BOSS CE-3 Signal Allowance HEPL ME!)

[Rob Strand]

I have one of these pedals.  I'll try it out tonight to see how much of a volume difference there.

From what I can see the Q2 thing only reduces the gain when  Effect=ON and Mode I.  It affects both output A and B  (A=D+E, B=D-E); see JFET state table in post #6.   It's achieves the same gain reduction as R37 on the MXR flanger,
https://photobucket.com/gallery/user/bainzy007/media/bWVkaWFJZDo1MDU3MzE2/?ref=

The Boss BF-2 and CE-2 on the other hand don't bother reducing the gain and just mixes both the Dry and Effect with gain = 1.

I think the original problem is fixed but now we have a secondary problem where the delay signal is not coming through.

[midiez1997]

I checked everything near D2 and D3. Everywhere is clean.
I don't know the impedance of my multimeter ..
I checked Q4, Q2, Q5, and write results.

[Rob Strand]

I checked everything near D2 and D3. Everywhere is clean.
I don't know the impedance of my multimeter ..
I checked Q4, Q2, Q5, and write results.

What combination of Modes (I or II) and Outputs (A and B) actually produce a chorus or vibrato effect?

[midiez1997]

vibrato- A output, II.
Chorus- A output, I and B, also mode I.

[Rob Strand]

vibrato- A output, II.
Chorus- A output, I and B, also mode I.

So it's all working normally *except* for the strange turn-on delay?

[midiez1997]

Well yes..

[midiez1997]

Q2: S: 5.26(byp/eff), D:4.7(byp/eff)
Q4: S: 5.26, D: 4.82/5.26(byp/eff)
Q5: S: 5.02(byp/eff), D: 5.26(byp/eff)

[Rob Strand]

Q2: S: 5.26(byp/eff), D:4.7(byp/eff)
Q4: S: 5.26, D: 4.82/5.26(byp/eff)
Q5: S: 5.02(byp/eff), D: 5.26(byp/eff)

I am now sure your multimeter has a 1MEG ohm input impedance.

I carefully analysed the circuit accounting for the 1MEG meter impedance and my calculated voltages
matched your voltage measurements very closely.

I am not 100% sure the problem only exists in mode II.   The problem might be still present in Mode I but it is more difficult to hear because in Mode I the clean signal is always present.

Can you try this test:
- Set the Chorus settings to Rate = Full , Depth = Full.  That will make Chorus sound very exaggerated and easier to hear.
- Set the Mode switch to Mode I (not II)
- Switch the pedal between Bypass and Effect.
- On Output A, can you determine if the chorus sound is delayed (the 0.5 second delay problem).  The clean sound is present all the time so you need to listen carefully for the chorus sound.
- Repeat the same test for Output B (mode I)

Is the chorus sound delayed in Mode I on Output A, Output B, both A and B, or not delayed?

My thoughts:  The original problem had some sort of loading due to contamination on the PCB.  I think there may be some contamination still on the board.  If the problem is in Mode I and Mode II it might help us narrow down the problem.

[midiez1997]

Did as Нou said, there is no signal interruption.
It is only in mode II, well, respectively, in output A
the fee was washed thoroughly. no snot or dirt. On both sides. The diodes are all clean, neatly soldered.

[Rob Strand]

Did as Нou said, there is no signal interruption.
It is only in mode II, well, respectively, in output A

OK, thanks for confirming that.

Have you confirmed all the diodes D2, D3, D4, D5, D6, D7 are all OK?  You can use the diode check on your multimeter.   Check the voltage is good and that the diodes are not shorted.

This issue is very subtle.  The problem is related to Q4, or the signals driving Q4.  However, the thing that makes is more difficult to debug is the signals driving Q4 are not directly affected by the Mode switch!

From Reply 8:

when in output A is turned on in the chorus mode (position 2), pressing the pedal on button the sound disappears for half a second and reappears.

