How to make boss ge-7 true bypass?

[DVB_master]

Hi to all! I'm a newbie here! 

I'm going to build the boss ge-7.
Here's the schematic posted by cd:
http://img267.echo.cx/my.php?image=ge7pg28oc.jpg

I have two questions:

1. Could someone tell me how to remove the fet switching in order to use true bypass?
Should it work if I do something similar?

Removing everything inside the red zone and connecting the components following the blue arrows

2. The pots used in the original scheme are graphic pots. Does it change something if I use linear pots?

Many thankx to all!

[Mark Hammer]

As a newbie, I will bounce this question back at you: Why do you feel you need to make it "true bypass"?

We have had much discussion about the relative merits of TB vs FET switching here over the past year and a half or so, and the emerging consensus has been that FET switching, and the accompanying always-on buffer, is not the universal demon spawn it's cracked up to be, and TB switching is not the universal panacea it's cracked up to be either.  Each are good for different sorts of reasons and applications and occasionally what people THINK will be better isn't necessarily so for their particular use or needs.

So, if there is no compelling reason for converting it to TB, then the sensible thing is to spare yourself the aggravation, cost, and risk, and just leave it as it.  I'm not saying there IS no compelling reason, or could never be.  I'm merely saying that you shouldn't automatically invite heartache and burden just because you heard from somebody or read somewhere, from a guy who heard from another guy, that TB would be the answer to all your prayers.  Taking a few moments to consider whether you NEED it might be time well spent.

By "linear pots", I assume you mean rotary type?  The slider pots in there already are linear in their taper.  Cut/boost situations such as found in an EQ are probably the most dependably likely to use a linear taper, simply because the mid-point of sliding or rotation has to *mean* something (i.e., no boost and no cut).

And welcome on board!

[Dirk_Hendrik]

a very good post as usual

But in the case you have a reason for TB other then expecting miracles;\

The first option is to drill a huge hole and dump in one of those 3pdt switches. After that (and doing the alternative wiring) the cosmetics of your pedal are ruined forever and possible resale value of the padal is close to zero.
The second option is using a small 2PDT relais for the signal switching and using the original flipflop circuit in the pedal for the control ogf this relay. By using this approach the mod is 100% reversibe in case of a screwup.  I did TB's this way some 20 times over the last few months. While I fully agree with Mark on the "why do you want your pedal TB-ed" I experienced that there is quite some money to be made in TB-ing and it does therefore generate the cash to maintain this hobby...

[R.G.]

Option 3 is a true bypass box...

[Mark Hammer]

Correct as usual your majesty (that one is for all the aging Mister Rogers fans out there.    )

The bypass box or "loop selector" treats whatever you stick between the send and receive jacks as if it were a single circuit board being bypassed.  The advantage is that you don't have to monkey around with the GE-7 or anything else you stick in the loop.  So, no scratches where the drill bit wandered.  No torn wires where you forget where they were originally soldered to.  No aluminum tidbits finding their way into pots.  No trouble-shooting.  And no loss of resale value.  Plus the loop selector will continue to have usefulness on its own, despite whatever you may have as part of your rig in the near or distant future.  When newbies ask for an essentially foolproof first project, I almost always recommend a loop selector, chiefly BECAUSE it is so useful in so many ways, and so easy to make.

I think there is a little something to what Dirk notes, but I think the market for untampered-with pedals is probably larger than the market for modded ones.  True, someone might pay more for a modded pedal, but they have to buy it from you first for that to be a worthwhile effort on your part.  Personally, I'm inclined to go with whatever results in the greatest number of potential buyers down the road, rather than what might extract the highest price from any single buyer.  But that's me.

[DVB_master]

MANY THANKZ for your replies.
I don't have a ge-7 but i'm going to build it, so i was wondering to eliminate the fet switching from the original schematic so i could have less components to solder and true bypass switching... 
do you think that i shouldn't remove the original switching method. In this case what kind of footswitch I should use and how I should wire it to the PCB?

[Seljer]

Ah, if you're building then you should leave out the Boss bypass stuff and just use a regular footswitch. Makes things easier.

Q4 and Q5 and the FETs that make the main switch. Q1 is an output buffer that you can drop out. I'm trying to figure out what goes where at the input of the pedal.

C19 would be the output of the pedal.


