Getting brand new effects running

The hard thing about new effects is that since they've never worked, there are often several mistakes or flaws, and worse yet, the more of a beginner you are, the more flaws you're likely to have, and consequently the more debugging skill this may need. It's a frustrating circle. At least with an older effect you're fixing, it's likely to only have one major flaw, since it used to work.

If you follow this set of steps, you will be able to filter out almost all of the bugs you can run into. If it seems a lot of work, remember that you're making up in doing individual steps what you may not have in experience. As your experience grows, you'll start knowing what to do to skip steps that your new-found experience tells you is not necessary. Remember that you may have to fix more than one problem. Just do it in order, getting one problem fixed and going to the next. At the end, it will work.

Pick out your symptom:

Works fine before mounting in box, then dies.

Something about putting it into the box is interrupting the signal, or a causing a short that radically shifts the bias. Check carefully for ANY place metal on the effects board or a wire end touches the box.
Incorrectly wired jacks. If you get the wires reversed on the input or output jacks, it may sound fine before you put it into the box because the grounds on the jacks may not be connected to each other. When you put them in the (presumably) metallic box, the ground on the jacks are shorted together, and the signal is shorted. The fix - check your wiring and the jack lugs again!

Doesn't work at all, or something is clearly wrong with the effected signal

Do this in order.

Before applying power:
1. Print a throwaway version of the schematic you used!!! You'll use this as a working tool to getting it running.
2. Verify the orientation of every part that is polarized. That is all:
  - diodes - make sure the band (which is the cathode or (-) terminal is pointed the right way
  - transistors - verify correct orientation of the body, and also of where each lead goes; find the pinout for the specific devices you are using, as pinouts for otherwise identical devices can vary. The web has data sheets with this information for most devices. (and I'll put some URL's here). Do not guess about which pin is which. Doing the research for the pinouts FIRST as you're building will save you a lot of trouble later. Verify that the collector, base, emitter, gate, drain, source, etc. go into the correct holes and/or are wired to the right other parts. If you've use any substituted part that's not the identical part number specified in the original schematics, this is particularly important, even (and especially) if it's an ECG or NTE replacement.
  - IC's - get the orientation notch placed the right way
  - polarized (+ and - marked) electrolytic capacitors.
  - Put a check mark on your schematic to indicate it's been checked
3. Using a highlighter or colored pen, visually trace over and verify the entire circuit. Start at the input, and *mark* on the schematic when you verify that there is a part in the right place, and that the value is correct. Do not assume you put the parts in the right place. Read the values, or if it's a resistor and hard to read, measure it. Do this for the entire circuit, including any controls wiring.
4. Visually inspect your soldering. All solder joints must be smooth, shiny, and you should be able to see the shapes of the wires under the solder surface. If the blob is too big, or there are rough places, cavities, discolorations, etc. Remelt and rework that joint until it is correct - warning!!! if you have germanium transistors in there, you have to take as much care not to overheat them when reworking as you did when making the joint first. These things can be ruined by soldering heat.
Get Power and Ground right:
1. Using a voltmeter, measure your battery or power supply and verify that it is indeed putting out the right voltage. One dead battery can eat a whole day of debugging! Connect the (-) or black lead of the voltmeter to the (-) battery terminal and the red or (+) lead to the positive one. Expect a battery to show a little more than the "nominal" voltage when really fresh, a little less when it's partially used, and a lot less when discharged. For instance, a fresh "9V" battery is often 9.5 to 9.9V; a moderately used one is 8.5 to 9.5. If it's under about 7.5V, your results can be affected badly by the low voltage it puts out.
2. Connect the battery or power source to the effect board and verify - is the correct lead (positive or negative) connected to ground?
3. Connect the power source to the board. Hook your voltmeter (-) or black lead to ground, and start measuring power supply voltages.
4. First, measure the voltage where the battery wires come onto the board. If the battery measured good before you connected it, and it now measures a very low or zero voltage, you know that something on the board is pulling it down. You may have a short in your wiring or a bad component.
5. Go over the schematic and find every single component that is supposed to connect to the power lead. Measure the voltage on that component lead, and verify that they're all the same as the power supply voltage. Yes, every one. I use a yellow or green highlighter to mark on the schematic where I've been and not lose track.
6. Remove the power source or battery. Switch your meter to ohms, and go back over the schematic and board, verifying that every single place that is supposed to connect to ground actually does. Mark the schematic as before. If some points don't show a dead short (0 ohms) to ground and they're supposed to be connected, go back over the wiring and make sure.
Connect power to the circuit. One by one, measure the voltage to ground (that is, hook the black/- lead of your voltmeter to ground and probe with the red/+ lead) for every transistor and IC. Write this down on your "throwaway" schematic as you go.
There are two major uses for all devices in effects circuits - linear and switching. Linear devices amplify and process signals. Switching operation is sometimes used to re-route signals, or to operated a Low Frequency Oscillator for modulation or something similar. If you can't tell from the schematic which devices are used for amplifying and which are not, ask someone. If you got your effect layout or board from GEO, it will have a description of how the circuit operates so you can tell.
If you have:
- NPN transistors
  - For linear amplifying, the collector must be more positive than the base; the base must be more positive than the emitter by about 0.4 to 0.7V for silicon, and 0.0 to 0.3 for germanium; the emitter should be the most negative pin. The voltage difference between the collector and emitter is the biggest voltage that the transistor can swing linearly. If the collector is not a volt or more above the emitter, or the base is not a base emitter drop above the emitter as noted above, the transistor cannot be acting as an amplifier. This is a very good pointer to what can be wrong. See the debugging example below. If the collector is at the same voltage as the base, or even closer to the emitter than the base is, the transistor is saturated and simply can't be amplifying. Likewise, if the base-emitter is below the cutoff voltage for the transistor, no current can be flowing.
  - For switching, the base- emitter may be held at 0V or even reverse biased (base lower than emittter) in one state; this condition should turn the collector Off, and it will float up to whatever voltage other parts pull it up to. The base- emitter may also be strongly positive, 0.6 to 0.7V, and the collector will be at the same voltage as the base or lower. The transistor is saturated, and cannot be amplifying signal
  - Debugging example: A fellow had built a replica of the Fender Blender circuit. It was not working, and I asked him to list the voltages on the transistor pins. He found:
    
