All my ASM-1 questions at once!
magnus at analogue.org
Thu Dec 4 11:24:35 CET 1997
>>>>> "KDB" == Kurt Dwight Bleach <gghegel at earthlink.net> writes:
KDB> So, I'm plugging away at my ASM-1 and have, fortunately, found the
KDB> answers to almost all of my questions in the archives. However, there's
KDB> a couple of things I can't figure out, so I thought I'd compile them all
KDB> into one e-mail, so as not to flood the list with a bunch of stupid
KDB> newbie questions. Any input would be greatly appreciated!
KDB> 1) In regards to the NPN pairs in the exponential converter, do 2N3904s
KDB> work okay or is it much better to go with the MAT-02s? If the 2N3904s
KDB> are okay (I have a bunch lying around) how do they go into the board? I
KDB> ordered some LM394s from Digikey, but stupidly got the 8-dip package.
KDB> Whats the correct package? TO-5? (Anybody need a couple 8-dip LM394's?)
The MAT-02's is much better paired and is also well thermally
connected, something that is worth noting since thermal differance
between those transistors transforms into detuning...
Personally I went for the MAT-02's and I feel happy about that decission.
Orienting the MAT-02 in metalcap is a little tricky, but not as your
second question below...
KDB> 2) Im using 2N4391 JFETs. In trying to figure out their orientation on
KDB> the board, I went through the archive and found what looks to me to be
KDB> conflicting information. Can anybody help me out with this?
KDB> First I saw this: The 2N4391 to 2N4393 does not have the right leg
KDB> pinout, you will have to bend the middle leg to end up on the other side
KDB> or put the transistor on the soldering side.
KDB> But then I noticed: Primitive ASCII pin diagram for 2N4391/2N4392 JFET:
KDB> TO-18 Metal Can Package
KDB> / \
KDB> S | o | G
KDB> | o o |
KDB> tab ---> /\ /
KDB> \/ ---
KDB> **** Bottom view - pins pointed towards your eyeballs ****
KDB> S = source, D = drain, G = gate
KDB> In the ASM-1 VCO: Source = goes towards CA3140 (-) input, Drain = goes
KDB> towards CA3140 output, Gate = comes from LM311 output
KDB> Comparing this information with the schematic, I ended up putting mine
KDB> on the board with the tab pointing straight up (away from the CA3140, no
KDB> crossing pins). Is this correct, or did I screw up?
I found an error that Gene did in the schematic/board transition... I
guess many people know of this by now:
He wrote 2N4391 but he designed the board for the PN4391!
They are the same transistor really, but the capsule and pinning is
diffrent, so you end up with a very quite VCO or you found out somehow
and bended the Gate pin so that you can put it in the PN4391 position.
I had to do this to my 2N4391's...
Here is some horrible ASCII graphic of it (bottom view again):
G D D
o o o
o o o
S S G
So you have to bend the Gate (pin2) along the diagonal in the square
for which the pins touches the corners...
Oh, you migth not get the VCO running cause a little other trouble...
If the output locks to +15V then you should increase the feedback
capacitor over the LM311. This capacitor controls the reset-time and
when it is too small will it not empty the capacitor long enougth and
therefore it locks to +15V.
KDB> 3) Regarding power supplies, I found a Harrison 802B twin low voltage
KDB> power supply at a prop house for $3.00. (They thought it was a power
KDB> amp.) It has knobs on the front to adjust the voltage. I tested it out
KDB> and it seems quite stable, outputting +15V, -15V and ground. Anybody
KDB> know if this will work or are there some important criteria of which Im
KDB> embarrassingly ignorant? (E.g. noise problems, hum, grounding problems,
KDB> 4) This is more of an aesthetic thing, but I want to add a couple LED's
KDB> to the Envelope generators. Can I just attach them to the ADSR out in
KDB> the same way that the LED is incorporated in the LFO? I'd like them to
KDB> flash when the envelope is triggered. Anybody tried this? (I know, it's
KDB> silly, but I have a couple bright blue LED's collecting dust...)
Put them to work then... it should be OK, but if you run into trouble
use a unity gain buffer with a spare op-amp to reduce the load at the
KDB> Well, any help would be grand. Thanks!!!
PS. In case people forgot or didn't know it there is a ASM-1 homepage at
I admitt that it is not refreshed for a long time... motivate me
to change that ;)
More information about the Synth-diy