PT-80 delay: how to add a direct out?

[geertjacobs]

I ordered a pcb and parts to make a PT-80, but I would really want to add a direct out as in the Boss DD-3 to be able to connect it to two amps.
Any suggestions on how to do this? Has anyone done this before?

G.

[geertjacobs]

Since it's a true bypass pedal, maybe I need to add a parallel buffered non-sound-coloring audio path that is in phase with the bypassed signal?

G.

[Mike Burgundy]

That's the right idea.
Add an input buffer (either JFET opamp, MOSFET or JFET discrete - high input impedance and low output imp. is what we want here), split the signal into the direct out and effect line, and hook up your pedal as if the buffer out-to-effect was the input jack, as it were.
You could run a direct line next directly from the input, but a buffer isolates whatever's before from radical impedance changes and makes sure you have enough juice to drive two inputs so level (and highs) doesn't drop when you switch the delay on.
hih

[Mark Hammer]

To turn the PT-80 into a separate wet/dry output unit, locate point A at the output of the first op-amp on the board/circuit.  Presently, it goes into a 10k resistor in the last op-amp at the mixing stage as well as to the compander chip that feeds the delay chip.  

Lift the end of the 10k resistor going to the output of that first op-amp, and run a lead from that point A to its own 1uf/1meg RC combination, just like you see on the existing output jack.  In this way, that first op-amp wil act like a splitter, sending one "copy" to the delay chip and its own output, and another "copy" to the direct out jack.

If you want to be able to go back and forth from a single dry+wet mono output to a stereo split output, wire up a SPDT toggle that connects the point S output to either the 10k resistor going into that final mixing stage OR the 1uf/1Meg RC combo and the other output jack.  Since that mixing stage at the output is unity gain, you should notice no level difference between the two output options, although there may be a little bit more top end in the separate dry output since the .015uf/1k feedback loop combo rolls off some high end for noise-reduction purposes.

[geertjacobs]

Mark,

I'm afraid I don't really get the part with the SPDT switch.
I' missing Output "S" on the schematic?

G.

[geertjacobs]

After re-reading I think I got the idea. Please correct me if i'm wrong.
Short summary:

-possibility 1: original pedal, true bypass or wet and dry mixed mono output

-possibility 2: no true bypass, separate wet output (disconnect 10K from point A) and dry output (add RC+output jack from point A)

-possibility 3: no true bypass, spdt switch switches between possibility 1 and possibility 2 by switching point A to either the 10k resistor or the added RC+output jack.

And I thought of another one to be able to play stereo all the time
-possibility 4: no true bypass. Always connect the direct out (RC+output jack) to point A. Put an spdt switch before the RC+output jack of the delay path and connect it to either the output of the final opamp stage or to point A.
This way you have either wet+dry stereo or dry+dry stereo.
Would this work?

G.

[Mark Hammer]

I read your last response quickly, but on the surface it appears to be excatly what I was just about to suggest to you.

So, dry always goes to ONE of the output jacks, and the stomp switch selects the output going to the other jack from EITHER that same dry signal OR the wet signal.

Note that the last op-amp stage is inverting.  That means that the gain of a signal in that stage is equal to the feedback resistance divided by the input resistance.  As shown, it is 10k and 10k or unity gain.  If the input resistance  is increased, its gain drops below unity.

So consider the following....

Leave the dry and wet signals connected to the final op-amp stage, as shown in the schem, but run a separate line from the 'A' point on the first stage to another output jack with a 1uf/1meg combo.  That will be your default clean output.

Instead of 10k input resistors on that last op-amp, use 1meg resistors.

What your stompswitch will now do is switch 10k resistors in parallel with one of those 1meg resistors or the other, thus dropping the level of the delay OR dry signal down to practically nil.

The advantage to this method over some others is that:

a) Your input signal (hence output too) is always buffered.
b) It's pop free.
c) There is a much better chance of balancing bypass and engage modes. (I *say* use a 1meg plus 10k parallel resistor but you may find that between the companding, filtering, etc., that some adjustment to the "engage" levels may be needed.  For instance, maybe a tiny bit of gain helps the delay signal match the clean/bypass in level, so use a 9.1k 1% resistor to parallel the wet path instead of 10k,)