You might also be interested in something like http://generalguitargadgets.com/index.php?option=content&task=view&id=216&Itemid=241 (which has a nice layout allready made)


edit: heres a cleaned up version of the boss that I think should work http://g0g0g0.dyndns.org/uploads/050114/ge7pg28ocmod4pe.gif
though theres one opamp too much in there to do it with just 2 quad opamp chips

[Dirk_Hendrik]

I think there is a little something to what Dirk notes, but I think the market for untampered-with pedals is probably larger than the market for modded ones.  True, someone might pay more for a modded pedal,

Couldn't agree more. The market I referred to is the market of people who want their stuff TB-ed. This indeed doesn't have anything to do with the market for selling modded pedals which is, unless your name is keeley or pierra, quite limited.

[DVB_master]

I agree with you also. I wouldn't touch at all my boss ge7 if I had it!

[Mark Hammer]

Well that kind of changes things, now, don't it?  Glad you supplied the missing critical information.

Yes, the circuit should work just fine without the circled components.  The TB is not critical to the circuit.  Our collective advisory assumed you had a functioning GE-7 and were set on ripping it apart. 

The following other things should be kept in mind about the circuit.  That whole circuit built around Q1 doesn't need to be there.  Indeed, you can go from pin 1 of IC4 to C19, run a 100k resistor to ground and be done with the everything else to the right of it.  Much of the stuff yu see adjacent to the blue lines you've drawn is just there because of the use of FETs.

Naturally, the bypass switch will go to R30 (pad '1' shown on the periphery of the drawing) and to the output, which will now be the junction of C19 and the 100k resistor to ground.

Note that since you aren't using FETs for switching, that path that starts from pin 1 of IC5 and continues along C25 is completely unnecessary (although you'll still need the R20/R23/C18 network for correctly biasing the op-amp).

The TL022 chips shown are low current chips and are not likely essential for the circuit working.  Install sockets and try out your garden variety TL072 or 4558s first, and if it doesn't work THEN order some TL022s from somewhere.

Likewise, the HA1457 is not super critical to the unit working.  Just about any decent single op-amp will be fine.  Just note that if you opt for a different op-amp, they will likely have a different pinout and C21 will not be needed.

In case you have not already commited to this circuit, consider the simple easy parametric EQ circuits posted over at geofex. I can't see where their performance would be any less than the GE-7.  You may find as well that a couple of bands of semi-parametric (i.e., cut/boost and tunable centre frequency) turn out to be more flexible and capable than 7 bands of fixed-frequency EQ.

[DVB_master]

OK Thank you very much, man! 

Should this work fine?

[Mark Hammer]

Bingo.  Everything inside the red line is expendable.  You learn quick.  I like that.

[DVB_master]

  Just another question: is C19 necessary or not?

[Seljer]

Yes, its the coupling capacitor, It stops the bias voltage for the opamps from getting out of the pedal.

You also wouldn't need anything thats right of C18 (where you drew that blue line) if you took out all the stuff circled in red

[DVB_master]

Do I need a 1M resistor between C18 and C19?

[Seljer]

Nope, thats there for the benefit of of Q4, which you're taking out of the circuit so it has no purpose and it'd mess with other things if you did put it there...

[DVB_master]

OK thanks to all. The connections made with the blue lines are ok?
If I use rotatory 10k linear pots instead of the sliders, where is the point of non cut/boost that is equal to +0db. I think the +15db (max boost) is at the end of the rotation, and the -15db (max cut) at the beginning... Am I right

[Mark Hammer]

Correct.  Make sure you buy linear pots.  These should show 'B' after the value (i.e., 10kB) in most instances.

[DVB_master]

edit: heres a cleaned up version of the boss that I think should work http://g0g0g0.dyndns.org/uploads/050114/ge7pg28ocmod4pe.gif
though theres one opamp too much in there to do it with just 2 quad opamp chips

Should I connect C18 with C19 ?

I have found another schematic of the boss ge-7. Here's the link:
http://www.freeinfosociety.com/electronics/schematics/audio/bossge7.pdf
It says that the HA1457 is for pre-emphasis and gives +3db @ 2 kHz and the Q1 is for de-emphasis.
If I remove them both, what will happens?

Each pot of the boss ge7 boost or cut the volume at a such frequency of 15 db. If I remove pre-emphasis (As Seljer as suggested.), will the gap of boost/cut be 12db?(that is also an acceptable value ).

I think I'll follow Mark's advice and I'll substitute TL022 with TL072.

Thanks

[MartyMart]

No, C18 now just connects to "ground" it's part of a "voltage divider" along with R20/R23 to give a 4.5v "Vb" line
This 4.5v "half voltage" biases the opamp through the resistor R29 ( 470k ) at opamp pin 3
From C19 you can go direct to the output/100k to ground ( R12? )

Marty.