Doepfer

I decided to try a number of signals patched to my A150's I/O2 
(Normally Open) jack. I wanted to see which ones would trigger an ADSR 
envelope by bleeding through the O/I jack, even though nothing was 
patched to the A150's CV jack to close the switch.

Signals that DID NOT bleed through were:

A145 LFO Pulse
A147 LFO Pulse
A155 Trigger
A156 Trigger
A190 Clock
A191 LFO Pulse

Signals that DID bleed through were:

A155 Gate
A160 Clock Divider
A162 Trigger Delay
A165 Trigger Modifier
A190 Gate

What's even more worrysome is that I'm fairly sure that I've 
successfully switched an A190 Gate signal with the A150 in the past, 
so now I'm concerned that the A150 is actually deteriorating in this 
regard. I wish I could be more certain about this.

There was an earlier response from Dieter on this issue, in which he 
ascribed the problem to user abuse of the A150. I'm certain that he 
misunderstood the symptoms, though, because he was talking about 
signals patched to the CV input, rather than the switch inputs. In any 
case, in was in the middle of preparations for the Frankfurt show and 
he probably didn't give the issue a lot of attention.

I've sent a copy of this post directly to Doepfer by email.

Joe

Damn shit I won the bidding at vemia for a Analog Systems 6u standard
system... just thought I'd jack up the price 680 euro was too low I
thought... so I got it for 690 ;)

Anybody have Asys here?

Hi Joe,
	I looked back over my old notes on this one: I still don't fully
understand it, but can offer the following. When you get the bleed
through, do you leave the other input unconnected? If so, try
grounding it, as this seems to improve matters considerably (I've been
partially inserting an unconnected patch cord to short the input to
ground - not the most reliable of connections - but connecting a CV
input set to 0V should do it too. Make sure you get your in's and
out's round the right way - shorting the gate or trigger signals to
ground through the switch would probably not be very nice!). You may
still get some bleed through, but from my experiments it is
considerably less, i.e. instead of the full signal, its down around
100mV or so. Also swapping inputs (I/O1 <-> I/O2) makes a difference
too - its worse one way than the other. The problem is exceeding the
+/-8V limit specified in the manual: despite some diodes apparently to
protect the chip, some excess voltage does get applied to it, and so
it's normal operating range is exceeded, and it thus stops working
properly. The 'DID NOT' signals you listed will almost certainly be
within +/-5V (I don't have a 155 or 191 so can't actually check these
two, the rest certainly are); of the 'DID's the 160 and 190 signals
are 0 - 12V, and I also strongly suspect the others are too (again I
don't have these). If anyone else has knowledge of/can point me to a
suitable reference for why the FET circuitry in the 4053 chip behaves
in this manner, I'd love to know it!

Tim

[The views expressed above are entirely those of the writer and do not
represent the views, policy or understanding of any other person or
official body.]



--- In Doepfer_a100@y..., "buechlerjoe" <buechlerjoe@t...> wrote:

> I decided to try a number of signals patched to my A150's I/O2 
> (Normally Open) jack. I wanted to see which ones would trigger an ADSR 
> envelope by bleeding through the O/I jack, even though nothing was 
> patched to the A150's CV jack to close the switch.
> 
> Signals that DID NOT bleed through were:
> 
> A145 LFO Pulse
> A147 LFO Pulse
> A155 Trigger
> A156 Trigger
> A190 Clock
> A191 LFO Pulse
> 
> Signals that DID bleed through were:
> 
> A155 Gate
> A160 Clock Divider
> A162 Trigger Delay
> A165 Trigger Modifier
> A190 Gate
> 
> What's even more worrysome is that I'm fairly sure that I've 
> successfully switched an A190 Gate signal with the A150 in the past, 
> so now I'm concerned that the A150 is actually deteriorating in this 
> regard. I wish I could be more certain about this.
> 
> There was an earlier response from Dieter on this issue, in which he 
> ascribed the problem to user abuse of the A150. I'm certain that he 
> misunderstood the symptoms, though, because he was talking about 
> signals patched to the CV input, rather than the switch inputs. In any 
> case, in was in the middle of preparations for the Frankfurt show and 
> he probably didn't give the issue a lot of attention.
> 
> I've sent a copy of this post directly to Doepfer by email.
> 
> Joe

[Apologies if this appears twice - seem to be having 'finger trouble'
with Yahoo.]

