Having Trouble with the GEO Tube Wah

[Paul Marossy]

I finally decided to build this thing. I etched my PCBs, wired it all up yesterday. It's turned out pretty well so far. However, I get no sound at all - absolutely nothing. I haven't found any wiring mistakes or shorts that I can see. I am getting a B+ voltage of about 100 volts, but I think that will improve as line voltage was low at 4:00AM this morning - only 106V. Anyhow, any ideas on where I should look? I guess I need to get out the audio probe...

Anyone who has built this, please chime in! 

Here's the last discussion I was in on regarding the GEO Tube Wah:
http://www.diystompboxes.com/smfforum/index.php?topic=43203.0

[R.G.]

Actually, "What to do when it doesn't work" still applies.

Measure the voltages to the pins of the tube and the MOSFET, give us a list. Sometimes shorts or opens and wiring mistakes are very difficult to see. But making exactly the same mistake (i.e. missing a wiring problem) when measuring voltages is extremely unlikely.

If the DC voltages are OK, then there is a signal short or open that's getting you.

[Paul Marossy]

R.G.-

Thanks for the reply. I just checked out the voltages. I think the problem is at Pin 1 & 7 on the tube - I get a reading of 0.5V, so that seems to be the obvious problem. What is not obvious is why that is happening. I used a brand new NOS 12AT7 that I had in my parts bin for testing, but I'm going to try another tube and see what happens.

[Paul Marossy]

Here's some pics for the curious. I made this completely out of things I had lying around that I wanted to put to use someday. The first few pics are my AutoCAD PCB layouts compared to the etched boards with my usual pre-population mark-ups. The power supply is housed in an old transceiver enclosure that was given to me a few years ago and the wah part is in my second DIY wah shell. I think I did a pretty good job of cramming all that power supply stuff in the enclosure. I'm going to make a 1/4" plywood base for it. When it's all done, it will be about the size of the Maestro Boomerang wah shell. If I get it working and I like it enough, I may try my original idea of putting it in a CryBaby shell, which was my original intent, but I don't have a spare CryBaby shell lying around at the moment. The main PCB is designed to go in a std. wah shell, and I designed a power supply PCB that can fit inside as well - I had to etch one that was as big as the last piece of unetched PCB material that I had lying around, so that's why it looks a little different than my CAD drawing.

[Paul Marossy]

OK R.G., here's a complete list of voltages for you (using a different tube):

B+ = 96V
Pins 1&7 = 1.2 VDC

NTE2398 (IRF820 "replacement")
S = 0 V
D = 96 V
G = 0 V

Tube heaters = 12.0 VAC

Is there something wrong with the MOSFET? Not sure what all this means... 

[R.G.]

B+ = 96V
Pins 1&7 = 1.2 VDC

NTE2398 (IRF820 "replacement")
S = 0 V
D = 96 V
G = 0 V

Tube heaters = 12.0 VAC

OK, that part is good. Do the heaters actually light up orange?

B+ = 96V
Pins 1&7 = 1.2 VDC

That's at least one problem.

Pin 1 is a plate. The pinout for the 12A?7 series is
1 - plate 1
2 - grid 1
3 - cathode 1
4, 5 - heaters
6 - plate 2
7 - grid 2
8 - cathode 2
9 - heater centertap

What you SHOULD see is about:
1 - 50 to 90V
2 - 0.0V
3 - 0.9 to 1.5V

6 - B+
7 - same as 1
8 - same as 1 plus a volt or so.

On the MOSFET, you should see drain = B+, gate = tube pin 8, source = a volt or two lower than the gate.

The first problem is that your first tube seems to have no B+ getting to it. What is the voltage across the 100K plate load? If it's almost full B+, then your tube may be shorted or there may be a wiring problem. If it's 0V (that is, both ends are at B+) then the resistor is not connected to the plate. Read the resistance with the power off to be sure that it is not internally open.

Verify that the voltage on the end of the 100K grid feedback resistor coming from the inductor is 0V.

It's not passing any audio because the tubes are not running in a linear area of their characteristics.

[Paul Marossy]

Do the heaters actually light up orange?

Yep, they sure do.

What you SHOULD see is about:
1 - 50 to 90V
2 - 0.0V
3 - 0.9 to 1.5V

6 - B+
7 - same as 1
8 - same as 1 plus a volt or so.

Sorry, I forgot to mention the voltages on the pins of the tube. Everything is where it should be except for Pins 1 & 7.