    Q1 E-0.03 B-0.56 C-2.69
    Q2 E-2.10 B-2.67 C-3.93
    Q3 E-1.89 B-2.53 C-1.96
    Q4 E-0.00 B-0.18 C-3.80
    Q5 E-0.00 B-0.21 C-3.70
    
    Here's what the voltages say about each transistor.
    
    For an NPN silicon transistor (these all are) to be working as a linear amplifier, it MUST have its base higher than its emitter by about one silicon diode drop, 0.5 to 0.7V, and the collector must be higher than either base or collector. The collector-to-emitter voltage is the size of the negative half signal swing, and the power-to-collector voltage is the size of the positive half signal swing.
    
    >Q1 E-0.15 B-0.70 C-3.22
    Base is 0.55 higher than emitter, collector a few volts higher than either one, and room to swing up and down on the collector - this one looks OK. We could tell how much current is flowing because the emitter to ground voltage must flow through the 15K emitter resistor, so the current is 0.15V/15K or 10uA. Looks all right.
    
    >Q2 E-2.66 B-3.24 C-7.75
    This one has its collector tied to the + supply, so its collector is equal to the battery voltage (and your second battery is getting near end of life, too ;-).
    The base is 3.24-2.66= 0.58V more positive than the emitter, the collector to emitter is 7.75-2.66V = 5.09V so there's room for a signal to swing. It can swing -2.66V before it clips, and more than that positive, so this one is OK, too.
    
    >Q3 E-3.95 B-4.63 C-3.99
    Hmmmm.... base is 0.68V higher than the emitter, within the range of operation, but higher than its brothers under similar conditions. That indicates a lot of current through the base-emitter. Further, we see that the collector is only 40mv higher than the emitter - not good! This one is probably saturated, so no signal will come through. I'd bet that either the R11 150K resistor is open (bad solder joint, open PCB trace, broken or wrong value resistor) or there is DC coming in through what should be an open circuit at C5 0.1uF - either a solder short, shorted capacitor or some other such condition. Temporarily unsolder one of C5's leads and bend it up out of the board. If this fixes it, at least c5 is bad. If not, you have a problem with either R11 or a bad transistor.
    
