Having hiss problem with millenium 2+ bypass

[henry33]

Hi,

I have a garagetone tremolo pedal which has developed a loud noise floor in the form of hiss & some static noises. It was sitting in the closet  for years because of this but I wanted to see if it can be fixed.

I have other pedals from the garagetone series which give a loud hiss if the input is not plugged in. This is with a 9V adaptor. The noise goes away when a plug is inserted at the input. I've not checked this using a battery. But is common with their phaser, drivetrain & chainsaw pedals.

The tremolo used to do the same as above & the hiss would go away when the input had a cable plugged in it. Now its constant hiss without a cable plugged at the input. I checked the in/out sockets, did not find problems there.

Recently I tried to reflow the solder at every point. Before that using a battery would give a lesser volume of hiss but since I reflowed the solder, the hiss is present while using the battery as well & its at the same volume level as using a adaptor supply.

I feel its a problem with the switching & was hoping if it were possible to remove the buffer bypass & hope the pedal works normally.

The issue is present with the pedal in bypass state. In effect-on stage I can still hear the tremolo effect masking the hiss & pulsing the hiss.

I could post pics of the board if required highlighting the traces. Any ideas if the MOSFET is malfunctioning? The visual sound stuff has used millenium 2+ bypass buffer I think.

[R.G.]

To the best of my memory, the Visual Sound stuff in general and the Garagetone series do not use the Millenium 2 bypass. I will go look. But I think it's something else. The Garagetone line was not particularly hissy back when they were made, so something has drifted off. It may take a bit for me to refresh my memory. I'll post back when I get that stuff pulled out and in mind.

Followup:
I remembered right. The Garagetone line does not use the Millenium Bypass in any form. The bypass is a hard switch driving two P channel JFETs on and off. The tremolo is an optical type, LED driving an LDR.

Just guessing, it may be that the PNP which switches power when the input jack is plugged in has become leaky and is letting a trickle of current leak through when it's not turned on. Or, possibly, your amp doesn't like the signal conditions when the circuit is not turned on. If there is no input plug, there's supposed to be no power to the circuits, and that was true when the pedals were manufactured, as I wound up personally doing first-run quality inspections.

It could also be an issue with the power. Two questions: which 9V adapter exactly are you using, and could you try it with battery power to get more info for debugging?

[R.G.]

The devil that sits on my left shoulder and jabs me with its pitchfork is pressuring me to ask: why do you listen to the pedals with no input plugged in?

[henry33]

Thanks R.G. Keen.

I've tried it with an EHX adaptor & duracell battery.

On battery power there was less hiss until I resoldered all possible joints including the smd components(using a micro solder iron) But it was still more than the other pedals in the series.

There is an audible pop when connecting/disconnecting a cable from the pedal into the amp input. Will have to check again if the others do this. Also the switch isn't silent, it gives a pop.

There's also some static that I can hear, frying sort of, like a component is failing, hard to describe, it's not as loud as the hiss though. Hiss is like being at a beach lol.

Well there have been times when I've tried listening to the VS buffer compared to MXR & Digitech. That's when I had noticed that the garagetone phaser would hiss without a input cable plugged into it hehe

I thought it was the buffer that had developed a fault since I had noticed a CD4049 IC on the board & it might have failed, so I assumed they were using the Millenium Bypass 2+ in there.

[R.G.]

I've tried it with an EHX adaptor & duracell battery.

On battery power there was less hiss until I resoldered all possible joints including the smd components(using a micro solder iron) But it was still more than the other pedals in the series.

Hmmm. With less hiss on the battery and more on the adaptor, it is possible that the bypass filter capacitor inside is starting to leak.

There is an audible pop when connecting/disconnecting a cable from the pedal into the amp input. Will have to check again if the others do this. Also the switch isn't silent, it gives a pop.

That's another vote for a capacitor starting to leak. It means that there's DC on the output jack, and that is definitely not true by design.

Switch pop is again unusual. The first batch of these used P-channel JFETs for signal switching, and we had to change to CD4053s for later batches because we could not get the P-channels in quantity in a reasonable time for production. The P-fet versions were dead quiet. The CD4053 switching is not totally quiet as I could see the transition on a scope, but it was remarkably hard to hear even at full amp volumes. So something is still wrong. That may mean that the Vbias circuit has started drifting off.

There's also some static that I can hear, frying sort of, like a component is failing, hard to describe, it's not as loud as the hiss though. Hiss is like being at a beach lol.

Well there have been times when I've tried listening to the VS buffer compared to MXR & Digitech. That's when I had noticed that the garagetone phaser would hiss without a input cable plugged into it hehe

I thought it was the buffer that had developed a fault since I had noticed a CD4049 IC on the board & it might have failed, so I assumed they were using the Millenium Bypass 2+ in there.
[/quote]

Is there a CD4053 in your unit? That tells the difference between the Pfet version and the CD4053 switched version.

In either case, I think the data points to one or more failing electro caps, and possibly a leaky power switching transistor.

If it were me doing the debug, I would replace the aluminum electro caps first, then see if the issues are still there. Replacing the caps will never hurt unless the replacement process damages something. Then I would check for DC on the output jack and read the actual voltage level of the DC bias and see if that changed when the effect is switched in and out.

Finally, for a belt and suspenders kind of thing, you might parallel the feedback resistor of the input opamp stage with a small capacitor, something like maybe 100pF to knock off some ultrasonic response. I've found that this helps with some opamps.

