Hello folks, I'm here again!
Today with an updated version of the Boss SG-1 "Slow Gear" attack delay.
I received many mails from people wanting for a tested verison of the layout, and
so it is. I completely redesigned the layout, with some improvements in the 
circuit, true bypass and, more important, the board is now 30% smaller than the 
other version.



General notes
-------------

There are few changes from the earlier version. 



Notes about the schematic
-------------------------

Unlike the earlier version, I've drawn using the service manual only, so, few components  value changed in the updating process.
These components are:

C2:  1uF EL   --> 0.022uf
C6: 0.47uf EL --> 0.47uf
R14: 1k       --> 470
R16: 3.9k     --> 390
TR1: 25k      --> 10k

For those who built the older circuit, I strongly suggest to change these
components, since they allow to have longer attack delay time (in my prototype I
reached amore than half second of  delay), as well as they improve the
sensitivity and the response of the circuit.
Note that I changed the components name, so pay attention in doing the 
substitutions.
As previously noted, I removed the bypassing circuit, since many people wanted a
true bypass verision of the SG-1. The effect is always engaged, so connect it
with a standard 2P2T switch.
Last, I used my default template for drawing the schem... it looks again ver
vintage...



Notes about the components
--------------------------

Almost all the transistors are all BC549 or BC550. Other NPN transistors can be
used, but pay attention to pinouts. BC549 and 550 pinout is different from 2N3904
and 2N5088, so if you plan to use these, change the orientation of the trans
(remember: BC549 is CBE and 2N3904 is EBC).
For best result, use matched transistors for T5 and T6, so use a couple of 
transistors with almost equal hfe (it's not necessary, but reccomended).
Take care of the Jfet: T2 is a 2SK30A, very common in Japan, but very rare in
Italy. I've had only two of these and I couldn't find them anywhere else.
Many reported that changing the fet would lead to bad functioning or not
operating behavior, so, be sure to use this particular fet. I tried BC264 and 
BF245, both with very bad results.
Also, twisting the trimmer will lead us to different behavior of the unit. There
was a very accurate trimming procedure, using a signal generator and a
multimeter, but I think that you can set-up the trimmer by ear (that's how I
did...). For a correct calibration, move the "Attack" pot fully clockwise
(maximum attack) and the "Sensitivity" pot to 12 o'clock (halfaway between heaven
and earth).
Lotsa capacitors are polaryzed.
Choose a low noise single op-amp IC, like a TLO61 or a TL071. TL081 or a 741 will
work nicely, too.
I reccomend to use sockets for T2 (jfet) and the IC.



Notes about the PCB and the component layout
--------------------------------------------

As always, that's a hand drawn layout. No auto-route software used. Just a blank
piece of sheet, a pencil and a very old version of Photoshop.
The blue line is the GROUND, the green line is +9 and the brown one is +4.5.
This board is based on the old one, so it's correct and tested. Many people had
built the other one with no problem, I hope this will happen with this version,
too.
The board is 30% smaller than the other, it will fit in Hammond boxes.


As always I'm very glad to hear any kind of comments, corrections and 
improvements of this project. My e-mail address is 
offramp.unoSPAMMERKILLER@tiscali.it 
write me and remove the word SPAMMERKILLER from the address!
Bye from Italy!