    >Q4 E-0.00 B-0.27 C-7.73
    Ack! Another problem. Base is not at least 0.5V higher than the emitter. That means no current flows, and we see that this is true, because the emitter and collector are at ground and power supply respectively , indicating no current flow through the resistors. This one's not getting enough bias on its base. I'd guess that one of the diodes or associated resistors (R15, R16) has a solder short to ground as a first try.
    
    >Q5 E-0.00 B-0.44 C-7.29
    And another one. Base only 0.44V, no voltage drop across the emitter or collector resistors, so no current is flowing. Something is not letting enough bias get to the base. There is a solder short/wrong value resistor/something else with R28/R26/C13
- PNP transistors
  - The debugging procedure for PNP's is exactly the same as for NPN's, but with the voltage directions reversed. A PNP needs its base biased a bit negative from its emitter to be a linear amplifier where the NPN needed positive. In both cases, what matters is that there is a bit of current flowing through this junction. To have some room to swing (hey, we all need that!) the collector must be at least a volt or two lower than its emitter (the NPN's collector needed to be more positive). Like the NPN, if the collector is at the same voltage as the base or even closer to the emitter than the base, the transistor is saturated, and no audio signal amplification can happen.
- N-channel JFETs
  - JFETs are weird. They're a bar of conductive silicon (I don't know of any available JFETs in germanium) that has a gate junction. The source and drain are at opposite ends of the bar of silicon. Where bipolar (NPN and PNP) transistors normally don't conduct until you turn them on, JFETs normally conduct until you turn them off. An N-channel JFET with its gate at the same voltage as its source will happily conduct all the current it can. To turn it off, you have to reverse bias the gate to a voltage lower than the source. If you keep making the gate more negative than the gate, the JFET will eventually cut off entirely.
  - For an N-channel JFET to be operating as a linear amplifier, the drain (kind of like a collector) must be at a higher voltage than the source pin. The gate must be at a lower voltage than the source to have any control of the current. A JFET with a gate voltage equal to or higher than the source pin is saturated. A JFET with its gate less than the source voltage may be amplifying or may be cut off. Best to debug this with the Audio Probe.
  - JFETs are sometimes used as variable resistors. In this case, the gate will be held in the range of operation voltages as though it were a linear amplifier; that is, 0 to minus a few volts.
  - There is no good way to say how much the gate of a JFET must be negative from the source to be linear or cut off. It varies a lot from JFET type to JFET type, and even as much as four to one in the same type number.
- P-channel JFETs
  - Again, weird. Just like N-channels, but with the voltages reversed. Gate must be more positive than the source to cut the channel current back into the linear region. Best debugged by Audio Probe.
- Operational amplifiers (op-amps)
  - For an opamp to be working as a linear amplfier, both the inverting and noninverting inputs and the output pin must be within a few millivolts of the voltage on the noninverting input. This is a consequence of the negative feedback..
  - Measure the +input, -input and output pins of a single opamp. If they're not within 10mv of the voltage on the (+) pin, either the opamp is dead, the biasing is seriously wrong, or it's being used as a comparator and switching. If you know it's supposed to be used as a linear amplifier, there's a problem in this stage.
  - When an opamp is used as a comparator, you can still tell if its bad. The output will bang very close to the + or - power supply depending on which input pin is higher. If the - input is more negative than the + input, the output must be at the + power supply, or the opamp is dead. If the -input is more positive than the + input, the output must be at the - power supply or the opamp is dead.
  - Be aware of the pinouts of the chips you use. Get on the net and find them and print them out. Many opamp IC's have more than one opamp inside. Dual opamps are very common, usually more common than single opamps, and have the advantage of always having the same pinout. Quad (4!) opamps are fairly common, but there are several pinouts for these, so be sure you know which pin is which.
  - Measure the voltages directly on the pins of the IC, not at a socket pin. Sometimes socket contacts are bad. As are solder joints.
Audio Probe Method
You'd like to use an oscilloscope, which shows you the electrical waveforms directly in the circuit, but these are expensive, and require some skill to work with properly. An approach that works almost as well is to use your ears. This method is an old one, but was most recently revived by Craig Ochikubo.

