Maestro FZ-1 HF noise

[andrewnelles]

Working on a Maestro FZ-1 build. Working off the layout below using the offboard power supply to get +9 to -3.

It's working and producing sound and fuzz. However there is some nasty HF noise, the attack knob seems to control it. What weird is when the effect is disengaged, the noise is still faintly present, and still controlled by the attack knob.

I have this wired as true bypass with an LED via 3PDT, standard method. Shouldn't this prevent the noise when the effect is disengaged?

I'm stumped as of right now, any suggestions?

[Cozybuilder]

I just looked over the Vero design, it matches up with the 1999 Jack Orman schematic, based on Vinnie Zumbo's 1962 pedal. So, the design is not suspect, this leaves the build. After verifying the component values and electro cap polarity connections, check for solder bridges and cold solder joints. Did you connect Attack 3  and Volume 3 to ground?

The voltage inverter vero looks OK, did you verify that the output is about -3? If not, then just use a couple of AA's to power your circuit (connect pos to ground), and try it. Still not working?

Please post clear photos from the front and back of your build, maybe one of us can spot something.

[andrewnelles]

Thanks for the reply.

Leg 3 on both pots is wired to ground. The problem seems to be originating on the power converter board. It is outputting a correct value, measuring to -3.133VDC.

They problem occurs with either a wallwart or a 9V battery. I need to dig up a AA holder to try by passing the board.

Here is a link to a video of the issue: http://tinyurl.com/qymqcsb

Here is a photo of the guts, need to clean up that wiring:

[andrewnelles]

Update, the noise is absolutely coming from the -3VDC board.

It's about a 6hkz tone, which I found reference too in other threads about L7660S chips. As you can see in the vero layout, Pin 1 is connected to VDC which should get the oscillation above audio level. It doesn't seem to be doing that.

I read elsewhere that a MAX1044 chip runs cleaner for this purpose, I put one on order, just in case. We'll see when that arrives.

My other thought, the datasheet seems to indicate the the oscillation is proportional to the input voltage. Since this IC is running after the 3VDC regulator, maybe it would have less noise running before the regulator? More input voltage = higher frequency?

Thoughts?

[GibsonGM]

Don't know how the power board works, but if it's switching, you may be coupling noise into your in/out lines. If you were to replace them with shielded wire or try to re-route them, it might help.  Try 'chopsticking' them around when on, see if the noise decreases or not.  It is very near your switch....


Just a thought.

[Cozybuilder]

If you've measured about -3V from your power board, the problem has to be in the circuit board or connections.

Its a little unclear, but it looks like the red wire on the LED is connected to the power plug ground? That should go to +9V. Where is the load resistor for the LED? You need something to avoid a quick flash and dead LED. Double check your wiring at the switch for the LED. The In/Out looks OK for both effect on and bypassed modes.

Its difficult to read the value on the resistor in the 1K5 spot (Q2-C to -3V), and the one in the 470K spot is obscured by the black wire from ground to ? The components appear to go to the correct holes according to the layout, I assume the 2 resistors standing up wired in series add up to 56K? Hard to read the in and out cap values, but they could be 103 and 332, which would be correct.

Assuming the transistors are plugged into the sockets OK, then the most likely suspect is solder bridges and/or cold joints. Good luck, let us know.

Edit: 2 Posts while I was writing this. Try with the 2AA batteries, hopefully that will confirm both your posts above.

[andrewnelles]

The LED housing has a limiting resistor built in.

Component values have been triple verified. The two resistors in series are 56K, didn't have any 56's on hand. 

Since the problem occurs faintly event with the transistors removed, I'm almost confident the effect board is healthy.

I'll try GibsonGM's advice and check to see if the 3V board is placed poorly.

Anyone familiar with L7660S think running it before the 3V regulator might help?

[R.G.]

It would be far quieter to make -9V, then regulate it to -3V. That way the regulator is reducing any power supply inverter noise.

Any time you have a switching thing - like a charge pump inverter - wire dress and grounding become critical. You want to keep wires with big signal voltage changes or big current changes away from signal wires, and connect grounds to force switching current to flow in wires that are not connected to the signal circuits.

All of these concepts are common threads in the steady stream of posts saying "my pedal makes noise".

[andrewnelles]

That makes sense. I'll build a replacement board with the regular after the inverter. See if that helps out. I've got tons of spares, so it's no problem.

As far as my grounding, is there anything glaringly wrong that you can see?

I relocated all the voltage wires, not even a slight change unfortunately.

[andrewnelles]

Alright, I think I went down the wrong path here.

