[sdiy] moog module designs

[Tim Daugard]

Extra long discussion on design issues for a module follow.

FIRST:

Thanks Kevin Lightner for all the help and an excellent website. I know
that these discussions can get time consuming, so anytime you want to
stop replying feel free.

second: all the survey results so far have been in the forum (except
maybe one - I'll have to double check)

GENERAL ITEMS:

> I think most of the folks here have more experience with building
> their own designs than working with original Moogs on an ongoing
> basis. We're all specialists and this is one of my specialities. :)

I have built many of my own designs. I'm now going through serge and
moog designs looking for interesting circuits / functions I've missed.

> I'm not one to lend much to a thread that asks which atmel processor
> has the most ports, etc.  ;-)

Neither am I, my prom programer project is sitting in a storage
container right now. To continue it, I needed a fairly accurate way to
measure pulse widths. This entailed bringing out a freq counter I built
a long time ago and converting it to an A to B event timer and a pulse
width timer. As I worked on the trigger circuits for that, I discovered
that I needed a greater range of pulses out my square wave signal source
(it had 1 pulse, f/2 and f/4 outputs). This required I build an
extension box so that I could get f/8 and f/16 so that I could use the
f/16 for a reset pulse and have enough pulses from the f/2 to test the
rest of the circuit. I hate working digital, I makes me think too hard.

MOOG 923:

> >LM308s in the filters,  with different compensation capacitors. This
> >means I have to find a LM308 and see what the different caps do to
the
> >
> Are 308's that hard to find?

No, just a non-standard part in my stock. I have a good stock of LM324s
and TL074s. When these don't work because of cross talk, I switch to
1448s? (whatever the half 324 part is) and 072s. If I need a single
chip, I have a some of 5534s and 071s. I also have a large collection of
other sampled parts, just no LM308s. I prefer to use my stocked parts so
that I always have a spare if a part fails (though the only chip failure
I have ever had is in the test box listed above. In fact I should retro
fit that box with the moog output protection diodes to supplement the
current limiting resistors and the EMI coils on the outputs.)

The moog design is so old that part of the quality of the sound could
come from the limitations of the LM308 opamp.

> >buffer. The circuit is just begging me to build it. And, I know I
don't
> >have exact matches for the capacitor values in my bins. I stock 1,
2.2,
> >4.7 as my standard caps.
>
> The values in the pink noise filter are all available.
> (Should I sell you a kit? ;-)

Sure, I'll give you $1 for all the caps, matched to 1% of the specified
value including shipping :-)

All the  modules I've built (40+ different - I stopped counting after
40) cost less than $5 a peice. The only exceptions to this were the
power supply modules and some of the sequencer modules (8 pots at $0.41
eats up the budget fast.)

> The 32 mfd caps decoupling the outputs can be "assembled" from two
> separate caps if you want to be that precise. The 2.5 mfd as well. I

I was going to sort through my Rat Snack bulk caps to look for the
matching values. Those caps are so off labeled that I can usually find
any value small cap I want to within 1%, it's just a bunch of work. It's
easier to put the module on the shelf waiting for the next parts buy.
The problem with that is that there are too many containers of projects
on that shelf already.

Second option is 32 mfd = 22uf + 10uf right! right?

Actually the output decoupling caps was one of the areas that would
probabily be redesigned to meet the standards of the rest of my modules.
I know that is hearesy, but my system uses a single wire patch cable and
current signals instead of two wire voltage signals with shielded plugs.
I live in an area with so much radio interferance that AC coupling is
just asking for trouble. I have had to add 10K to 100K resistors and pf
bypass capacitors to ground on some equipment just to keep the inputs
from picking up noise.

> know there's some loading after them, so you may well want to be
> precise.

. . . and this is where one of the questions of the 308 vs the 071 comes
in. The input impediance of the 308 is 10 to 40 Megs, the 071 is much
greater.  I thought about hanging a 10M resistor on the 071 input to
reduce its impediance.

> >Another problem with the 923 is finding something to replace the SD10
> >noise diode. I'm probably going to go with a zener.
>
> That *is* a zener in the circuit.
> If I could recommend, buy several zeners and do listening tests.

That's good to know. My preliminary search for the SD10 hadn't come up
with a data sheet. All my noise modules (5+) have been built with zener
diode noise sources. I haven't built any with transistor noise sources
yet. Is transistor noise that much better?

> There can be a huge variance in their spectrums and this is probably
> more important than getting the caps perfect in the filter(s.)

Actually, it was the pink noise filter, the 6db variable low pass and
the 6db variable high passes that intrigued me.

The final question is still the 308. The 308 has caps for freq
compensation. If only some one could run a 1Khz Square wave, a 10Khz
Square wave and a Sine wave large enough to overdrive a 308 with a 3.3
pf, 30pf and without a compensation cap. It would be interesting to know
the results. There must be some reason for the caps with the 308,
otherwise Moog wouldn't have uses 3.3pf for the high pass and 30pf for
the lowpass.

If someone wants to donate a 308, I could do this. Otherwise it will
have to wait till my autumn parts order.

After reading the description of the upstate NY moog plant in Analog
Days, MOOG would not have spent money on those parts unless they were
necessary for the design to function the way he wanted it.

END OF THE MOOG 923 DISCUSSION

> >The SOS article on moog modulars states that the 902 has a punch
useful
> >for bass that hasn't been equalied by anything else. And that
statment

> The Moog does have an expo/lin switch and the expo setting can be
> used for some nice percussive curves, but the one thing I've learned
> after playing with synths so much is not to trust what people write
> about for the most part. There's a ton of false rumors out there and
> money changes hands everyday over expectations, disappointments,
> fear, etc.

Exactly, that was why the survey questions. In the past, I've bought
things only to be disappointed. My first commercial computer was the TI
99/4 for the same price I could have bought the RS color computer. I
trusted TI more than RS. The TI ultimitley proved to be a
disappointment. I got my first CoCo as a freebe tossed in when I bought
a printer off a guy. It was easier to learn the CoCo then build an
interface for the printer for the TI.

My absolute favorite thing (currently) of the internet world is the user
reviews on amazon.com. When many people say something is great, and I
was thinking of buying, I feel more confident in the purchase. When they
say that the sound transfer is poor or the book is only average, I can
wait or forgo the purchase. There is enough clutter in life that I don't
need to add to it by buying poor products. And there is enough room on
my bench (but not in my rack at the moment) that I can build one more
killer module. If several synthdiyers say a module is great, it's at
least worth looking at. If several people say a module is no good, then
that can go way down the list. Does anyone actually use the Serge TKB
touch pads to play?

Tim Daugard
AG4GZ 30.4078N 86.6227W Alt: 3.7 M
http://home.sprintmail.com/~daugard/synth.htm

. . . and the one thing everyone quoted out of the last message had a
misspelling not caught by the spell checker . . . sigh