So, I think I fixed my FAB Metal, but... what happens in the long run?

[goulashnakov]

OK, So I'd had some trouble with my Danelectro FAB Metal pedal, and I finally got around to trying to ID and fix the issue.

The pedal would function as it should at first.  Hit the stompswitch, the effect is engaged.  Hit it again, it's "bypassed."  (It's not a true bypass, but an electronic/solid state)
What started happening was that, after the distortion was engaged for a few minutes, the sound would start to intermittently cut in and out.  It would only cut out a little at first, but it would become progressively more severe, until it would reach a point where the sound would be gone most of the time and intermittently cut in.  Finally, it would reach the point where no sound would come out at all while the effect was engaged.  Bypass would still work fine, i.e. I'd still get clean signal with the effect "off."

So now, the debugging:

Lots of audio probing, voltage readings, visual tracing (with the help of a jeweler's loupe.  Mad props to whoever made the suggestion to invest in one of those).  All as per the guide to debugging laid out in this forum.  Long story short, I trace the problem to the bypass switching circuit:

The switching circuit uses two chips; a CD4013 flip-flop circuit, and a HEF4053B analog multiplexer.  The HEF4053B uses magic and voodoo to switch between two channels; bypass and distortion.  The flip-flop circuit switches the voltages of the input control pins on the 4053 to determine which channel the signal goes through.  This is all detailed much more than I care to elaborate in the 4053's datasheet.

So, I determined that the part of the circuit that was behaving oddly was at Pins 9 and 10 on the HEF4053B (the control input pins for the drive/distortion/"effect on" channel). Pin 10 is directly connected to one of the outputs on the CD4013 flip-flop, while pin 9 is coupled to pin 10 by way of a 470k resistor.  There is a diode between pin 9 and supply voltage (the way it's connected, it is always reverse-biased, so I'm guessing it's there either for protection against voltage spikes when the flip-flop circuit flips and flops, or maybe as a reverse polarity protection).  Finally, there is a small capacitor from pin 9 to ground.

When the pedal was in "bypass" mode:
pins 9 and 10  V<<4.5 volts, (which I believe 4.5 was the voltage at the reference pin 6 aka "E-bar" in the datasheet.  For the channel to be ON, the voltages on pins 9 and 10 have to be greater than "E-bar") 

When the pedal's "drive" channel was engaged:
pins 9 and 10:   V > 8 volts 

However, after giving the pedal time to do its thing and start acting up, I checked those voltages again:
pin 10:    V > 8 volts
pin 9:     V < 4.5 volts

In fact, as pin 9's voltage got lower and lower, the pedal's sound would cut out more and more.  By time the pedal had reached its state where the sound was completely cut out, pin 9 was at less than 2 or 3 volts at any given point.

My hypothesis was that the capacitor had gone bad, and that after the drive channel of the pedal was on for a few minutes, that capacitor would start leaking, and eventually pull pin 9's voltage down to where it could no longer effectively hold the drive channel in the ON state.  But, when the pedal was shut off for a while (like an hour or so) and then I turn the pedal on again, that cap would no longer be leaky, and it would again work fine for a while before it would start leaking again. 

So I de-soldered the cap, and so far it seems to have cleared up the issue.  Now the pedal works fine again.  The voltage on pin 9 stays where it should, and there is no more cutting in and out with the sound.





So, I think I fixed my FAB Metal, but... what happens in the long run?

In other words, for those of you who are still awake, here is my main question for all of this:  What was the purpose of that capacitor?  (Obviously, it wasn't there just to fail and cause my pedal to misbehave).  Could it have been there to filter out noise, or maybe to control voltage spikes?  I didn't really notice any new noise after I took the cap out, nor did I notice any popping when pressing the stompswitch.  But, before I finish putting this pedal back together, I want to make sure it would be safe in the long run to leave this capacitor out (as I have no replacement for it), or if this capacitor is so crucial to the circuit that leaving it off could eventually cause damage in the long run?

I apologize for this post being so long-winded, but I wanted to see if there was anyone familiar enough with this type of HEF4053B circuit to determine how essential to the circuit design that capacitor was.

[goulashnakov]

I went back and looked at the circuit again, and re-checked some of the voltages to make sure I'd remembered everything correctly.  I discovered a couple of errors in my previous statements.  Also, I decided to make a quick schematic of the bypass circuit section in question to better describe the situation (I shoulda done this to begin with, to save words.  )



(Value of capacitor Cx is unknown)


Relevant Voltages (re-checked):

Supply Voltage:  9.6 V (dano "zero hum" battery eliminator)

Effect Bypassed:
Pin 10:  0V
Pin 9:  0V
Pin 6:  0V
Diode's Cathode:  0V (Diode Anode connected to pin 9)

Effect Engaged:
Pin 10:  9.45 V
Pin 9:  8.85 V
Pin 6:  0 V
Diode Cathode:  9.45 V

Prior to removing capacitor Cx, voltages after effect has been engaged for some time (enough time for it to start it's misbehavior):
Pin 10, & Diode Cathode:  9.45 V
Pin 6:  0 V
Pin 9:  less than 4 V (after initially being at 8.85 V, it would slowly start dropping.  Sound would start cutting cutting out at about 6 volts.  By time Vpin9 had hit <4 V, sound was completely gone.)

After removing capacitor Cx, voltage at pin 9 was back to 8.85 volts with effect engaged, and did not drop off over time.