Hi Joe,
	I looked back over my old notes on this one: I still don't fully
understand it, but can offer the following. When you get the bleed
through, do you leave the other input unconnected? If so, try
grounding it, as this seems to improve matters considerably (I've been
partially inserting an unconnected patch cord to short the input to
ground - not the most reliable of connections - but connecting a CV
input set to 0V should do it too. Make sure you get your in's and
out's round the right way - shorting the gate or trigger signals to
ground through the switch would probably not be very nice!). You may
still get some bleed through, but from my experiments it is
considerably less, i.e. instead of the full signal, its down around
100mV or so. Also swapping inputs (I/O1 <-> I/O2) makes a difference
too - its worse one way than the other. The problem is exceeding the
+/-8V limit specified in the manual: despite some diodes apparently to
protect the chip, some excess voltage does get applied to it, and so
it's normal operating range is exceeded, and it thus stops working
properly. The 'DID NOT' signals you listed will almost certainly be
within +/-5V (I don't have a 155 or 191 so can't actually check these
two, the rest certainly are); of the 'DID's the 160 and 190 signals
are 0 - 12V, and I also strongly suspect the others are too (again I
don't have these). If anyone else has knowledge of/can point me to a
suitable reference for why the FET circuitry in the 4053 chip behaves
in this manner, I'd love to know it!

Tim

[The views expressed above are entirely those of the writer and do not
represent the views, policy or understanding of any other person or
official body.]


--- In Doepfer_a100@y..., "buechlerjoe" <buechlerjoe@t...> wrote:

> I decided to try a number of signals patched to my A150's I/O2 
> (Normally Open) jack. I wanted to see which ones would trigger an ADSR 
> envelope by bleeding through the O/I jack, even though nothing was 
> patched to the A150's CV jack to close the switch.
> 
> Signals that DID NOT bleed through were:
> 
> A145 LFO Pulse
> A147 LFO Pulse
> A155 Trigger
> A156 Trigger
> A190 Clock
> A191 LFO Pulse
> 
> Signals that DID bleed through were:
> 
> A155 Gate
> A160 Clock Divider
> A162 Trigger Delay
> A165 Trigger Modifier
> A190 Gate
> 
> What's even more worrysome is that I'm fairly sure that I've 
> successfully switched an A190 Gate signal with the A150 in the past, 
> so now I'm concerned that the A150 is actually deteriorating in this 
> regard. I wish I could be more certain about this.
> 
> There was an earlier response from Dieter on this issue, in which he 
> ascribed the problem to user abuse of the A150. I'm certain that he 
> misunderstood the symptoms, though, because he was talking about 
> signals patched to the CV input, rather than the switch inputs. In any 
> case, in was in the middle of preparations for the Frankfurt show and 
> he probably didn't give the issue a lot of attention.
> 
> I've sent a copy of this post directly to Doepfer by email.
> 
> Joe

According to the response I got from Dieter, the A150 User's Manual 
has been UPDATED to describe the -8V - +8V restriction. I guess if a 
product doesn't meet it's prospectus, update the prospectus ;)

Also, the FAQ on the Doepfer home page now describes this problem, and 
suggests either using attenuator patch cords with a resistor in the 
plug, or a user-applied modification to the A150 which would make it 
operate over the 0-16V range instead of -8V - +8V. This would make the 
A150 unusable for AC - audio or LFO signals.

I'm wondering if it's possible to mod only half of the dual switch.

