Tube Driver Clone - Hum

[pedalbob]

 Hi All-
I built a Tube Driver clone using the pcb's and circuit from here:

https://guitar-electronics.eu/en_US/p/TUBE-DRIVER-PCB-SET/310

Sounds great (using EHX 12AX7) but there is a 60 cycle type hum when I hold the guitar and have effect on. If i touch the enclosure or one of the pot shafts the noise goes away. If I put the guitar down with effect on the noise goes away. The circuit used 12V AC input and rectifies it with caps and diodes rather than using 12V dc input.
I tried two different power supplies with no change. I tried grounding the board directly to the enclosure with no effect.
When the effect is off but powered there is a low frequency hum but not the higher 60 cycle type hum.

I haven't tried a different tube yet because seams more like a grounding issue. 
Anyone else ever built one of these or had similar issue with AC powered effect  ?

Thanks for any suggestions.

[GibsonGM]

Does the guitar do this with ANY other amp/effects?

[pedalbob]

Hi-
No just this effect.  It is quiet with other effects I have ... battery powered, DC output wall wart and older MXR/DOD with direct 110v AC input etc.
Thanks   

[GibsonGM]

I'd have a serious look at the grounds, the continuity from output jack thru circuit and back to input/enclosure grounding, just to rule them out.

Can you post a pic of your build, as well as the schematic you worked from?  Nice pics of the diode bridge or whatever it is (showing diode orientation), w/filter caps please. Need to know where DC takes off and AC comes in, and how ground is established.  The 'hum' from power rectification worries me a bit.  Didn't get into the docs - is this the way it is designed or a mod you did?

Being sure the outlet is wired right is 'a thing' (ok, that's grasping, but hey you never know and these things only take a minute to do).   Let's get basics out of the way so there is no confusion lol.

[mozz]

60hz or 120hz?

[pedalbob]

Here are pics of the build.  A bit hard to show rectifier section because boards mount vertically.  No mods to build instructions except I top mounted the power jack instead of on side ... wondering if this is issue now because it results in much longer wire run which passes many component. Will probably remove the jack and pull wires around to outside enclosure to see if any effect on the bypassed hum.

[pedalbob]

Here are pictures of the build doc layout & schematic ... better copies at link in first message

[mzy12]

Sounds like a heater supply issue, if I had to guess. If not, there are plenty of other things that the board and layout get very wrong (not your fault).

First, there is no centre tap, artificial or otherwise, for the heater supply. Bad move. It's also got a 10R resistor on only one side of the heater supply, basically guaranteeing a mismatch in the voltage/current consumption. Double bad move. Secondly, the heater supply wires going to pins 4 and 5 of the tube need to be given proper attention regarding the positioning and twisting. Those are the pads named HT and GND next to the pins 4 and 5 on the tube. Make sure they're twisted together like the wires in an ethernet cable, zip tie them and move them away from any other wires carry an audio signal or power. The fact that a heater supply is named HT gives away the fact that maybe the person designing the board doesn't necessarily understand how tubes work. I don't mean to be a negative nancy, but this is far from the only issue I see.

The artificial centre tap is a little bit more difficult. Check valve wizard's website on heater supply. https://www.valvewizard.co.uk/heater.html
I don't think you can do it because the negative rail is derived from a common ground point, among other things, which means you can't change the side of the AC that's referenced to ground. Ah well.

Next is the problem with the DC power supply. It seems to just be floating, with the ground not being referenced to the chassis at any point. This is less than ideal. I think this is because they don't want either side of the AC power to be referenced to chassis, seeing as one of the the AC wires is going directly to ground, and there is no direct Earth pin from a power cable to protect you in a pedal. On the other hand, it's 12VAC, which you'd have to actively try to make it harm you in any way. My recommendation is another hole for something like an M3 machine screw, a grounding tag of an appropriate size, then a nut and a lock washer. Then ground one and only one point of the board to the new ground tag.

Then there's the fact that there's no bypass cap in close proximity to the positive or negative rails for the opamp. *sigh* Get a 100nF ceramic cap on both of those rails to ground. It won't be neat and it's definitely a bit of bodge, but it's important. Oh, and another 100nF cap from right after each rectifier diode to ground also won't hurt. (Make sure it's on the DC side.)

Overall, the board design is a bit of a mess. Not your fault. Just very annoying.

[R.G.]

mzy12 has some good insights - look at the power supply. There are design problems.

Focusing on one additional issue: it has to use a 12Vac supply because it uses the AC to make a +18-ish V and a -1-ish Vdc supply. The tubes run from +18 to -18V to get a total of 36-ish Vdc from cathode to plate supply. The cathodes run at -18Vdc, which is the result of half-wave rectifying the 12Vac. The grids are referenced to the -18V, not to the signal "ground" that the opamps are referenced to. This difference in reference will make hum all on its own. Capacitors C3, 4, 5, 6, 7, and 9 are an attempt to hold ground and -Vpp together.

These are issues that are built in, not necessarily implementation.

[pedalbob]

Thanks mzt12 and R. G. for all the input !  I guess I have a bunch of work ahead to get this thing to maybe stop humming.  I've built a lot of pedals but this was my first tube build ... thought I bought the board from a reputable source but alas .  Obviously have some time invested in this with all of the offboard wiring etc. and I had to by a not very common sized Hammond enclosure to get the boards to fit ($$)... so I'd like to get it working better if possible.
I'll try what you recommended and report back... it may be awhile    

[mzy12]

The cathodes run at -18Vdc, which is the result of half-wave rectifying the 12Vac

Off-topic, but the only calculations I can do on half-wave rectifying 12VAC gives me  ± 5.5VDC. I'm not really sure what I'm missing here to get that to ±18VDC.

[R.G.]

"12Vac" usually means "12Vac rms". A sine wave, like AC mains power lines try to be, with an RMS value of 12Vac has a peak voltage of 12*1.414 = 16.968V. This gives a peak voltage of one diode drop less, about 16.3V. Rectifying that into a capacitor input filter lets the capacitor charge up to the peak of the incoming waveform, minus that pesky diode drop.

You were missing that peak-to-RMS factor of the square root of two.

I mentioned 18-ish because line frequency transformers typically have greater than their rated output voltage to allow for their own internal resistances causing sag. So it would be common for a small transformer rated, for instance, 12Vac at 200ma to be wound a volt or two higher  with no load. The higher-than-12V output only sags down to 12V at full rated load.

[mzy12]

Ah, thanks. I was doing the math of 16.98 * 1/π. I knew I was wrong, but I wasn't sure how!

[PRR]

calculations I can do on half-wave rectifying 12VAC gives me  ± 5.5VDC.

Note the difference UN-filtered half-wave, and CAPacitor filtered. Naked half wave makes a lumpy wave with average something less than 6V; lightly-loaded capacitor smoothing approaches the 17V peak.

[Clint Eastwood]

Looking at the build doc, the input and output jacks are not grounded to the enclosure. Are you using the same kind of jack sockets?

[GibsonGM]

I'd have a serious look at the grounds, the continuity from output jack thru circuit and back to input/enclosure grounding, just to rule them out.

[mzy12]

Note the difference UN-filtered half-wave, and CAPacitor filtered. Naked half wave makes a lumpy wave with average something less than 6V; lightly-loaded capacitor smoothing approaches the 17V peak.

I was following my college notes on rectification and had a brain fart haha! Thanks for clarifying. 

[pedalbob]

Looking at the build doc, the input and output jacks are not grounded to the enclosure. Are you using the same kind of jack sockets?

Yes I used the same style jacks they had in their build layout (Neutrik enclosed jack with the weird male threaded nut insert with plastic insulating washer) 

[mzy12]

That's also why I said that there was no DC ground reference to chassis. Which is defo part of your problem. The entire thing is basically unshielded!

[PRR]

Alligator-clip leads are essential audio debugging tools.

https://www.amazon.com/Alligator-Leads-ELEGOO-Double-end-Jumper/dp/B0739YWFQ6
https://www.amazon.com/WGGE-WG-026-Pieces-Colors-Alligator/dp/B06XX25HFX