As for my mistakes:
First of all, Pin 6 (E-bar) is NOT 4.5 Volts like I originally thought.  That was a misremembery on my part. 
Pin 6 is actually appx. 0 Volts all the time.

Second, the diode attached to Pin 9 is NOT connected to supply voltage.  It is actually connected in parallel with the 470k resistor.  Anode on pin 9, cathode on 10.  So it sees the 9 Volts from the output of the CD4013 flip-flop only when effect is engaged (which must have been the only time I had measured that voltage before).  So I guess that clears up the mystery of the diode to me; (if I understand correctly) it helps pull down the voltage at pin 9 when the effect is bypassed and CD4013 goes from flip to flop.


What still is certain:  That capacitor Cx was probably leaking DC.  When this happened, it formed a voltage divider with the 470K resistor and pulled down the voltage of pin 9 until it was no longer sufficiently higher than pin 6 to keep that channel open.  Taking the capacitor out immediately solved that problem.


What is still not known:
What was the purpose of that capacitor, and could having removed it potentially cause problems in the future?

[PRR]

> those of you who are still awake

zzz.... zzz.... zzz.... zzz.... zzz....

I've spent 20+ minutes reading, but I do not know the pedal and don't have the 4053's logic memorized.

You musta spent 40 minutes writing all that.

If our time is worth a penny a minute, this is already a 6-dollar question.

The cap is 20 cents. Put the cap in.

[amptramp]

Pins 9 and 10 are select pins for their respective channels on the 4053.  It appears that the designer wanted pin 9 to go down almost as soon as pin 10 when the input from the 4013 went down but wanted pin 10 to come up earlier than pin 9 when the 4013 input went up.  Pin 9 controls a single-pole double-throw output at pin 4 to either pin 3 or 5.  Pin 10 connects a similar switch at pin 15 to either pin 1 or 2.  Inputs and outputs may be interchanged.  It looks like a break before make or make before break function.  Pin 9 high makes 4 connect to 3.  Pin 10 high makes 15 connect to 1.

This kind of connection is usually a method to prevent switch pop.  Replace the capacitor.

[goulashnakov]

so, then...

The capacitor is crucial to the circuit because... it's only 20 cents and my posts are too long? (JK.  I'm glad my humor isn't completely lost on you, Paul)

The capacitor is crucial to the circuit because... it may prevent switch pop?  (But I haven't encountered any pops since I removed it.)

Or am I completely on the wrong track?  I'm sorry if I'm a little slow to pick up on this.  (I write/talk too much, but don't think enough).

How about this one:  The capacitor is crucial to the circuit because... it will irreversibly damage the circuit if it is left out? (Or not?)

Advice is nice, but answers will suffice. 

I can't replace the capacitor because it's one of those teensy weensy surface mount dealies, smaller than a fraction of a grain of rice.  Removing the bad one was easy, but putting in something to replace it?  Not with my limited skills and equipment.

So if I can get away with not replacing the capacitor, then all the better.

But, if the circuit absolutely cannot be without it's beloved capacitor, then I am hosed.  Time to buy/build a new Mettle Peddel.

[DougH]

Advice is nice, but answers will suffice. 

See, the thing about DIY is you will have to weigh all the advice, read what you need to read, and produce the answer yourself.

Sounds like if the switch isn't popping, you're done.

[goulashnakov]

See, the thing about DIY is you will have to weigh all the advice, read what you need to read, and produce the answer yourself.

Sounds like if the switch isn't popping, you're done.

Right on.  And I entirely agree.  (Sorry if I come across like I'm brushing off anyone's advice or wisdom, that's just my bad social skills.  No hatin' is intended.     )  As part of weighing all the advice, I feel I have to question everything presented in order to probe the logic behind the statements.  It's the only way I can really wrap my tiny little brain around the concepts.  (Ever hear "there are no stupid questions, only inquisitive idiots?"  Yeah, I was never a teacher's favorite.  ) 

Anywhoozit, what you folks have said seems to concur with my suspicions that the mystery capacitor is just another noise filter.  But, y'know, I figured it wouldn't hurt to ask the community anyway.  Thanks, community!  You guys have been great.  Rock on \m/ (>_<) \m/

An afterthought:
The pedal actually started its malfunction back in the summer.  I'd written it off as a total loss at the time, but didn't want to get rid of it yet because of sentimental value?/couldn't help but feel like if I tossed it I would later learn of a way to fix it after the fact.  Anyone else experience this (urge to keep your junk pedals for either of those reasons)?
Before working up the motivation to try and decipher the chaos inside, I came up dry while searching for schematics/info about the pedal.  But a post from another forum that popped up in a google search claimed that the FAB Metal is basically a copy of the DOD FX69 / Digitech "Grunge" circuit, with only the treble working for the tone control, and the bass pot replaced with a couple of resistors for a fixed setting.  That and, of course, different switching circuits.  A trip to the music store and a chance to plug into a Digi Grunge was enough to convince me, and an employee at the store who'd also heard that claim noticed a similarity when he AB'd them.  And now, more recently, after getting a chance to closely examine the FAB's circuit and briefly compare the topology to a DOD Grunge schemo, I think we've got a winner.

So, I figured I'd pass that little bit of info along, so that if anybody is looking for a schemo to build a FAB Metal clone, just go Grunge.  For that matter, anybody looking to buy a FAB Metal, just go Grunge.  Even if they aren't the same distortion circuit part for part, well, horseshoes and hand grenades I say.