On the MOSFET, you should see drain = B+, gate = tube pin 8, source = a volt or two lower than the gate.

Something's definitely not right with the MOSFET. I can't see why it's not working unless it's a defective part.

The first problem is that your first tube seems to have no B+ getting to it.

Yep. I can't see where I've made any mistakes, though.

What is the voltage across the 100K plate load?

At B+ side, 96V, at plate 1 side, it's 0.5V

If it's 0V (that is, both ends are at B+) then the resistor is not connected to the plate. Read the resistance with the power off to be sure that it is not internally open.

But it is connected to Pins 1 & 7, that's what I don't understand why it's not getting B+! It does measure 100K with the power off. Already checked that.

Verify that the voltage on the end of the 100K grid feedback resistor coming from the inductor is 0V.

I'll check that.

It's not passing any audio because the tubes are not running in a linear area of their characteristics.

Yeah, the tube is basically off!

[R.G.]

...pins of the tube. Everything is where it should be except for Pins 1 & 7.
...
At B+ side, 96V, at plate 1 side, it's 0.5V
...
But it is connected to Pins 1 & 7, that's what I don't understand why it's not getting B+! It does measure 100K with the power off. Already checked that.

Just for ease in checking, let's lessen the mystery. Disconnect the path into the MOSFET gate through the 220 ohm resistor. This does two things. First it lets us concentrate on the tube, which is the least familiar part of the circuit, and two, if the MOSFET is the issue, it tells us that.

With the gate stopper resistor opened, what are your tube voltages? If they suddenly pop up to normal, then we can concentrate on the MOSFET.

What makes me suspect the MOSFET holding things down is that a 12AX7 can't "saturate" its plate-cathode voltage to less than about 50V. I've tried. So it's not the tube that's pulling its plate down to under a volt, it's something else. That's why I needed to know if the plate resistor was open. If it was not, something in the second tube section was pulling things down. If the MOSFET is shorted, if its gate protection diode is shorted, if there's bad wiring, solder, etc. that holds down the cathode of the second tube, then its grid is pulled positive by the first tube plate and goes into grid conduction. The grid CAN pull down the plate of the first tube, so that might explain things.

Cutting the MOSFET loose lets things get back to normal. If the voltages don't pop up to normal, it's an issue with the second tube section.

[Paul Marossy]

It's working! After my last post, I checked for things being grounded that weren't supposed to be and I found that pin 8 had continuity with ground. Apparently, something was shorting out on the PCB while it was in the enclosure. Now I get 117 VDC for B+ and right at 50 VDC on pins 1 & 7. MOSFET is working fine, too.

Now I have a new problem: a HUMMMMMMMING BUZZZZZZZZZZZZZ 

I know that's one of a few things - the heater supply, lead dress, the inductor picking up EMI from the transformers or not enough filtering on the B+ supply. Moving the inductor around in relation to the power supply does not have any affect on the hum/buzz. I know that using shielded cables on certain key wires can help with that, but this is far beyond the kind of hum that you can pick up over a wire in the signal path. I am really wondering about the B+ filtering. The schematic I have indicates a pair of 22uF filter caps? Is that correct? I remember the last time I looked at the schematic, I thought that seemed like really small filter caps compared to other tube circuits I have built... 

[R.G.]

Pin 8 short to ground - yep, that would do the same thing as a MOSFET sucking that cathode to ground. Congratulations!

Now - welcome to the wonderful world of tube amp hum. 

First - are your heater supplied referenced to ground? If it is not, you'll get massive hum.

Does your 12Vac heater supply have a center tap? If so, ground that. If not, hook two 220 ohm/1% 1W resistors in series across it and ground the center of the resistors.

22uF is quite small if you're used to solid state power supplies, but in this unit, the plates are pulling around 1ma each, so with only a few milliamperes of current, the 22uf filters should be good enough.

The exact details of grounding matter, too. We'll get to that.

[Paul Marossy]

Pin 8 short to ground - yep, that would do the same thing as a MOSFET sucking that cathode to ground. Congratulations!

Thanks! 

Now - welcome to the wonderful world of tube amp hum.

Yep, been there before. It's a pain in the rear. 

First - are your heater supplied referenced to ground? If it is not, you'll get massive hum.

No, but the heaters being referenced to ground did also cross my mind last night but I forgot to mention it. I kind of forgot about that - it's been a while since I built my last tube amp. 

