MOTM

I am in the process of converting my Blacet Time Machine to MOTM format and 
pose this question. Since all of the grounds on the Blacet schematic 
individually "star" back to one point on the pcb, can't I simply connect all 
the jacks with unshielded wire in a star type web and run one connection off 
this web to the pcb ground for the same overall effect? I used the 
unshielded wire method and it appears to work fine. Although in looking on 
Larry Hendry's page I noticed that his model uses Blacets wiring scheme 
while Dave Hylander uses the "Oakley Type" unshielded wire approach as I 
did. Comments?

Thomas White

PS. Nice to not have to fiddle with 1/8 cables anymore :)



_________________________________________________________________
MSN Photos is the easiest way to share and print your photos: 
http://photos.msn.com/support/worldwide.aspx

OK, I'll offer up an opinion and disclaimer.

I did not exactly follow John's recommendations.  But, I did use up all the
ground pads available on the PCB.  You will notice I combined some grounds
and not others.  I did not mix CVs and audio or anything else that logically
made sense to me might interfere with one another (which seemed to be John's
basic approach too from my casual observation).

Here was my small brain's thought process. In an unbalanced scheme, there is
as much signal current in the shield as the signal conductor.  So, if you
have only one wire from jacks to PCB, then all these currents mix on the way
to the PCB.  And, no matter how hard you try to keep them separate, you
cannot completely because they jacks are all quite well bonded at the panel.
So, some interaction between CVs and audio caused by the voltage drop in the
ground connection will occur.  Ideally, there would be no common path
(isolated grounds, isolated path on the PCB).  But, that of course is not
the case.

My thought was that I wanted as little interaction between I/Os as possible.
So, that meant to keep the R of the ground connection from jacks to the PCB
very low.  Therefore, I went with the multiple wire method.  I assume the
single wire method might work just as well, provided the wire was not
significantly undersized. Since there is a limit to the wire size that can
connect to the TM PCB, the multiple wire path made sense to me.

Disclaimer: I don't posses the educational accreditation to support my
theory.  This is not meant to be an instruction on how it should be done,
but rather an insight on what I was thinking when I largely just followed
the designer's recommendation and made my own decisions.  I have never built
any of my modules with the "strap 'em all together, one ground wire" method.
In fact looking at my Multimix, I see I used shielded I/0 and used heat
shrink near the PCB where I tapped the PCB end of the shield to run a ground
wire over to the SCRN connection (unless using the adjacent ground
connection).  I am also not offering any criticism to the other methods.

Larry (now you know how my crazy brain works somewhat) Hendry

P.S. Since the TM is basically a "lo-fi" device AND according to our quiet
list friend Harry B, "BDDs suck" does it really matter?

----- Original Message -----
From: thomas white <djthomaswhite@...>
To: <motm@yahoogroups.com>
Sent: Tuesday, February 26, 2002 8:33 AM
Subject: [motm] Conversions Question


I am in the process of converting my Blacet Time Machine to MOTM format and
pose this question. Since all of the grounds on the Blacet schematic
individually "star" back to one point on the pcb, can't I simply connect all
the jacks with unshielded wire in a star type web and run one connection off
this web to the pcb ground for the same overall effect? I used the
unshielded wire method and it appears to work fine. Although in looking on
Larry Hendry's page I noticed that his model uses Blacets wiring scheme
while Dave Hylander uses the "Oakley Type" unshielded wire approach as I
did. Comments?

Thomas White

PS. Nice to not have to fiddle with 1/8 cables anymore :)



_________________________________________________________________
MSN Photos is the easiest way to share and print your photos:
http://photos.msn.com/support/worldwide.aspx





Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/

Actually mine is a partial star.  Two points anyway.  The top and bottom 
row are on separated ground wires.  The second TM I built I did as Larry 
and ran all the grounds out, keeping the CV and audio separate.  But I'm 
not sure which is better.  I can not hear, see, feel, or taste 
anything  different between the two.  But then I  don't know if I'm a 
qualified connoisseur of fine  circuitry :)

dave

At 08:44 AM 2/26/02 -0600, J. Larry Hendry wrote:

>OK, I'll offer up an opinion and disclaimer.
>
>I did not exactly follow John's recommendations.  But, I did use up all the
>ground pads available on the PCB.  You will notice I combined some grounds
>and not others.  I did not mix CVs and audio or anything else that logically
>made sense to me might interfere with one another (which seemed to be John's
>basic approach too from my casual observation).
>
>Here was my small brain's thought process. In an unbalanced scheme, there is
>as much signal current in the shield as the signal conductor.  So, if you
>have only one wire from jacks to PCB, then all these currents mix on the way
>to the PCB.  And, no matter how hard you try to keep them separate, you
>cannot completely because they jacks are all quite well bonded at the panel.
>So, some interaction between CVs and audio caused by the voltage drop in the
>ground connection will occur.  Ideally, there would be no common path
>(isolated grounds, isolated path on the PCB).  But, that of course is not
>the case.
>
>My thought was that I wanted as little interaction between I/Os as possible.
>So, that meant to keep the R of the ground connection from jacks to the PCB
>very low.  Therefore, I went with the multiple wire method.  I assume the
>single wire method might work just as well, provided the wire was not
>significantly undersized. Since there is a limit to the wire size that can
>connect to the TM PCB, the multiple wire path made sense to me.
>
>Disclaimer: I don't posses the educational accreditation to support my
>theory.  This is not meant to be an instruction on how it should be done,
>but rather an insight on what I was thinking when I largely just followed
>the designer's recommendation and made my own decisions.  I have never built
>any of my modules with the "strap 'em all together, one ground wire" method.
>In fact looking at my Multimix, I see I used shielded I/0 and used heat
>shrink near the PCB where I tapped the PCB end of the shield to run a ground
>wire over to the SCRN connection (unless using the adjacent ground
>connection).  I am also not offering any criticism to the other methods.
>
>Larry (now you know how my crazy brain works somewhat) Hendry
>
>P.S. Since the TM is basically a "lo-fi" device AND according to our quiet
>list friend Harry B, "BDDs suck" does it really matter?
>
>----- Original Message -----
>From: thomas white <djthomaswhite@...>
>To: <motm@yahoogroups.com>
>Sent: Tuesday, February 26, 2002 8:33 AM
>Subject: [motm] Conversions Question
>
>
>I am in the process of converting my Blacet Time Machine to MOTM format and
>pose this question. Since all of the grounds on the Blacet schematic
>individually "star" back to one point on the pcb, can't I simply connect all
>the jacks with unshielded wire in a star type web and run one connection off
>this web to the pcb ground for the same overall effect? I used the
>unshielded wire method and it appears to work fine. Although in looking on
>Larry Hendry's page I noticed that his model uses Blacets wiring scheme
>while Dave Hylander uses the "Oakley Type" unshielded wire approach as I
>did. Comments?
>
>Thomas White
>
>PS. Nice to not have to fiddle with 1/8 cables anymore :)
>
>
>
>_________________________________________________________________
>MSN Photos is the easiest way to share and print your photos:
>http://photos.msn.com/support/worldwide.aspx
>
>
>
>
>
>Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/
>
>
>
>
>
>
>
>
>Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/

Thomas White writes:
I need to apply the Hendry and Bradley mods anyhow.

Larry Hendry:
Speaking of Hendry mods:  Sometimes you cannot see the forrest for the
trees.  I started my Time Machine mods with the thought of protecting the
two inputs from negative voltage separately.  My first shot at the delay CV
mod didn't work so well but the one to protect the external mod CV input
did.  So, I ended up changing the circuit for the delay CV mod to the one
that is published on my website. But, what didn't hit me until Scott Juskiw
was kind enough to point it out is that my delay CV negative input voltage
protection modifiction actually protects both inputs.  Doh !! (slapping my
head like Homer Simpson),

So, if you have already modified your Time Machine, you now are double
protected on the mod CV input.  If you have not, sounds like you can skip
the second part that calls for the changing of R57 and R58 with the
piggyback diode.  I will change the website information to reflect this when
I get a chance.

Thanks to Scott for helping me see the light.

Larry Hendry

How do you go about creating a Super Saw(tooth) wave
as found on the Roland JP-8000/JP-8080 on the MOTM (or
any analog modular for that matter)?  This appears to
be 7 detuned sawtooth waves.  How are these 7 related?
 Do I need to use 7 VCOs to achieve this?  Is Roland
using a static wave to do this or is the relationship
between the indiviual sawtooth waves varying over
time?

thanx
jeph

__________________________________________________
Do You Yahoo!?
Yahoo! Greetings - Send FREE e-cards for every occasion!
http://greetings.yahoo.com

At 1:33 PM -0800 2002/02/26, VCOVCAVCF wrote:
>How do you go about creating a Super Saw(tooth) wave
>as found on the Roland JP-8000/JP-8080 on the MOTM (or
>any analog modular for that matter)?  This appears to
>be 7 detuned sawtooth waves.  How are these 7 related?
>  Do I need to use 7 VCOs to achieve this?  Is Roland
>using a static wave to do this or is the relationship
>between the indiviual sawtooth waves varying over
>time?

I can think of two ways to do this with analogue gear:

1. use 7 VCOs
2. use fewer than 7 VCOs, sum them, and put them through a delay line

Digital gear cheats by generating a waveform that sounds like 
numerous saws detuned, and can do this with very little computational 
overheard. A saw wave is just a counter that goes up and then wraps 
around to zero. Detuned saws are similated by partially resetting the 
counter before the counter wraps. By changing the reset point on each 
cycle, it sounds like detuning. If the wave is partially reset N 
times each cycle, it sounds like N+1 detuned saws.

There must be some other way of getting something close to this with 
analogue gear. If an oscillator had a hard sync input that didn't 
completely drain the cap then the detuned saw wave could be created 
with an appropriate hard sync signal. But generating the hard sync 
signal would require 6 detuned pulse waves, so you're back to 7 
oscillators again.

> There must be some other way of getting something close to this with
> analogue gear. If an oscillator had a hard sync input that didn't
> completely drain the cap then the detuned saw wave could be created
> with an appropriate hard sync signal. But generating the hard sync
> signal would require 6 detuned pulse waves, so you're back to 7
> oscillators again.

i don't know how to get an osc to partially drain the cap, but as for the
other part of the problem, you could put a 310 pulse through a 120 which
would give you 5 detuned pulses. or you could get *ten* by putting a
second 310 pulse through the B input as well (i think).

bleep.
out.

---
www.mp3.com/leichenfeld
www.mp3.com/captainhotrod
www.mp3.com/silenttristero

At 4:15 PM -0600 2002/02/26, blip wrote:
>  > There must be some other way of getting something close to this with
>>  analogue gear. If an oscillator had a hard sync input that didn't
>>  completely drain the cap then the detuned saw wave could be created
>>  with an appropriate hard sync signal. But generating the hard sync
>>  signal would require 6 detuned pulse waves, so you're back to 7
>>  oscillators again.
>
>i don't know how to get an osc to partially drain the cap, but as for the
>other part of the problem, you could put a 310 pulse through a 120 which
>would give you 5 detuned pulses. or you could get *ten* by putting a
>second 310 pulse through the B input as well (i think).
>

But I believe the pulses coming out of the 120 would not be detuned 
with each other; they'd be locked to the 310 driving the 120. So 
that's not quite the same thing as having them all drifting freely.

On the other hand, that's a good place to start looking for a super 
saw substitute. I've made patches using the 120 that have all manner 
of weird beating going on that I can't explain.

You could just take a '300 oscillator and send its pulse output to the '120,
and take the output from the '120 and plug it back into one of the '300's FM
inputs.  The '120 changes state once per cycle, so each cycle of the 300
ends up with a different period, repeating over 16 cycles.  Tweaking the
various sub levels and the overall FM level gives you a variety of effects.
Changing the FM amount changes the "average" tuning because (I think) the
'120 doesn't average to zero, using AC coupled FM1 reduces that effect.

I don't know if its the effect you're looking for, but it sounded
interesting to me.

                    ...Craig

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

From: "blip" <bleep@...>
To: "MOTM list" <motm@yahoogroups.com>
Sent: Tuesday, February 26, 2002 2:15 PM
Subject: Re: [motm] Creating Super Saws???


>
> > There must be some other way of getting something close to this with
> > analogue gear. If an oscillator had a hard sync input that didn't
> > completely drain the cap then the detuned saw wave could be created
> > with an appropriate hard sync signal. But generating the hard sync
> > signal would require 6 detuned pulse waves, so you're back to 7
> > oscillators again.
>
> i don't know how to get an osc to partially drain the cap, but as for the
> other part of the problem, you could put a 310 pulse through a 120 which
> would give you 5 detuned pulses. or you could get *ten* by putting a
> second 310 pulse through the B input as well (i think).
>
> bleep.
> out.
>
> ---
> www.mp3.com/leichenfeld
> www.mp3.com/captainhotrod
> www.mp3.com/silenttristero
>
>
>
>
>
> Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/
>
>

>Digisound had one called the "Waveform Multiplier."

I have been meaning to do this one for quite a while after Chris
Crosskey (Atari madman and generally nice bloke) told me about it. It
should be easy enough to do, although the Digisound's PCBs set up was
horribly complex with all the little daughter boards. I tried the basic
principle and it certainly does work. If I get around to it, expect it
sometime in the Summer.

Regards,

Tony Allgood                   Penrith, Cumbria, England

Oakley Sound Systems                 www.oakleysound.com

If you want Super Saws, why not pick up a JP-8080 and use it as an oscillator 
bank among a million other things?  It's 10 voice polyphonic, so 10x7 gives 
you the equivilent of 70 sawtooth oscillators.

-Elhardt