DOD Phasor 595 mystery problem

[ovnifx]

Hi all,
I recently picked up a DOD Phasor 595 which works but has very low output, both when engaged and when bypassed.  If I really turn up the gain on a preamp following the pedal, it sounds exactly the way a phaser should sound.  But with normal levels (eg. inline between guitar fx), it is very low volume.  I spent some time with pal "soggybag" trying to diagnose it, with not much luck.  We removed a pulldown resistor from the output, and that didn't seem to change anything, so I haven't replaced that yet (you'll see two empty holes on the board).  We also replaced the resistor across the output TL022 with a higher value (it's the one sticking up near the number "92" in the second photo); that DID increase the gain at the far side of that resistor, but it did not perceptibly change the actual output signal.  I'm a novice at all this, so there's probably something obvious I'm missing- please let me know!
Here is the schematic: http://www.muzique.com/schem/dod595.jpg
Here's the circuit:

[bioroids]

Hi!

I think a good way to tackle this is using an audio probe:
http://www.geofex.com/FX_images/audioprb.gif

You keep testing the audio path from the input jack to discover in which stage the sound gets lost.

Luck!

Miguel

[Transmogrifox]

take to the board with some electronics cleaning spray.  Unsocket the op amps and clean the sockets , brush the op amp pins.  I'm suspicious that there is some kind of a short formed by some grime or similar dust build-up.  This would either directly divide the signal down to a low level, or it would make one of the op amps go unstable at a very high frequency.

Another easy thing to try is to just replace the dual op amps since they're socketed and easy to access.

Finally, you may find it beneficial to re-solder the jacks.  And what the hey--clean the input and output jack contacts.  It may be that simple:  Input/output jacks may be dirty.

Looking at the schematic, I can't see any other ways you would lose signal level in both bypass and phaser mode (assuming all the components are stock.

You would be best off putting the output pulldown resistor back.

If everything else fails, I can tell you an easy way to boost the gain on the pedal.  It would not be a good idea to do this since it would not address the real problem--we'd just be creating a work-around for the symptom.

What's the input impedance on your amp, by the way?

[ovnifx]

Thanks for the tips guys...
I will spend some time cleaning everything this weekend.  I didn't try an audio probe per se, but I did follow the signal with a scope probe while a test tone was running through the pedal to a guitar amp.  Weirdly (prob. because of the design of the bypass switch), there were several points in the circuit where touching it with the probe would cause the effect to switch in or out.  This made it tricky to determine exactly where the signal drops off, because the effected signal has a much wider Vpp swing than the uneffected signal, and varies widely.  So the apparent "good" and "bad" levels were hard to pin down.  I'll try that again too.  I don't know the impedances, but I have been using a variety of test interfaces including other pedals, bass preamps, mic preamps, etc.

I''ll post again with the results after it's all clean- but if anyone else has any ideas in the meantime, I'm all ears!
Thanks!

[soggybag]

The pedal is not true bypass, so I think the signal level would be set by the output buffer. Everything sounds like it is working, the problem is the output, which is low whether the pedal is engaged or not.

I think adjusting the 22K resistor on the op-amp in the lower left of the schematic should set the gain of the output buffer.

There is a 4K7 in series with a .055 cap right above this 22K resistor. I'm not sure what this does? It might have something to do with the gain.

Also, it appears to me that the input buffer uses a non inverting op-amp, while the output buffer is an inverting op-amp. Is this done on purpose to enhance the phase shift sound or not?

[christian]

Also, it appears to me that the input buffer uses a non inverting op-amp, while the output buffer is an inverting op-amp. Is this done on purpose to enhance the phase shift sound or not?

I guess a design flaw. non-inverting amp gives you higher input impedance for the input buffer, and its cheaper (well, one resistor) to make a mixer from inverting opamp, as in the output op-amp. What I wonder is why the "Tone" knob affects the signal even when bypassed?

Have you tried to change a fresh battery in?

ch.

[George Giblet]

Assuming the effect works when switched to the effect position....

Check for short across 33nF (?) cap that is in series with the 4k7 resistor in the feedback loop of the output opamp.  The cap could be faulty or there is short on the PCB.

Beyond that measure the DC voltage on the supply pin of one of the opamps then check that the DC voltage at the output pin of all opamps (except the LFO stuff) is at half that voltage.

[ovnifx]

What I wonder is why the "Tone" knob affects the signal even when bypassed?

