MOTM

----- Original Message -----
From: <mark@...>
The controllers I have that send a positive voltage or
CC#65 (whose value I can easily convert to a voltage),
such that the portamento would be on when I'm stepping
on a pedal or playing legato.  To me, it seems very counter
-intuitive to keep my foot on a pedal unless I want an effect.
---LH-----
I agree it seems like you should step on the switch when you want to bring
ON the effect.  If you plan to do it directly with a foot pedal instead of
running the foot pedal through the controller, just be sure you use a pedal
with a normally closed switch.  Then you press on the switch, the ground is
lifted and the portamento effect is ON.   I forget, but Roland and Yamaha
are backwards.  One is normally closed.  And, 99% of any foot switches I
have ever used have a SPDT switch in them.  So, if your polarity is wrong,
open the switch and swap the connection from NO to NC.

I agree with Paul that using CV control is the best, certainly if
sequencing.  then you can not only turn on and off, but vary the times as
part of your work.

However, (WARNING DIY CONTENT) in the the DIY spirit, I will offer my
opinion of how the bypass circuit works.  It seems to me, that under normal
(not-bypassed condition), the positive voltage applied through R3 causes Q1
to operate in saturation.  The saturation C/E current is limited by R5. B/E
current is limited by R3 and R6.   Because, Q1 is conducting, D2 is
effectively shorted out and does not light.   In this condition, a positive
voltage is applied to the switch pins (8 and 9) of DG213, the analog switch.
This positive voltage causes the normal closed switches to open and the
Normal open switches to close.  So, pin 6 and 7 are approximately shorted
(some resistance) sending the lag-output voltage to the output driver
circuit.  Pins 10 and 11 are open so that the input is not directly coupled
to the output driver.

When the bypass switch is shorted (jack or switch), the voltage applied to
the base of Q1= ground.  Q1 is cutoff because the B/E junction has no
forward bias. In this condition, D2 is allowed to conduct and light
indicating a bypass condition.  Current is still limited by R5.  With pins 8
and 9 of the DG213 at ground, the normally closed switch (10 to 11) is
closed connecting the input directly to the ouput driver.  The normally open
switch is open (pins 6 and 7) so the lag output circuit is not connected to
the output driver.  Notice that the input is alwasy connected to the lag
input circuit.  So, the circuit continues to lag whether bypassed or not.
All the bypass does is select the output.

Oh!  D1 looks important to protect Q1 and DG213 from any negative voltage
that might accidentially be applied to the bypass jack by shorting that to
ground (through R4).

Here is an idea to modify the circuit as you suggested.   Disclaimer:
WARNING!!! I have not tried this.  This is just what makes sense to me
looking at the schematic.  Remove R3.  That takes the forward bias off of Q1
and causes DG213 to be in the "normal" condition.  The 820 will be in bypass
mode and the the LED will be lit.  If you apply a positive CV of sufficient
magnitude at J1, you will force Q1 back into saturation and the bypass light
would turn off. DG213 would switch to cause the lag out to be connected to
the output driver and you would have lag.

What is not clear to me (no data sheet) is how much positive voltage is
required at the DG213 switch pins to activate the switch.  It might be more
than 5 volts.  I see Paul is applying considerably more than that.  So, that
research is required.

To restore the operation of the panel switch, you could disconnect the
ground side and take that to +15V through 10K (removed R3).  But, it would
not be exactly the same. The override would be opposite.  As it is today,
either in bypass overrides the other.  The new way would be either turned on
overrides the other.

Disclaimer:  This modification description is strickly for the purposed of
education and discussion.  I do not recommend you make this modification to
your 820.  I have not tested it and my logic could be flawed.

Larry Hendry