There typically isn't one "The Visual Sound buffer". They're similar, but were all custom designed to fit the circuits. Sometimes they're identical, sometimes not.

You might try plugging your guitar into the input and seeing if that helps the sound any.   

[tubegeek]

You might try plugging your guitar into the input and seeing if that helps the sound any.   

I know that when *I* try that, I get a whole slew of NEW problems.

[henry33]

That's another vote for a capacitor starting to leak. It means that there's DC on the output jack, and that is definitely not true by design.

Switch pop is again unusual. The first batch of these used P-channel JFETs for signal switching, and we had to change to CD4053s for later batches because we could not get the P-channels in quantity in a reasonable time for production. The P-fet versions were dead quiet. The CD4053 switching is not totally quiet as I could see the transition on a scope, but it was remarkably hard to hear even at full amp volumes. So something is still wrong. That may mean that the Vbias circuit has started drifting off.

There's also some static that I can hear, frying sort of, like a component is failing, hard to describe, it's not as loud as the hiss though. Hiss is like being at a beach lol.

Well there have been times when I've tried listening to the VS buffer compared to MXR & Digitech. That's when I had noticed that the garagetone phaser would hiss without a input cable plugged into it hehe

Is there a CD4053 in your unit? That tells the difference between the Pfet version and the CD4053 switched version.

In either case, I think the data points to one or more failing electro caps, and possibly a leaky power switching transistor.

If it were me doing the debug, I would replace the aluminum electro caps first, then see if the issues are still there. Replacing the caps will never hurt unless the replacement process damages something. Then I would check for DC on the output jack and read the actual voltage level of the DC bias and see if that changed when the effect is switched in and out.

Finally, for a belt and suspenders kind of thing, you might parallel the feedback resistor of the input opamp stage with a small capacitor, something like maybe 100pF to knock off some ultrasonic response. I've found that this helps with some opamps.

There typically isn't one "The Visual Sound buffer". They're similar, but were all custom designed to fit the circuits. Sometimes they're identical, sometimes not.

You might try plugging your guitar into the input and seeing if that helps the sound any.   

Thanks R.G Keen, the board says Second Issue. It has three op amps, LF353, 0728(B)eZ945 & OPA 2137U. It has the CD4049UBCM & VTL5C3/2 vactrol. These vactrols are no longer in production, so I doubt VS will be reissuing this pedal again, perhaps they'll come up with a better one  The CD4053 is in the phaser I think..

I changed all electrolytics in there. It took a long while, never had to put so much effort in sucking out the solder from the holes lol Working on smd based PCBs is, well, aggravating because the board itself is tiny & stuff cramped in. Anyways, I did one cap at a time hoping to nail the offending one (I find it better to change one thing at a time while debugging) 
Results at the end proved to be more or less negative. The hiss went down slightly, could be my mind playing tricks lol but the intermittent frying eggs sound had not changed in any form. Well atleast I don't have to worry about any filter caps going bad soon 

However, good news is that I think I've found the issue. Remember when I said I reflowed the solder joints everywhere & after that the hiss level became constant & loud. Well, I removed the in/out wires connecting to the PCB this time. This was also a time consuming job cause of the glue that VS used at the contact terminals of the pcb & wires. I had done this before once & the problem had reduced but then came back on its own. Anyways, I cleaned it all up & resoldered  them back, the hiss all went away    I feel the problem was(hiss) or is(frying eggs) at the in/out points of the board, probably a bad solder joint or maybe a failing smd component.

Now I've not checked if the frying sounds are still there, the pop soud is still there when connecting to the input of two different amps. I checked to see if there was DC voltage on the output, I got a reading of 0.08V once but I didnt get it again, could be the meter, will check it again in the morning. Haven't checked if the switch pops, too sleepy hehe

Will update if more gremlins are at work in the night 

[antonis]

I checked to see if there was DC voltage on the output, I got a reading of 0.08V once but I didnt get it again, could be the meter..

Meter impedance discharged output capacitor..
(which capacitor shouldn't carry DC charge..)

[R.G.]

Good work. As you say, the caps are now all new and guaranteed not to be the problem. No harm, and some benefits in longevity. If cleaning up and/or resoldering the lead wires cleans up hiss - good. Bum solder joints on lead wires or nearly-cracked-through wires are a known source of noise.

80 mV on the output is plenty enough to cause a pop. You'll want to do some more measurements of that, switching the effect in and out to see which variant causes the DC offset. And as Antonis says, your meter can pull down a DC offset all by itself, so you have to note the voltage at the instant you touch the contact.

You're well on your way to fixing it!

[henry33]

Sorry took awhile to get back.

I took some measurements.

With the power off to the pedal:
Power On= as high as ~129mV to ~64mV to ~29mV were the highest readings I observed got at the output over a few power on/off tries. It would then continue to fall to around -0.08mV to -1.2mV while the power was on.

Turning the effect On= it seems to constantly jump between 1.2mV to 0.2mV or so.

Turning Power off= it keeps dropping from 10mV to 0V, like if one were to meter an electrolytic cap out of circuit & watch it drain. The output cap reading this one I guess.

What could cause DC voltage to leak on the output of the pedal  & is it bad for the amp its connected to in any way?

Also quick question, this IC labelled 072(B)eZ945 is it a tl072 based opamp or opa728 or njm072B? The pinouts on the njm072B are not the same a tl072, just curious, my gut says the eZ945 is a tl072 variety. I couldn't find much(or any) info about it online.