King Nothing - Oscillating

[drew.spriggs]

Howdy,

Built a King Nothing. Everything works fine, voltages are spot on (254v on the plate resistors, and 29v heading into the J201), but there's significant amounts of oscillation that decreases as the gain is turned up. Low switch has no impact on the noise, but the left position of the mode decreases the range where it oscillates from about 1-9 to 1-3 (assuming each tick is about 30 degrees of rotation).

I've tried multiple power supplies, from isolated DC bricks, to cheap SMPS to an old unregulated power supply I've got around (all capable of +500ma), and no difference. Tried multiple guitars, multiple leads, etc and there is no difference.

Tube is a brand new Shuguang 7021 (confirmed to be good), J201 is a onSemi MMBFJ201 on a SOT23>T092 adapter board.

Thoughts?

 

[drew.spriggs]

Ok, tried a few things:

1. Decreased R5 (plate resistor on second stage) from 220k to 100k. No change
2. Tried an old EHX 12AX7B I had spare from my amp. No change.
3. Tested a 100pf cap in parallel with R13/R14 for RF suppression coming in, as well as across D/G of the J201 (in case it's oscillating there). No change.

I've also noticed that when it's oscillating, it is MORE noticeable in bypass (didn't notice that originally).

I've confirmed every component is correct, and there are no sketchy solder joints.

Semiconductors:

1. TE NE555P
2. Central 2N3904
3. Vishay IRFB17N50 (have used on previous builds and works perfectly).

 

[Smrtokvitek]

By oscillation, do you mean high-pitch squeak overtaking the overall sound or just a background "noise"?
It would be nice to share a picture of your build.

 

[drew.spriggs]

By oscillation, do you mean high-pitch squeak overtaking the overall sound or just a background "noise"?
It would be nice to share a picture of your build.

Squeal significantly louder than signal levels

 

[drew.spriggs]

Ok, have some spare time so decided to poke around with this.

I've now changed every active and the tube, and it's EXACTLY the same. I also put the board through a IPA wash in an ultrasonic cleaner so it is as clean as it could every possibly be with no changes.

I'm pretty much calling this a wash - I have seen at least one other post with people having the same issue that was never solved.

 

[drew.spriggs]

After somebody else pointed out the grid stopper for the first triode (which should be as close to the grid as is possible) is not only miles away, but has traces on both sides that run right through the middle of the power supply on both sides of the board, I thought I'd relocate it as a final hail Mary.

I removed R9, ran a shielded wire offboard, then straight onto the daughterboard. I cut the trace from the daughterboard and installed the 470k directly to the socket of the grid pin.

It DID give me a pretty substantial reduction in noise floor (so much so that I really, really hope it gets changed on any new boards as it was a massive difference), but the oscillation is still there - albeit substantially reduced from where it was originally.

For shits and giggles I wrapped the tube in alfoil and grounded it directly to the power socket, which reduces the oscillation even further - but still, it's on par with the signal coming through any time the tone control isn't all the way down.

She's in harvest for parts pile as I've wasted enough time on this.

 

[TucsonSean]

Quick question, did you ensure that your inductor is 100uH (101 marking is typical)? Your pics show 104 (100mH). Big difference if that is the case and maybe a source of your oscillation?

 

[Creach]

After somebody else pointed out the grid stopper for the first triode (which should be as close to the grid as is possible) is not only miles away, but has traces on both sides that run right through the middle of the power supply on both sides of the board, I thought I'd relocate it as a final hail Mary.

I removed R9, ran a shielded wire offboard, then straight onto the daughterboard. I cut the trace from the daughterboard and installed the 470k directly to the socket of the grid pin.

It DID give me a pretty substantial reduction in noise floor (so much so that I really, really hope it gets changed on any new boards as it was a massive difference), but the oscillation is still there - albeit substantially reduced from where it was originally.

For shits and giggles I wrapped the tube in alfoil and grounded it directly to the power socket, which reduces the oscillation even further - but still, it's on par with the signal coming through any time the tone control isn't all the way down.

She's in harvest for parts pile as I've wasted enough time on this.

Hello drew.spriggs,

I'm a little bit late to see your posts.

