Troubleshooting my Valvecaster - First post!

[rave0035]

Hi all!  Intros first: first post here, as well as my first pedal build.  Not new to electronics - I've been tinkering with instrument and hifi amps for several years and have intermediate skills and gear availability.

What I'm doing: building a Valvecaster pedal for a friend/bandmate entirely from spare parts I have lying around.  I used this layout:


...with a few "mods" based on available parts:
- VR1 is 60k, VR2 and VR3 are 15k, all linear
- Currently testing with a 12v, 300mA wall wort since my buddy pilfered all my 9v's 

What it looks like:






What it's doing: VERY weak signal at output when not bypassed... I'm talking a pure sine wave that usually rips speakers apart at 6 that is barely audible at 100.  Picking up some RF activity, though. 

What I've Done: I'm loosing signal at the grid of the first triode - I can trace a strong signal to that point.  After that, it's dead as a doornail.  I've double checked all the wiring and looked for solder blobs and obvious shorts and found neither.  Heaters glow, and grounds check out.  I pulled pin voltages this morning and then came here:
(All readings referenced to COM, with 12v DC input)
1:-11.6v
2: -0.4v
3: 10.8mV
4: COM
5: -12v
6: -11.9v
7: -369mV
8: COM

Where do I go next, o great pedal gurus?
Mike

[duck_arse]

first off, rave0035, welcome.

second off, rave, you MUST connect that meter for the pedal to function correctly! MUST! any method of connecting to provide indicating will do.

thirdly, all your voltages show as (-), is this correct? is your adaptor perhaps backwarding your volts?

fourthly, vr2 and vr3 are way low value, may load in a bad way.

[rave0035]

Agreed on the meter, DA!  Trouble is, it's a true AC meter with a 0-25 scale, so I'm looking at a significant amplification factor to get signal to make that needle bounce.  First things first - gotta get the rest of the thing working.

The neg readings had me scratching my head, too.  Haven't investigated much yet.  I did try a supply with the opposite plug configuration and couldn't get the heaters to light, but that's as far as I went.

I will report back.
Mike

[tubegeek]

Plate MUST be positive compared to cathode. 9 to 12-ish volts in this design. Grid MUST be negative compared to cathode, 1-to-2 volts in this design. Keep the 12V supply, forget about going back to a 9V - 12V will light the tube properly, 9V will not.

Until you get that straight, nothing's gonna happen. Polarity reversal in your power supply and a ground connection you aren't accounting for sounds like the problem.

Are you sure the tube is good?

Welcome.

I'll take 50 of those meter/enclosure units if you can spare 'em. Damn.

[rave0035]

Lesson of the day: check your DC input polarity with a meter instead of just assuming which pole leads where.

Reversed leads in the PS.  Flipped 'em and everything's hunky dory.  In case anybody is wondering, the pin voltages when functioning are:

1.-5.7v
2. -11.5v
3. .67v
4. COM
5. 12v
6. 3.7v
7.-350mV
8. COM

Thanks for the warm welcome, all.  I've got so many things like this that I need to build into stuff that you guys will likely see more of me this winter.

Anybody got ideas on a signal amplification circuit for that meter?  I believe it's a true AC moving iron meter, but I need to test it a bit to see.

Mike

[duck_arse]

well I hate to say it, rave, but now only half yr volts are negative. and you say it's working? nice one.

squint real close at the meter scale, and see if it has a f.s.d. marked. maybe on the meter back? is there a resistor across the meter terminals? there are ways to work out the full scale deflection value of the meter, then mess with it to get the range you want, but the method, along with many other things, eludes me.

scaling resitors, multplier resistors, meter/current shunt; something like that.

[rave0035]

Yes, she's working!  And sounds pretty grimy with the gain up  .  Still getting a little bit of RF through it on my test bench, but it's not a metal case so I guess that's to be expected.

No external resistor, DA, and the thing is sealed up tighter than a bank vault so there's no easy access to the innards.  Here's what I DO know: It doesn't appear to be a true AC meter.  I have an xfmer on my bench that puts out about 7vac, and connected to the meter it shoots it immediately to the (max) 25vac range.  Current operated?  DC batteries (C cell) don't move the needle an inch.

Ideal state would be to get this needle to jump with signal, similar to a VU meter.  I could care less about accuracy or any of that hooey - purely an cool effect.

Mike

[Ben Lyman]

Cool. Just don't let one of your friends mistake the tube for a bypass switch! 

[PRR]

> doesn't appear to be a true AC meter.

It is an AC current meter.

As-marked, it had a series resistor (now lost) to make it read "0-25 Volts AC".

Try 10K to 1K from your 7VAC source, you may find the sensitivity. If 7VAC plus 7K brings the needle almost to the "7" on 0-25, then it is a 1mA meter. (There are a few hundred ohms inside the meter.)

It is surely too hungry to feed from a tiny tube booster.

An op-amp output thru 1K may twitch visibly on loud chord.

Any of the LM386 mini-amps ought to drive it authoritatively.

True AC movements in small values are rare. (Audio dancing-needles are usually DC movements plus a rectifier.) I fear it may be fine for 25Hz-400Hz power (and low audio) and fall-off bad for faster (treble) frequencies. As eye-candy, that may not be a problem.

It seems odd for a small AC meter to be linear-scale. A permanent-magnet movement swings both ways on AC, average "zero" (or beat-up pivots). AC meters usually feed the AC to *both* armature and field. This fixes the polarity issue, but the response is the *square* of the current. If 0-25V, then "half" comes out 6.25V (not 12.5V), and worse further down. They can compensate some of this by shaping the poles.

No response on DC is also odd. I may have to shovel into old books to figure what that type is.