Have you tried to change a fresh battery in?

Interesting you should ask about the tone knob... I hadn't mentioned it, and I hadn't even noticed it before, but you are exactly right: the tone knob does affect the signal regardless of whether the effect is engaged.  Strange!

I've been using the original DOD 20V ac adapter power supply, not batteries.

I gave everything a really good cleaning: Caig's Pro Gold, toothbrush, abrasive pad.  Pulled the ICs and scrubbed their pins.  No change to the signal!  Oh well, at least now it's all clean and we can (more or less) rule out the grime factor.

I put the pulldown resistor back in- actually, I put in a 100K as spec'd on the schem; there had been a 10K installed originally.  I tested the 33 (or 55?) nF cap leads with a continuity tester and found no short, and no shorts anywhere else in the vicinity.  I tested the source and output voltage for all opamps, and they came up clean: 13.08V source and 6.55V output on all of them.

To recap, the signal is very low level both when the effect is engaged and when bypassed.  There is no appreciable volume jump or drop when switching the effect on/off.  The LED and footswitch operate normally.  The signal quality is very good- the tone is nice and full-range and low-noise, and sounds really sweet.  The effect sounds like it is operating exactly as it should.  Even when I feed it into a mic preamp and crank the gain to get a "normal" signal level, the majority of the noise present is the noise inherent in the preamp (similar noise level with nothing plugged into the preamp).  So I guess I just have to redouble my efforts to find exactly where the gain drops.

I have put in an order for some TL064's (per a recommendation in another thread) and a replacement 4007, so we'll see soonish whether they have an impact.

[George Giblet]

> I have put in an order for some TL064's (per a recommendation in another thread) and a replacement 4007, so we'll see soonish whether they have an impact.

There's seems to be too many good points for the opamp or switch chip to be fried - nonetheless these are possibilities.

I'm thinking you should check everything around the output. ie.  1k resistor, especially the 1uF cap, and the output wire and the output socket.  The do the same at the input ie. input socket 10n cap and 10k resistor.

A simple test would be to wire a cap, 100nF to 1uF, directly to the lead on your amp, then touch the other end ("probe") of the cap to the output of the last opamp.  This bypasses all the suspect components at the output.  If the signal is still low, try connecting the "probe" end of the cap to the output of the first opamp.  If still bad connect the "probe" to the input of the first opamp.  Progress through to the input socket.  If you get way back to the input socket then it could be the input socket or a short.  Desolder the input socket wire if it comes back to life then it's something around the input stage of the PCB if not it's the input socket.

[ovnifx]

OK, problem solved!      It was the last 1uF cap right at the output, dead as can be.  I don't know why I missed it the first few times I checked the parts in that area, but somehow I did!  Embarrassing, but that's how it goes.  I found it using the good ol' audio probe, so it is conceivable that I was just confused by the readings I got by using the scope probe (how I tested it the first few times).

NOW my problem is that the output peaks are too hot!  When the resonance is turned up past 50%, the peaks are huge!  I don't want to lower the overall output level, so I'm just going to have to run a limiter after this pedal.  It sounds killer, though.  I had planned to modify it for bass frequencies, but no mods are needed- this thing has great low end.

Of course it's not true bypass (or any other really good bypass) and there's the darn tone knob affecting the "bypassed" signal...  Maybe I should try to clone it into a box that solves those two problems, with a limiter built in!  Can anyone recommend a good simple/small limiter circuit?  (I will go search the archives too.)  Thanks guys!

[christian]

NOW my problem is that the output peaks are too hot!  When the resonance is turned up past 50%, the peaks are huge!  I don't want to lower the overall output level, so I'm just going to have to run a limiter after this pedal.  It sounds killer, though.  I had planned to modify it for bass frequencies, but no mods are needed- this thing has great low end.

You could try higher values for the 100k resistor after the Regen knob that goes back to the first allpass stage. Make a socket for that resistor out of 16-pin DIP socket and plug in different values, higher the resistor, less feedback.
For a rough guess, if you say after 50% its getting loud, 200k resistor would kill it nicely, maybe slightly lower like 180k if you wanna go wild someday

ch.

[Transmogrifox]

That would be kind of irritating about the tone knob.  I don't know why on earth they designed it that way.  I guess a good re-housing job would be the way to go.  Do a true bypass switch and add a volume knob to kill some of the peaking on resonance.

It must be a relief to have found the problem, though.