I also have a King Nothing so if you want to have your unit to work i can try to help you.
A squeal should be easy to spot even if you just have a multimeter, make it squeal and take AC measurements after c3, c5, c6. It's the first step.

 

[drew.spriggs]

Quick question, did you ensure that your inductor is 100uH (101 marking is typical)? Your pics show 104 (100mH). Big difference if that is the case and maybe a source of your oscillation?

Ok, things are....very weird. I ordered 100uh inductors from Mouser. This is the same listing from every single distributor out there, as well as the Coilcraft website. They even have 1mh inductors available with a code one multiplier designator higher (which you'd expect if they were labeled correctly).



Datasheet shows it as 100uh


However....



...its code is 104. Every single other 100uh Coilcraft inductor listed also has a code of 104. This appears to be the ONLY manufacturer out there that does this.

Am I missing something obvious?

 

[drew.spriggs]

Ok, so Coilcraft clearly just use whatever component code they want - I have a couple of spares, and they all measure at 110uh.

 

[TucsonSean]

No, your not missing anything obvious. In these types of cases, you always double check values to make sure that you didn't install the wrong value component on the board.

I also checked to see if Coilcraft is the only manufacturer to label 100uH as 104. MuRata also uses 104 to denote 100uH. Always good to check your sanity.

A couple of questions for you:
Does the pedal oscillate even when nothing is connected to the input and only the output is connected?

Do you have an oscilloscope or at the very minimum a mult-imeter that can measure AC volts? Anything to help trace the source of the oscillation.

Also, I checked the datasheet for the FET that you used, compared to the IRF740, the capacitance values are about twice that of the IRF740. This may be an issue for the 555 timer driving the gate. I would suggest that you place a 100 ohm or larger resistor in between the 555 and the gate of the FET. The other option is to use an IRF740.

-Sean

 

[drew.spriggs]

A couple of questions for you:
Does the pedal oscillate even when nothing is connected to the input and only the output is connected?

Do you have an oscilloscope or at the very minimum a mult-imeter that can measure AC volts? Anything to help trace the source of the oscillation.

Also, I checked the datasheet for the FET that you used, compared to the IRF740, the capacitance values are about twice that of the IRF740. This may be an issue for the 555 timer driving the gate. I would suggest that you place a 100 ohm or larger resistor in between the 555 and the gate of the FET. The other option is to use an IRF740.

-Sean

Pedal oscillates regardless of what's happening on the input - whether it's grounded, floating, or whether the pedal is on/off it oscillates (and is audible on the output). Turning it on only amplifies the oscillation.

Swapped to a 740, no difference (also have used these on a few 555-based PSU's with no problem). Supply is very clean after the cap (with only ~200mv or so of ripple at ~49khz.

Finally found my oscilloscope. Triode 1 is the culprit - even with nothing but some low noise going into it, I'm getting 10v p2p of oscillation out. Multiple tubes give the same output.

 

[drew.spriggs]

Out of interest, I superimposed both sides of the PCB on top of each other, and drew the input/output jack traces (green), the first grid stopper (white) and the output to the 3PDT (orange).

• There is a single shared ground plane for everything - input/output and switching supply. There's about 1A of junk from the inductor dumped into the ground plane that radiates across the entire PCB and joins in with the signal ground.
• The input jack trace goes right next to both plate resistors and the coupling caps to get to the switch PCB.
• The output jack trace goes right underneath the inductor, MOSFET AND switching diode (the origin of all noisy ground currents)
• The trace from the first grid stopper hits most of the high points of the onboard supply including the snubber cap/resistor (which I've moved since), and is miles away from the tube (where it is meant to be).
• The output trace from the volume pot is bang in line with the output jack trace, going right through the guts of the switching supply to the switch.

I've spoken to a number of fairly experienced guys who have probably forgotten more than I'll ever know, and they all said they're shocked that anybody has managed to build one that doesn't oscillate.