Because there is no sound we know Q3 is turning off immediately but Q4 is turning on slowly.

In Mode II the mode switch is *closed* (the schematic is wrong, see Reply 6).   That only affects Q3, the dry signal.  Q3 will turn on quicker.  In Mode II,  R11 (1M) is added to the output of R12 (100k) which makes the voltage to the gates of Q3 and Q5 increase by about 0.26V.  That makes the JFETs turn on harder and quicker.

I have a theory what might be happening:

In order for the BBD IC (IC3) to work correct the voltage on C20, which is set by VR3, must be adjusted to the correct value.  *** I am assuming this voltage is currently set to the correct voltage ***.    The voltage on C20 also goes to the JFETs.  The voltages from the switching circuit (Q11, Q12 upto C31, C9)  are not affected by VR3.  So the problem is, if the voltage on C20 must be set high to a high voltage (for IC3) the switching signal to the JFETs is effectively reduced.  This means the turn on for the JFETs could be delayed, or, they might not turn on fully.

The high voltage from R11 and the smaller value of R12 (100k) compared to R61 may be adding to the problem.  One side of the JFETs is switching fast and one side is switching slow which causes a gap in the switching.

So we have a few possibilities:
1) VR3 is set *too* high.  However, if the delayed signal is not distorted we cannot assume this.  We can only assume IC3 needs the high voltage.  So at this point don't change VR3.  (It might fix the switching but it will break the signal quality of the BBD.)
2) The output from the switching circuit at C31 does not switch high enough to turn on the JFETs.
3) The small R12 value making the problem worse.

I think the next step is to try adding a 270k (or 220k) resistor in parallel with R58 (56k which goes to Q11).  This will increase the drive signal to the JFETs and fix possibility #2.    If you can see an improvement but it is not enough try reducing the added resistor in parallel with R57 to 120k (or 100k).

Now, if those changes improve things but it is still not good enough I suggest adding a 1MEG or 470k resistor across R61.  This compensates for possibility #3.  Ideally we don't want to make R61 too small because you might get 'tick' sounds when switching.

If those changes work you could remove the parallel resistors are replace the original resistors with lower values.  I'll let you decide what to do here.

Good luck!

[midiez1997]

I set the correct value on the oscilloscope(vr3), and everything sounds great by ear)
I forgot to say that I redid the pedal to supply 9 volts by soldering a jumper between points 2 and 3 on the board
Further. At the same points on the board where R57 is soldered and resistor is soldered at 220k. And it gave nothing (
Then I soldered a 1M resistor to the R61, nothing has changed either (
I looked under the pcb glass board, I did not see anything bad.

[Rob Strand]

At the same points on the board where R57 is soldered and resistor is soldered at 220k. And it gave nothing (

Just to confirm was it R58 or R57?  The correct resistor is R58.

I think we need to try a more "heavy handed" approach.  We need to find out what affects that switch-on delay at all.

Then I soldered a 1M resistor to the R61, nothing has changed either (

Try a crazy value like 47k and see if there is *any* effect.

I set the correct value on the oscilloscope(vr3), and everything sounds great by ear)

Measure the voltage across C20 then record it on paper so you can set it back to that voltage later.

Now, try setting VR3 so the voltage across C20 is a low as possible where the delay is signal working but optimal.   Does this have an effect on the 0.5sec switch-on  delay between bypass and effect in Mode II on output A?   In other words does it fix or improve the problem?   Are there any positions of VR3 that improve the problem?  If yes measure the voltage across C20 at the bets position.

OK, I didn't realize you had an oscilloscope.  At this point it might be very useful.
Do you have oscilloscope probes with a "x10" switch?

If you don't have a x10 probe it might be better to leave the 47k across R61 as this will prevent the oscilloscope loading down the circuit.

[midiez1997]

Yes, to R58.
I tried to turn vr3. voltage turned 5.24v. only in the extreme position was it possible to avoid a delay. But there was no virbrato effect)) It's probably easier to live with this problem than to try to solve it.