Take an enclosed phone jack that fits the end of a guitar cord and solder a 0.1uF capacitor onto the signal lug, and a wire with an alligator clip on the end to the "ground" lug. Then wrap the jack in electrical tape so you can still plug a guitar cord into it but the ground wire and one lead of the capacitor stick out the end. Here's a picture. Then you can fire up your amp, plug in a guitar cord, and slip the audio probe onto the free end of the cord. Clip the ground wire clip onto the effect's signal ground, and you can poke the free end of the cap onto connections in the effect and hear whatever audio is there - if any! It's the "if any" that tell you stuff, of course.
Then do this:
1. Set the amp's volume low - you may get BIG signals at some points.
2. Power up the effect, plug in your guitar to the effect input.
3. Probe the input jack where the guitar cord connects. Signal? If not, the guitar cord you're using from the guitar could be bad, the jack could be bad, or the guitar could be turned down (I won't tell you how I found those... ;-)
4. Now probe the input to the effects board - does signal get though to there when the effect is engaged by the bypass switch? If not, could be bad wiring or a bum bypass switch.
5. On your throwaway schematic, follow the signal through the schematic, probing points where the signal is supposed to be. Mark signal present on your schematic with something like a red squiggle for "signal OK". When you find a place where there's supposed to be signal and there's not, you're just past the problem - or just past a problem at least. Here's some common ones you may find:
  - Bad solder joint not making good contact
  - Excess solder ball or thread shorting something
  - Broken wire or PCB trace
  - Reversed diode, or transistor pinout not right (although you should have found these from the voltage checks)
  - Electrolytic capacitor reversed. If you think you found one of these, get your voltmeter out and measure the actual voltage on the capacitor leads. If the voltage is not more positive on the (+) lead than on the (-) lead, the capacitor is backwards! Note that in some circuits this can be the right way for the capacitor to be inserted, but some other part is bad or wrong and putting the wrong voltage on the capacitor!
  - Wrong parts placement or values (I'm a bit color blind, so I'm forever getting 4.7K's mixed up with 47K's)
6. Here's some things you should expect:
  - Signal is bigger (louder) at a transistor's collector than at its base if there's a collector resistor bigger than the emitter resistor.
  - Signal is the same loudness at a transistor's emitter as it was at its base IF the emitter is not tied to ground with a capacitor
  - Signal is the same loudness or louder at an opamps output than before it went into the inputs.
7. When you've worked your way all the way to the output, there's bound to be something coming out the end!! Congratulations! You're an effects builder.
General mechanical/electrical problems
- Get yourself a 3X or 5X jeweler's loupe. These are the little black plastic cylinders with the lens in one end that you can hold in your eye socket by squinting. These will let you see great details on the PCB. Go over the whole soldering side of the PCB with it to find soldering and other problems.
- Solder joints can be intermittent. If the wires or leads in a solder joint were not clean, one of the wires can actually have an insulating covering of dirt, oil, or oxides that keep it from connecting to the other wires even though the solder covers the wire. This can also be intermittent, sometimes making contact, and sometimes not. The same thing can happen if you did not clean a PCB before soldering. The copper pad will not "wet" with the solder, an the wire component lead may hold a blob of solder there that looks OK, but is not making good connection to the copper pad on the PCB surface.
- Solder joints can be "cold". Cold joints will look grainy, not smooth, and usually happen when you move one of the wires while the joint is cooling down as you remove the soldering iron.
- PCB's may have hairline cracks that open intermittently; you'll need a loupe to see these.
- PCB's may have hairline shorting connections that are very hard to see, but are great shorts and keep your circuit from working right. Often a loupe is needed to see these.
- If the effect works when it's out of the box but quits when you put it into the enclosure, check to find where a wire, component or solder joint may be touching the enclosure and shorting out.
- Sometimes putting an effect in a box makes it quit working and it's NOT shorting out. This can happen when the mounting for the board twists or flexes the board and opens a bum solder joint or hairline crack in the board.
- The opposite can be a useful debugging trick; gently bend or twist your board while listening to it through an amp. This may make an intermittent contact or crack open or close.