[sdiy] Modular patching rules

[Florian E. Teply]

Tim Parkhurst wrote:
> On 2/11/07, Dave Kendall <davekendall at ntlworld.com> wrote:
>> Hi.
>>
>> Getting slightly indecisive here with the modular getting slowly
>> bigger. Some advice from the experts would be very welcome.
>>
>> Regarding gate and trigger levels, should I aim for +5V gate/trig or
>> +10V as a standard -  What do you guys do?
>>
Well, that's about the same thing I'm currently thinking about. Have yet
to make up my mind on it though.

>> A lot of the boards I've got waiting around (assembled or otherwise)
>> are EFM, which look like they *mostly* have 100k input impedance, but
>> are missing 1K resistors on the output (easy enough to fix). These seem
>> to often need +10V for gates,and up to 10V or so for control.
>> Is 0 to+10V common for a module CV input range, or is it  worth
>> tweaking the CV control ranges to make them 0 - +5V? That would make
>> sense for a VCF being swept by a +5V EG for example, but if swept by a
>> ±5V LFO, the LFO would need attenuating to cover the whole range
>> without dead zones at the top and bottom - maybe that's just standard
>> practice?
>>
> Regarding CV ranges: the inputs on modular gear usually have an
> attenuator and some sort of "offset" or "initial tune" setting. The
> Offset or Tune control usually applies a DC voltage to the CV summer.
> What this means is that between the attenuator and offset controls,
> you should be able to get a nice, full sweep from a 0 to 5V, 0 to 10V
> or a +/-5V mod signal. The attenuator sets the mod depth, and the
> Offset control sets the mod start point. 

Umm, maybe I'm a bit off track there, but wouldn't an attenuator break
the 1V/octave rule? Okay, that may or may not be a real rule, but in my
opinion it would be quite handy to retain that value wherever possible.
Would be cool though to hear from someone more experienced than me...

> Basically, properly designed
> gear should accept all of those signals without any problems. In the
> case of gear with a microprocessor powered on only 5V, then input
> protection should be added so that a signal greater than 5V doesn't
> fry the micro.
> 
> Trigger levels can be all over the place. A lot of stuff with logic
> circuits will output a 5V gate. OTOH, my Odyssey puts out a +13V gate!
> Again, a properly designed piece of gear should accept something from
> 5V or above without damage. Worst case, it's not impossible for a
> piece of gear running on +/-15V to put out a signal
> as hot as 15V. Design for that "worst case" and you'll be okay most of
> the time ;-) Until someone wires 36V phantom power into a gate input!
> 
Well, clamping diodes to the power rails somewhere along the input
resistors could be a possible solution for that scenario. Just make
sure, a voltage in any meanigful range won't kill the diodes ;-) I'm
tlaking about something like this:

Input Jack  o
            |
            1k
            |
  +15V o-|<-+-|<-o  -15V
            |
           99k
            |
          OpAmp

So in case the input voltage goes beyond +/- 15V, the diodes basically
clamps the voltage on the middle node to the power rail. Even a cheapo
1N4148 should survive this treatment for input voltages of up to 200+
Volts without turning into a puff of smoke. Still they won't affect
input resistance too much as long as the input voltage stays within the
power rails as they are reverse biased. For microprocessor or TTL stuff
just trade the +/- 15V for +5V and Ground...

Just my 2 cents
Florian

-- 
Langeweile? Lust auf ein fesselndes Spiel?
-> Das MorgenGrauen erwartet Dich!!
http://mg.mud.de/newweb oder
http://telnet.morgengrauen.info