Joe

--- In Doepfer_a100@y..., "stinchcombe_t" <tstinchcombe@q...> wrote:
> [Apologies if this appears twice - seem to be having 'finger 
trouble'
> with Yahoo.]
> 
> Hi Joe,
> 	I looked back over my old notes on this one: I still don't 
fully
> understand it, but can offer the following. When you get the bleed
> through, do you leave the other input unconnected? If so, try
> grounding it, as this seems to improve matters considerably (I've 
been
> partially inserting an unconnected patch cord to short the input to
> ground - not the most reliable of connections - but connecting a CV
> input set to 0V should do it too. Make sure you get your in's and
> out's round the right way - shorting the gate or trigger signals to
> ground through the switch would probably not be very nice!). You may
> still get some bleed through, but from my experiments it is
> considerably less, i.e. instead of the full signal, its down around
> 100mV or so. Also swapping inputs (I/O1 <-> I/O2) makes a difference
> too - its worse one way than the other. The problem is exceeding the
> +/-8V limit specified in the manual: despite some diodes apparently 
to
> protect the chip, some excess voltage does get applied to it, and so
> it's normal operating range is exceeded, and it thus stops working
> properly. The 'DID NOT' signals you listed will almost certainly be
> within +/-5V (I don't have a 155 or 191 so can't actually check 
these
> two, the rest certainly are); of the 'DID's the 160 and 190 signals
> are 0 - 12V, and I also strongly suspect the others are too (again I
> don't have these). If anyone else has knowledge of/can point me to a
> suitable reference for why the FET circuitry in the 4053 chip 
behaves
> in this manner, I'd love to know it!
> 
> Tim
> 
> [The views expressed above are entirely those of the writer and do 
not
> represent the views, policy or understanding of any other person or
> official body.]
> 
> 
> --- In Doepfer_a100@y..., "buechlerjoe" <buechlerjoe@t...> wrote:
> > I decided to try a number of signals patched to my A150's I/O2 
> > (Normally Open) jack. I wanted to see which ones would trigger an 
ADSR 
> > envelope by bleeding through the O/I jack, even though nothing was 
> > patched to the A150's CV jack to close the switch.
> > 
> > Signals that DID NOT bleed through were:
> > 
> > A145 LFO Pulse
> > A147 LFO Pulse
> > A155 Trigger
> > A156 Trigger
> > A190 Clock
> > A191 LFO Pulse
> > 
> > Signals that DID bleed through were:
> > 
> > A155 Gate
> > A160 Clock Divider
> > A162 Trigger Delay
> > A165 Trigger Modifier
> > A190 Gate
> > 
> > What's even more worrysome is that I'm fairly sure that I've 
> > successfully switched an A190 Gate signal with the A150 in the 
past, 
> > so now I'm concerned that the A150 is actually deteriorating in 
this 
> > regard. I wish I could be more certain about this.
> > 
> > There was an earlier response from Dieter on this issue, in which 
he 
> > ascribed the problem to user abuse of the A150. I'm certain that 
he 
> > misunderstood the symptoms, though, because he was talking about 
> > signals patched to the CV input, rather than the switch inputs. In 
any 
> > case, in was in the middle of preparations for the Frankfurt show 
and 

> > he probably didn't give the issue a lot of attention.
> > 
> > I've sent a copy of this post directly to Doepfer by email.
> > 
> > Joe

> According to the response I got from Dieter, the A150 User's Manual 
> has been UPDATED to describe the -8V - +8V restriction. I guess if a 
> product doesn't meet it's prospectus, update the prospectus ;)

Ah ha - I hadn't realised that there was a version in existence which
*didn't* describe this problem, as my manual entry does!!

> I'm wondering if it's possible to mod only half of the dual switch.

Almost certainly: I don't have the full PCB diagram with me, but there
must be a suitable point at which to cut the +8V and -8V tracks
between both halves of the 150 circuit, and then strap the appropriate
one (of the now isolated rails) to +12V and the other to ground. You
would really also need to place new 10uF and 100nF capacitors between
the new 12V rail and ground.