Does your 12Vac heater supply have a center tap? If so, ground that. If not, hook two 220 ohm/1% 1W resistors in series across it and ground the center of the resistors.

As a matter of fact, those transformer secondaries are center tapped, so I will try that first. That might take care of it right there. 

22uF is quite small if you're used to solid state power supplies, but in this unit, the plates are pulling around 1ma each, so with only a few milliamperes of current, the 22uf filters should be good enough.

OK, sounds good. I'm just used to looking at tube amp schematics, and there is quite a bit of filtering going on there.

The exact details of grounding matter, too. We'll get to that.

Yeah, I know. I'm just going on my mostly successful past experiences with the three tube amps that I have built and the few tube based stompbox pedals I have built. 

[R.G.]

As to the hum/buzz: is it 60 cycle or 120 cycle buzz?

60 Hz is pickup of line frequency, 120Hz comes from inadequate B+ filtering or rectifier buzz.

[Paul Marossy]

It sounds like a little bit of both. That's one reason why I was wondering about those 22uF caps. I'll mess with it some more tonight when I get home from work and see what I can narrow down.

[Paul Marossy]

OK, here's an update. I tried connecting the secondary center taps to ground, and it helped, but only a tiny bit. I still think it's the power supply filtering. I see that on the schematic 1N4007 rectifier diodes are indicated. I used 1N4004s in my build. Would that make any difference? Does a 1N4007 diode switch faster than a 1N4004 does?

[R.G.]

I see that on the schematic 1N4007 rectifier diodes are indicated. I used 1N4004s in my build. Would that make any difference? Does a 1N4007 diode switch faster than a 1N4004 does?

Data sheets are our friends. The 1N400x series of diodes are all 1A general rectifier diodes, and almost always the 1N4001 through 1N4007 are listed on the same datasheet. They are not particularly fast, being designed for 60Hz power line rectification. There is almost no difference between them except for the reverse voltage specification. The 1N4004 is rated at 400V, so it's good enough for the application. I buy 1N4007's (1000V) in 100 unit lots as I need them and use them for everything that's power rectification or protection under 1KV.

It's not the diodes, unless something's wrong with one.

Just for comparison, the Vox AC30 used an 8uF first filter cap, then 20uF for succeeding stages.

[Paul Marossy]

The 1N400x series of diodes are all 1A general rectifier diodes, and almost always the 1N4001 through 1N4007 are listed on the same datasheet. They are not particularly fast, being designed for 60Hz power line rectification. There is almost no difference between them except for the reverse voltage specification.

I didn't think there was any difference between them except for the voltage ratings. I've looked at the data sheet before and never noticed anything different other than the voltage ratings. But, I thought I'd ask the question anyways.

The 1N4004 is rated at 400V, so it's good enough for the application.

That's why I used them. And also because that's all I had four of at the time. 

Just for comparison, the Vox AC30 used an 8uF first filter cap, then 20uF for succeeding stages.

Wow! That seems tiny. I often have a hard time telling the difference between 60Hz hum and 120Hz buzz. It sounds more like a buzz to my ears. I'll try some more things to try and figure out where it's coming from. I think I'm going to do the battery powered heaters for testing trick next.

EDIT: Here's an idea - if you run a longish B+ wire in bundled in parallel with a twisted pair of 12.6 VAC heater wires, can that inject hum into the B+ supply?

[A.S.P.]

what a difference 20 Henries between 2 caps make... 
(even @ 50 times the current consumption)

[A.S.P.]

btw: what does an AC filament at ground-niveau do to a grid at ~100V niveau,
which is circuited in a (high-impedance) cathode-follower?

(search: "lifted filament ground", or similar)

[Paul Marossy]

So, are you suggesting that even though the secondaries are center tapped that the heater supply may need to be lifted above ground?  A pair of 220 ohm resistors could take of that. I'm still going to do the battery test when I get home tonight.

[Paul Marossy]

OK, well I tried running the tube heaters on batteries - still got the buzzing noise. I think I'm going to try bigger filter caps next because I don't think that it's the inductor picking up EMI. I've heard that before and it doesn't sound the same as this...

It's almost finished except for getting rid of the buzzing noise. If I get it working right and I like it, then I will probably re-house it in a CryBaby shell like I originally planned. Or it might stay like this and be a studio fixture so to speak. Don't know right now. Anyhow, here's a few more pics: