[sdiy] asm-1/asm-2

[Magnus Danielson]

From: Gene Stopp <gene at ixiacom.com>
Subject: RE: [sdiy] asm-1/asm-2
Date: Wed, 2 Jun 2004 16:56:30 -0700 
Message-ID: <15FDCE057B48784C80836803AE3598D503A55ACF at racerx.ixiacom.com>

Hi Gene,

> Yes good idea (hi Magnus!)

Yeah, hi! ;O)

> What happened is that the LM358 input comparator is powered between ground
> and V+ and negative voltages drive it batty. OK in pre-patched systems
> (where the gate can be 0 --> V+) but bad in modular systems.

Indeed. In a mixed power-supply device like the ASM-1 effectively is (the
ADSR is GND to +15V powered) one should be aware, but general input protection
is one of the things which we tend not to take very serious. Not very
supprising we see things deepfried a bit into the design when things go
terribly wrong. A resistor and a pair of well-placed diodes can do minor
wonders to save much more expensive components and hard work.

So, in order to follow that general idea, add another diode to the +15V rail,
so that over-voltage (above the +15V rail) gets directed to that rail. The
resistor stops major current flows to that rail. If you end up with 230Vrms
into that resistor it will for a 1k resistor create a current of 230mArms.
This creates a nice little heating of 52,9 W so it's fusing time! The power-
supply only needs to keep the voltage stable enought for it not to cause
damage. Some elaboration with different resistors from my finger-in-the-air-
value of 1k should be recommended. A little higher like 4k7 just heats the
resistor with 11,26 W, so it takes longer to burn, but on the other hand has
the current load it dumps on the PSU-line gone down to 48,94 mA which is more
manageable.

You don't want to make the misstake of using the diode as the fuse, it's there
to divert the current until the chock goes away or the resistor fuses. The
resistor is there to make sure the curren't isn't too high and that also to
drop the voltage to a manageable level.

WHEN a resistor fuses, check or even replace the diodes anyway. You don't want
a defect diode sitting there and failing on you the next time. They _do_ take
alot of stress if the resistor isn't current't limiting enought.

Hmm... where's that book? I got a book ful of good thinking and advices on
this.

Cheers,
Magnus - who hasn't badly burnt anything for a little too long time, heated yes, but not burnt...