[Rob Strand]

I tried to turn vr3. voltage turned 5.24v. only in the extreme position was it possible to avoid a delay. But there was no virbrato effect)) It's probably easier to live with this problem than to try to solve it.

It's up to you if you want to continue.  For me a 0.5sec delay would be very annoying.  It's not normal so the unit does still have a fault.

I think the way forward is to put a constant signal like a sine wave on the input then probe the points around Q3, D3, R36, IC1 (pins 1,2,3).   Do that in Mode I and Mode II, then try to see a difference between Mode I and Mode II.  If Mode I works and Mode II doesn't something must be changing.  If you can find out what is changing you might be able to work out what is causing the change, then after you can fix it.

One thing about the Boss circuit is if you probe the gates of the JFETs you can get strange readings because the gates float when the JFET is on.   As soon as you put a probe there the probe will cut-off the gate.   Any leakage or contamination on the PCB can affect the gate signal and I would not be surprised if the problem is caused by leakage between gate of Q4 and the Mode switch.

[midiez1997]

The oscilloscope is at the university and quite difficult to access it.
Can a simple way: replace any components with new ones?

[Rob Strand]

The oscilloscope is at the university and quite difficult to access it.
Can a simple way: replace any components with new ones?

The problem is a difficult one to solve and it is not obvious what components are at fault.  The oscilloscope would help trying to diagnose the problem.   We could change many parts and still not solve the real issue.

What we know is the problem is related to:
- The Mode switch set to Mode II
- Output A
- Changing from Bypass to Effect.

One idea is to:
-  Remove Q4
-  Solder a new JFET in place of Q4.  Only solder the source and the drain.
   Leave the gate hanging in the air.
- Get parts:  1M, 47n, and a small silicon diode.  Wire them together in the same way
  as R61, C31, D3.   Don't replace the existing parts.  Just wire the new parts together.
- With all the new parts off the board, solder the following wires:
    - a wire from the collector of Q11 to the new R61
    - a wire from the new C31 to ground
    - a wire from the anode of the new D3 to the gate of the new Q4

What we are doing is creating a new switch circuit for Q4.

Check the new circuit works in Mode I.   Then see if the 0.5 sec delay problem with Mode II and Output A is still present.

[midiez1997]

Hello again!) At the rehearsal I found out that the vibrato mode sounds very interesting in several songs, so I decided to finish our work)
Once again, I carefully re-read everything, and something became incomprehensible to me:
I need to drop one leg in the gate transistor.
and put 3 new parts instead of the existing ones?
then solder the wiring to them, as You said?

[Rob Strand]

Hello again!) At the rehearsal I found out that the vibrato mode sounds very interesting in several songs, so I decided to finish our work)
Once again, I carefully re-read everything, and something became incomprehensible to me:
I need to drop one leg in the gate transistor.
and put 3 new parts instead of the existing ones?
then solder the wiring to them, as You said?

You need the gate lead of the new Q4 JFET to be bent out into the air.  The gate for the new Q4 is not soldered to PCB. The source and drain are soldered to the existing PCB points for Q4.

The old R61, C31, D3 are left in the PCB.  The new parts R61, C31, D3 are mounted in the air.  The ground side of the new C31 is solder to a ground the PCB.  The new D3 goes to the new Q4 gate.

The added part duplictate what is already on the PCB.

[midiez1997]

I asked one owner, he said that he, too, has an interrupt signal ... Maybe this is a mistake in the early pedals?

[midiez1997]

I thought, suddenly they had a series of defective parts. And so I tried to solder instead of Q4 2SK117, and it gave nothing. (Of course I took into account its polarity)
Ok, but I still don’t understand how to connect with what.
I did as you advised, everything worked, but ... The problem remained !!!!
I do not know what else can be done with this pedal ....