 

[Tassieviking]

There are usually 4 copper layers in most of the C2CE PCB's that I have noticed, the 2 internal layers are often flood filled so they act as shields as well as conducting electricity.
It would be impossible to really make a comment on the signals without seeing the how all the layers are made compared with each other, there is only one person I know of who can know the layout of the layers and I personally don't think he should reveal them since it is his copyright material.
You can see which solder pad is connected to a flood filled layer by holding the PCB up to a bright light and look for the typical 4 legs coming of the pad in the layer, but you can not make out if the flood layer is covering the whole PCB or if there are any ruled out areas where there is no flood fill.
Something I was not aware of is that the metallized capacitors we use are actually polarized in a way, and if you install them the wrong way they can pick up a lot of noise and interference.
I am going to check all mine from now on with a scope before I use them.

Where To Connect the Outside Foil on Capacitors

Guitar amplifiers

aikenamps.com

 

[Smrtokvitek]

Out of interest, I superimposed both sides of the PCB on top of each other, and drew the input/output jack traces (green), the first grid stopper (white) and the output to the 3PDT (orange).

View attachment 95005

• There is a single shared ground plane for everything - input/output and switching supply. There's about 1A of junk from the inductor dumped into the ground plane that radiates across the entire PCB and joins in with the signal ground.
• The input jack trace goes right next to both plate resistors and the coupling caps to get to the switch PCB.
• The output jack trace goes right underneath the inductor, MOSFET AND switching diode (the origin of all noisy ground currents)
• The trace from the first grid stopper hits most of the high points of the onboard supply including the snubber cap/resistor (which I've moved since), and is miles away from the tube (where it is meant to be).
• The output trace from the volume pot is bang in line with the output jack trace, going right through the guts of the switching supply to the switch.

I've spoken to a number of fairly experienced guys who have probably forgotten more than I'll ever know, and they all said they're shocked that anybody has managed to build one that doesn't oscillate.

I have built it successfully some time ago, but you are right about the signal routing. Not ideal, but that is the cost of having to put it in the 125B-sized enclosure with the footswitch and jacks in places where you'd expect them. The 4-layer PCB helps a lot here.

In the last few years, I have laid out several tube-based projects (more or less successfully) and built a few from other designers. Tube circuits have this big disadvantage of signals easily crosstalking each other. The layout has to be spaced out or very clever, and from my experience, it is quite difficult to put everything on a small PCB (in one flat plane). To be fair, even the Kingsley pedals use shielded wires in many places inside their pedals because there is not enough space inside the small box, even though you can use the space in all dimensions when handwiring it.

 

[drew.spriggs]

It would be impossible to really make a comment on the signals without seeing the how all the layers are made compared with each other, there is only one person I know of who can know the layout of the layers and I personally don't think he should reveal them since it is his copyright material.

You can literally see where the traces go on both the PCB and the build docs. There is no confusion about where it's going, and why it probably shouldn't be.

 

[Tassieviking]

There are hidden traces on the internal layers you can't see, and you can not see the amount of shielding the internal layers provide between the front and the back.
On this design you can see most of the traces but not all that are there, I think there is a full layer of copper internally that is at 0V so it acts like a ground plane.
Once the ground plane is put down it changes everything.

 

[drew.spriggs]

There are hidden traces on the internal layers you can't see, and you can not see the amount of shielding the internal layers provide between the front and the back.
On this design you can see most of the traces but not all that are there, I think there is a full layer of copper internally that is at 0V so it acts like a ground plane.
Once the ground plane is put down it changes everything.

My dude, you can literally follow the traces, through the vias, on both the PCB as well as the build docs. There is zero confusion about where any of it goes for the couple of traces I mentioned as the entire path is visible with both sides of the PCB visible on an image.

Either way, I fixed it. I just removed the bypass cap off the first triode, and increased Rk to 1k8. I've tested this with multiple tubes and it's dead quiet, as well as going to back to 1k5 to test it's not a fluke.

Nope.

1k5 for Rk instantly makes every tube I tested oscillate.

It does halve the voltage gain for the first triode, but at least it's usable now.

I've also confirmed the oscillation coupling in bypass is due to the trace for the input jack which wraps around the coupling caps/voltage divider after the first triode (where all the noise is coming from). I did initially run these offboard straight to the switch PCB which eliminated any oscillation in bypass, but now it's not oscillating I went back to the PCB trace to the switch PCB just to neaten the offboard wiring up.