Where Is My Signal?

[KazooMan]

I am having a bit of a problem with my build of The Valve Wizard's U-Boat Sub-Octave.  It works great with a signal generator, but barely passes a guitar signal.

When I tried it with a guitar, it didn't seem to be passing any signal in bypass or effect on.  I have since determined that it passes a bit of clean signal if I dime the amp.  This is a new problem for me. 

I did my usual initial check of the board and I do not see any problems with the traces or my solder job.  All of the component values are correct.  I have confirmed the values of key components that could be draining the signal off with a meter. 

OK, time for the audio signal probe.  When I have my sine wave generator on the input the pedal works great!  The traces on my oscilloscope while adjusting the pot that blends the dry and wet signals is really sweet.  It starts with a very nice sine wave and transitions through some interesting combinations of the wet and dry and finally on to a complete wet octave down signal.  There is a switch for regular or synth mode.  This makes subtle effects on the oscilloscope trace, but both settle in on the octave down.

So, I have a pedal that works great with a significant level input signal, but not with the guitar.  BTW! I did confirm that the  guitar works fine straight into the amp.   The audio probe doesn't offer much help. But perhaps I can go back and measure the actual level of the signal along the path. 

Can anyone offer a gentle push in the right direction to figure this out?  BTW, it is not the usual sort of pedal with a true bypass.  The dry signal passes through.  Hmmmmm........ Perhaps THAT is telling me where to look!  The circuit and opamp that control the mix of dry and wet may have an issue. 

[Govmnt_Lacky]

From what you describe, I can be fairly certain that the circuit is NOT true bypass correct?

If you are getting very low output when in bypass mode, then it narrows the problem area pretty significantly.

Schematic LINK??

[KazooMan]

Yep, it is not true bypass.  Also, the signal is low in bypass or effect. I forgot to mention that I swapped out all of the ICs with no success.  

I am certain that I am in for a "dope-slap" moment when I finally figure this out.  

You should see the scope traces while turning the mix pot. It is amazing how well this pedal tracks the dry signal and mixes a perfect wet signal.  I am just losing the level somewhere along the path.  I have used an audio probe many times to see where the signal dies, but I haven't needed to monitor the actual level of the signal before.  Perhaps this is the first time!

[KazooMan]

I am on my Ipad and pasting the URL seems to be a chore.

Try

www. valvewizard.co.uk/uboat.html

[puretube]

http://www.diystompboxes.com/smfforum/index.php?topic=93387.0

[Govmnt_Lacky]

Double check all the resistor values....

You sure you have a 100R after the output buffer? Not a 1K, 10K, or 100K?

[MrStab]

IF all else fails (^wrong resistor values seem to be a common cause of this sort of thing), what happens if you remove the JFET nearest the output?

[KazooMan]

Thanks for the suggestions.  I have confirmed that all of the component values in the dry signal path are correct.  I'll try pulling the JFET this evening if I don't have any other success.  I plan to follow the signal with my oscilloscope to see just where the loss sets in.  My ICs are socketed, so I guess I will try one more swap.

[KazooMan]

Interesting.  With the signal generator input I can't find any place along the wet signal pathway where the level drops off.  As I indicated, I get a nice octave down effect on the wet side.  The input level is a couple of millivolts.

I reflowed all of the solder joints.  No change.  However!  I was pushing all of the ICs firmer into the MillMax sockets and I could get the guitar signal through, albeit momentarily.  I have pushed, flexed, etc. and I cannot locate a specific spot that is the problem.  I have looked at all the joints and traces with a magnifying glass and Measured continuity with a meter and I don't see any issues.  I swapped out two of the ICs with no joy.  Why the higher level input signal passes well but not the guitar remains a mystery.

I have never had a problem socketing ICs, but I am beginning to think that I have my first case.  These are good MillMax sockets not cheap leaf spring style, but perhaps one is not making a good contact.

So, I am at the point where further messing around with this build seems to be a waste of time.  It will be quicker to etch a new board and transplant the components (not the sockets!).

Thanks again for all of the help.  I will let you know when I am playing really LOW in the future.  

BTW, I just ordered components for RG's Neutron Filter.  Should be fun.

[duck_arse]

don't chuck it yet!

how much signal do you have on the mix pot in front the out buffer, both ends, and middle? pull it out and short the input buff output to the out buffer input. what happens?

are you getting the diagrammed sig at the first j112/10k/100k trimm? and at its wiper?

[anotherjim]

Check against the schematic linked to above. I'd guess you have too low an input impedance - the guitar won't like that but the sig gen won't really care. ONLY the 1Meg resistor should be on the input opamp + input - NOT the 10k resistors forming the bias supply.

[KazooMan]

Thanks for the additional suggestions.  With the signal generator hooked up I have a good signal at the output of the input buffer, at all of the terminals of the mix pot, and at the input and outputs of the output buffer.  The mix pot seems to work fine, smoothly transitioning from 100% dry to 100% wet signal as measured at the output jack.  I measured the resistance of the pot, wiper to both ends and it is fine.  Bridging the input of the pot and the wiper out to the output buffer had no effect.  I do have a good signal at the trim pot and it's wiper.

As far as the input impedance goes, that is a great suggestion and it is where I first looked for a problem.  The template for the board is from Merlin's site and it appears to provide the proper connections.  My R1 is 10 M from the input to ground as indicated in the schematic.  Merlin's block diagram (in the link above and in the link above to another thread on this forum) does not show a 100 ohm resistor from the input to C1, but it is on the schematic.  I have this resistor in place.

Your comment led me to notice something odd.  If you go to the other thread in the forum from the link above, the first posting is from Merlin.  He has a link to a set of images for the board, block diagram, and schematic.  On those, the resistors R4 and R5 that are the voltage divider for the bias supply are listed as 10K.  I downloaded a PDF from Merlin's site.  In that documentation the block diagram shows the resistors as 10K, but the schematic and parts list have them as 1M.  My build uses the 1M resistors. I don't see how this would be causing my problem, but it is odd.

From the poking and prodding and flexing the board I have had moments of passing the guitar signal.  This leads me to think that the problem I have is mechanical.  For the life of me I cannot find any transient shorts or any hairline cracks in traces that would open with flexing.  As I indicated, I have reflowed all of the solder joints.  They looked fine and still seem fine. 

I promise not to chuck the pedal, but I do think I will etch a new board and transfer the components.  The only times I have had to do this in the past was when the original template had very thin traces.  I printed this one straight off the PDF.  I think I will take it into Photoshop and look for any places where there may be a problem with narrow traces and beef them up a bit where I can.

Smallbear says the Neutron parts are in the mail today, so I had better get this one wrapped up fast!

[duck_arse]

I saved a local copy of the thread, so this is the circuit I always see from his initial post ....



?measure the resistance from the in-buff input pin to "4.5V", and from in-pin to V+ and to ground? did you show us photos of your board?

[KazooMan]

Yep, I now have that copy of the circuit and the one from the PDF that shows the bias supply resistors as being 1Ms.

Sorry, but I bit the bullet and already started moving components, so I can't give any resistance reading.  I appreciate all of the interest and helpful suggestions.  I will report back when I get the new board populated.

Thanks