Removed the -3VDC board and just wired it up to 2 C batteries (didn't have AA holders on hand). Same exact problem. The board doesn't seem to be the culprit.

Earlier I thought I was getting the problem with the transistors removed also, but I can no longer replicate that.

Back to square one.

Thinking a transistor issue? They were sold to me by another build as being matched for this pedal.

I've got these:

Q1: NKT275 53hfe .32ma leakeage
Q2: AC128 82hfe .58ma leakage
Q3: AC125 128hfe .79ma leakage

[LightSoundGeometry]

I just noticed something for the 1st time which I would like to utilize if i could - the ground rail on the vero, he is using it to ground a lot of other joints. for example, I can run my ground from lug 1 to ground rail on vero instead of one of the 6.5mm jacks and so forth?
any exceptions to this rule?

[lowbrow]

It's the chip. Not set up for over clocking so you can hear the converter whine. You need one of these:

http://www.taydaelectronics.com/icl7660cpaz-icl7660-cmos-voltage-converter-ic.html

[andrewnelles]

Ruled the chip out, the problem persists on battery power with the chip removed.

[andrewnelles]

Still troubleshooting this. On -3V battery power now, problem persists. Verified all components again.

Any thoughts?

Many thanks.

[andrewnelles]

Alright, since this is a simple circuit, I made a whole new replacement board. Same exact issue, high freq noise. Even when pedal is bypassed. Symptoms are identical with either the 9v to -3v conversion, or running off AA batteries.

Rerouted all wiring while the pedal was running, no noticeable change heard.

The noise rapidly changes if my finger is anywhere near Q1 or the 10nf cap. Problem persists with different transistors in Q1.

After some advice elsewhere, I built an audio probe.

What I've found.

- The HF noise is slightly present on both the tip of the input and output jack when the pedal is disengaged.
- Q1 - B and C are quiet
- Q2 - B is quiet, C has HF noise.
- Q3 - B and C are both very noisy.

Found a leaky 2n404A in my collection and tried swapping it in. 194hfe .36ma leakage

The following combo got it working, as long as I keep the attack knob at like 9 o'clock, any higher and I have problems...

Updated measurements:

Q1 NKT275 55hfe .32ma
Q2 AC125 124hfe .70ma
Q3 2N404a 194hfe .36ma

Unused: AC128 82hfe .55ma

Has a nice satisfaction fuzz tone with the attack knob way low. If I turn it up past 9 o'clock  I get crazy high freq feedback, or at the very least unflattering high freq crackle. The faint noise with the pedal disengaged remains if the attack knob is up.

So, it's closer, but still far from working. Seems like the transistor pairing is the issue. This has been a more frustrating build than I had imagined.

Any additional thoughts on how I should continue troubleshooting would be greatly appreciated. Buy more transistors? Modify the circuit to get these working?

I can post voltages also if that helps.

Thanks for the help.

[JerS]

Maybe try the old trick of taking a small cap (50-200pf) and tacking it accross the B-C or B-ground of Q1 (or try it on all the transistors) to help eliminate RF.

I am wondering if you are picking up some RF at Q1 and the other two transistors are amplifying it. This would only take a couple of minutes to try.

[andrewnelles]

I'll give the cap trick a try to see if that helps out, sounds like it might.

The guy I got the transistor set from is sending me a second set. I'm hoping that just squares it away.

[andrewnelles]

Didn't have much luck with the cap idea, no significant change.

I picked up a lot of germaniums, and had a handful of leaky ones that were in the ballpark for this circuit. Swapped them into this, same problem.

So it now looks like the transistors are not the issue....

Maybe this circuit just is not stripboard friendly???

I'm open to any other suggestions here.

[R.G.]

See the ground and -3V traces on the effect circuit board? Place (1) a 22uF electrolytic cap (2) 1 0.1uf **ceramic** cap and (3) a 0.01uF **ceramic** cap between those two lines, observing the polarity on the electro.

If that doesn't help, it's time to get out the oscilloscope.

[andrewnelles]

Got a package in the mail today, the buyer I got my original transistor set from set from sent me some replacements. With that set the pedal is just fine.

This is the third mix of transistors I tried in it, and it's golden, as of now at least.

I guess this lesson is that this circuit is incredibly picky about transistor matching...


My only complain is that there is some high freq noise when a note is ringing out. It's quiet otherwise. The noise is a bit annoying. Anyone have a suggestion on a switchable cap to tame this a bit? Maybe a switchable 50pf from B to ground on Q1?