BUT if you are going to go to this extent, it's not really such a
great step to sorting the problem out properly (if I'm right about
it's cause I must add!!!), and then the whole module would work with
+/- 8V *and* 0 - 12 V (although the switched gate/trigger would no
longer be 0 - 12V, but I've seen nothing in any circuit so far to
suggest that this would be a problem). The problem as I see it is
because the diodes 'protecting' the chip won't conduct until the
inputs are around 8.7V: this voltage is then greater than the supply
to the chip (8V), which is what (I'm assuming) is causing it to
misbehave. A solution would be to lower the +8V to the diodes to, say,
+7V: then nothing above 7.7V would reach the chip, which being less
than 8V would (hopefully!) mean it behaves properly. This could be
achieved with another regulator of appropriate rating, but would
require quite a few straps and cuts to tracks because of the way the
diodes and chips hang off the +8V rail. In fairness, the FAQ solution
is described as 'simple' - perhaps you could persuade Dieter to post a
more 'Rolls-Royce' solution (after checking the above 'theory' out of
course)?

Tim

[The views expressed above are entirely those of the writer and do not
represent the views, policy or understanding of any other person or
official body.]

> perhaps you could persuade Dieter to post a
> more 'Rolls-Royce' solution 

I doubt that I could persuade Dieter to do anything at all, but if I 
could, I would persuade him to fix the bug, rather than just updating 
the documentation to mention the problem.

This will never happen, of course. They're too busy developing a 
friggin' Nintendo controller and the like.

If someone works out a schematic of either your solution, or the 
"split" approach, I'll buy a spare A150 and give it a try. Its way 
beyond my capability to work it out myself. 

Joe

I have done some experiments using a breadboard and a Zener diode, and
by supplying a lower voltage to the diodes ?protecting? the 4053 chip,
I was able to ensure that for a 0-12V input signal no voltage greater
than the supply reached the chip. This confirmed what I expected, i.e.
that the chip then behaves the same for a 0-12V signal in as it does
for a normal +/-5V (albeit with the 12V chopped down to about 7.5V,
but with no problematic ?bleed through?).

I could suggest a modification to the lower switch (the right half of
the board) which would allow it to work with 0-12V signals as well as
the standard +/-5V (extending to cover both switches wouldn?t be such
a great step). However this will require: 2 new components ? a
resistor and a zener diode; drilling the PCB and inserting 3 pins on
which to mount the new components; cutting 3 tracks (6 for both
switches); adding 5 wires (8 for both). This is quite involved, and
wouldn?t be for the faint hearted!
The price paid in using the Zener rather than another regulator (the
Zener has a better choice of voltage over the regulator) is an
increase in current consumption of the A-150 by about 5mA ? it also
means the mods to the board are probably easier though.

Tim

[The views expressed above are entirely those of the writer and do not
represent the views, policy or understanding of any other person or
official body.]

> > I'm wondering if it's possible to mod only half of the dual switch.
> 
> Almost certainly: I don't have the full PCB diagram with me, but there
> must be a suitable point at which to cut the +8V and -8V tracks
> between both halves of the 150 circuit, and then strap the appropriate
> one (of the now isolated rails) to +12V and the other to ground. You
> would really also need to place new 10uF and 100nF capacitors between
> the new 12V rail and ground.
> 
> BUT if you are going to go to this extent, it's not really such a
> great step to sorting the problem out properly (if I'm right about
> it's cause I must add!!!), and then the whole module would work with
> +/- 8V *and* 0 - 12 V (although the switched gate/trigger would no
> longer be 0 - 12V, but I've seen nothing in any circuit so far to
> suggest that this would be a problem). The problem as I see it is
> because the diodes 'protecting' the chip won't conduct until the
> inputs are around 8.7V: this voltage is then greater than the supply
> to the chip (8V), which is what (I'm assuming) is causing it to
> misbehave. A solution would be to lower the +8V to the diodes to, say,
> +7V: then nothing above 7.7V would reach the chip, which being less
> than 8V would (hopefully!) mean it behaves properly. This could be
> achieved with another regulator of appropriate rating, but would
> require quite a few straps and cuts to tracks because of the way the
> diodes and chips hang off the +8V rail. In fairness, the FAQ solution
> is described as 'simple' - perhaps you could persuade Dieter to post a
> more 'Rolls-Royce' solution (after checking the above 'theory' out of
> course)?
> 
> Tim
> 
> [The views expressed above are entirely those of the writer and do not
> represent the views, policy or understanding of any other person or
> official body.]