hybrid device.  As
far as a modular is concerned would this be something I would build into
the PS or would I build one off the rails in each module as needed??  Is
there a limit on the number of "dustbins" I could run off of one PS??
Would I have a quieter ground from the get go if I sent the filter caps
following the regulators to the "dustbin" rather than the ground??

>But as I don't want an extra mixer stage
>for the (planned) 4 modules, I have implemented symmetrical outputs
>that you can simply short together. I used the Minimoog VCA
>circuit. The last stage is a linearized (degenerate emitter) PNP
>differential amplifier with rather high current (10mA if memory serves).
>So you can directly drive a balanced cable from the two collectors
>of the diff amp (with an elko and a small protection resistor in each
>connection, of course). It doesn't have a higher impedance than
>your usual opamp stage with 1k series resistor, it's full Class A,
>and you can mix several of these by simply shorting them together.)

Wow!! I'm going to give that more thought.  Being able to combine outputs
without a summing amp would be powerful.  I'm not sure what you mean by
elko (here an elco is a 56 pin connector used for consoles and multitracks).

So what happens to this VCA when one of the collectors goes to ground??
This is what will happen if you stick a TS plug into a TRS jack.

>>True, but how much noise gets into a six foot cable in the
>>first place?? Wouldn't raising potentials be easier than
>>balancing the system??   Given high enough voltages, any RF
>>interference would be inaudible or irrelevant.
>
>One word: Headroom.
>This isn't much of an issue for internal synth signals, but becomes
>a problem as soon as you want to process external signals.

Now, I'm confused.  If it were designed to process hotter internal signals
without clipping, wouldn't it also be able to handle hotter external
signals??  I can forsee the opposite problem that external signals would
not be hot enough, and the necessary gain would also amplify the input
noise.

>BTW, even for internal signals, I would never design a synth for
>+/- 10V swing with +/-15V supply voltages.
>I designed JH-3 for +/-5V swing which gives me 6dB headroom.
>I would rather choose an even lower level and greater headroom next
>time.

Isn't 10Vpp and +/-15V PS typical for synth??  Why don't they run closer to
the rails??  Wouldn't using hotter signals greatly reduce these noise
problems??

>>>Another thought: Maybe (I' only guessing here) it would even
>>>help to use unbalanced output signals, but balanced input stages.
>>>The outputs would drivethe tip with the signal, and connect
>>>ring and shield with their *local* GND. The inputs jacks would
>>>have the shield unconnected, and a differential
>>>input stage would amplify the voltage between the tip (=signal)
>>>and the ring (=local GND of output module). This way you'd
>>>loose the advantage of radiation immunity (but we have short
>>>cables anyway), but you'd still avoid GND loops (the ususal
>>>unbalanced problem) and crosstalk (the potential banana problem).

Isn't that just as bad or worse than going unbalanced??  For all practical
purposes, you are connecting the shield to the cold (out of phase) input.
Not that ground loops are exclusive to unbalanced connections, but I don't
see how you are necessarily avoiding a ground loop or how this is
symmetrical.

In this particular example, if you connected the shield to ground and the
differential inputs for in phase and out of phase signal to each other
conductor at the input, and disconnected the shield at the output (to avoid
ground loops)  that might be better than just going unbalanced.   At least
the screen wouldn't be draining into one of the inputs.

However, wouldn't it be better to balance the outputs and leave the shield
unconnected at the input??

>It's not RF what I want to fight in the first place, but hum and
>crosstalk.

Will the CMRR remove the hum, and is crosstalk less than RF interference??
I do not think I would want to lose the radiation immunity of a balanced
system -- I think that is the primary function of a balanced system.

>You may be right. I just want to point out that this (non standard
>levels) is exactly the same constraint that forbids raising your
>standard level (see above).
>BTW, a 3080 and a handful of resistors may be much compared to
>a simple pot, but remember that the OTA does the balanced input job
>as well as the attenuation. And it could be the first step to give your
>module some preset or even storage functions (;->) (;->)

I don't know what the trade-off would be between adding noise from another
IC and applying common mode rejection to six foot cable.

As far as trying to build this, I think you are correct that adding
balanced outs is "the problem."  You have explained how many synth circuits
have natural differential inputs.  Beyond that adding a differential input
with an op-amp can't be that hard -- you seem to get yours "for free" by
including it with a new feature :)

Also, you should seriously consider having all of your balanced in and outs
built in such a way that they will accept unbalanced connections.  Firstly,
your existing modules are most likely unbalanced.  Secondly, most synths
and effects pedals are unbalanced.  Thirdly, it is always easier to find
1/4" TS cables than 1/4" TRS cables.  Fourthly, you can't keep other people
from using the wrong cables on your equipment when your not looking!!

I think the big question is whether adding balanced connections is going to
reduce hum or be a potential source of even more hum.  Perhaps balancing
some connections but not others is a good idea.  Choosing the outputs of
VCA's and VCF's being sent to other equipment would seem sensible.

Even though such balancing a modular is extreme, I don't think we're
tilting windmills :)

PEACE OUT :)
MARK


--
"I know that I can come off as a little proactive, and for that I'm sorry."
                                                  --  Homer Simpson