DOD FX-50 Clipping Diodes

[soupbone]

I bought a DOD FX-50 for 30 bucks.It sounds pretty good,but it's a little fuzzy for me.I was going to mess with the clipping diodes,and see what i can do with it.I can't find a schematic anywhere,but i found this site that has a gut shot of the inside;http://www.americaspedal.net/fx50/fx50_279k.jpg  You'll see 2 pairs of diodes at the top,and there's 2 more right behind the big yellow cap at the top to the right.Anybody know which ones are the clippers?

[kristopher612]

looking at the schematic as well, you'll see 2 sets of clipping diodes.  one set is in a feedback loop like a TS, and then a set to ground.  i personally would take out the ones in the loop, they're in parallel with the 1M resistor. give it a shot. 

[kristopher612]

where you at in Oklahoma, BTW?  i live right outside Tulsa myself.

[Mark Hammer]

If you look at the schematic for the DOD250 and MXR Distortion+, you'll see considerable overlap.  Both of those have a 4k7/47nf+500k pot trio for the ground leg of the first op-amp, and both have a 10k,2*diode, plus .001uf arrangement on the output of the op-amp.  Where the FX50 differs is that it also has a diode pair in the feedback loop.  I'm not exactly sure how an identical diode pair to ground will introduce any additional clipping when there is already a diode pair in the feedback loop, since the feedback loop pair will effectively set the ceiling on the output amplitude of that gain stage.

However, whether it makes sense to me or not, many of the same recommendations that will make the 250 and Dist+ more usable pedals also apply here.  So, increasing the value of the .047uf (47nf, 473) cap ny the gain pot will add more bass, and avoid losing it at higher gain settings (.22uf is a nice value to aim for).  As well, adding a small value cap in parallel with the feedback loop components will tame the fizz (33pf-39pf is a good value to aim for).

What this particular circuit allows you to do, is experiment with feedback loop clipping vs ground-diode clipping.  So, you could lift one end of each diode pair, and use a DPDT toggle  to place the one pair back in circuit, or the other.

Another mode to consider is to make the clipping threshold higher for the op-amp, so that the second set of diodes will add more audible clipping.  So, you might add a pair of germanium or schottky diodes in series with the existing 4148 diodes in the feedback loop to raise the clipping threshold i the gain stage.  The pair to ground would still clip when the signal amplitude was high enough, but the first set wouldn't until the gain was up enough.    Once you dime the gain, you get a harder clip out of the second set of diodes since the output is not only already clipped but has a higher forward voltage than the diodes to ground.

[amptramp]

I'm not exactly sure how an identical diode pair to ground will introduce any additional clipping when there is already a diode pair in the feedback loop, since the feedback loop pair will effectively set the ceiling on the output amplitude of that gain stage.

Witht he non-inverting amplifier stage, the diodes in the feedback loop will clip and essentially connect the output of the op amp to the inverting input (with a fixed offset of a diode drop) which reduces the gain to unity for the signals coming in.  There may still be enough signal to clip if the input is above the threshold, which is possible if the signal comes from a buffer with some gain.  Some very hot pickups may be able to drive the diodes to ground directly at unity gain.

All four dides are shown to be 1N4148 silicon but you could get interesting effects from using LED's in the feedback and silicon to ground - the silicon to ground could clip first in that case.

[soupbone]

where you at in Oklahoma, BTW?  i live right outside Tulsa myself.

Tulsa!

[soupbone]

If you look at the schematic for the DOD250 and MXR Distortion+, you'll see considerable overlap.  Both of those have a 4k7/47nf+500k pot trio for the ground leg of the first op-amp, and both have a 10k,2*diode, plus .001uf arrangement on the output of the op-amp.  Where the FX50 differs is that it also has a diode pair in the feedback loop.  I'm not exactly sure how an identical diode pair to ground will introduce any additional clipping when there is already a diode pair in the feedback loop, since the feedback loop pair will effectively set the ceiling on the output amplitude of that gain stage.

However, whether it makes sense to me or not, many of the same recommendations that will make the 250 and Dist+ more usable pedals also apply here.  So, increasing the value of the .047uf (47nf, 473) cap ny the gain pot will add more bass, and avoid losing it at higher gain settings (.22uf is a nice value to aim for).  As well, adding a small value cap in parallel with the feedback loop components will tame the fizz (33pf-39pf is a good value to aim for).

What this particular circuit allows you to do, is experiment with feedback loop clipping vs ground-diode clipping.  So, you could lift one end of each diode pair, and use a DPDT toggle  to place the one pair back in circuit, or the other.

Another mode to consider is to make the clipping threshold higher for the op-amp, so that the second set of diodes will add more audible clipping.  So, you might add a pair of germanium or schottky diodes in series with the existing 4148 diodes in the feedback loop to raise the clipping threshold i the gain stage.  The pair to ground would still clip when the signal amplitude was high enough, but the first set wouldn't until the gain was up enough.    Once you dime the gain, you get a harder clip out of the second set of diodes since the output is not only already clipped but has a higher forward voltage than the diodes to ground.

Thanks Mark!I clipped the 2 diodes right above the 1M resistor to see what it sounded like.Not a big noticable difference.I'm going to try some differn't combo's and see what i like(Led's,1n34a's,1N4001's etc.).One question;Which cap on the schematic were you saying to change to 33pf-39pf?Thanks Mark!

[soupbone]

looking at the schematic as well, you'll see 2 sets of clipping diodes.  one set is in a feedback loop like a TS, and then a set to ground.  i personally would take out the ones in the loop, they're in parallel with the 1M resistor. give it a shot. 

Thanks fellow Oklahomie!

[kristopher612]

no problem!  i think mark means to addd one in parallel with the 1M resistor in the feedback loop of pins 1 and 2.  you could also try jack orman's warp controls or saturation controls on the diodes to ground if you don't really want to change them.   you might be able to do the same with the diodes in the feedback loop, but i don't know, and someone else might be able to say, if it would change the overall gain of the circuit.
http://www.muzique.com/lab/main.htm
those controls are in this page, just look for them

[teemuk]

Usually when things sound "fuzzy" tampering with clipping devices is the most ineffective solution. Try mods that reduce low frequencies before clipping.

[Mark Hammer]

kris is correct.  The nice thing is that small ceramic caps of that value range can be easily tacked onto the copper pads of the 1M feedback resistor, such that you can try out a few different values for the one you find optimal.

Note that, if you DO use such a cap, and remove the feedback loop diodes, as kris suggested, you can still have a bit of fizz.  Keep in mind that the clipping diodes to ground will add harmonic content...because that's what clipping does.  The suggested feedback cap reduces the amount of gain applied to the top end, using the formula:  Freq = 1 / [2 * pi * R * C]

When the clipping diodes are the pair going to ground, the "pre-filtered" signal from the output of the gain stage hits those diodes and generates additional harmonic content during the clipping.  The .001uf cap in parallel with those diodes does some treble cutting, but not all that much, since it starts at a fairly high frequency.  So, if you DO use the ground pair for your clipping - DOD250 and Dist+ style - then not only will you want to add the feedback cap, but you will probably also want to increase the value of the cap to ground from .001uf to somethng higher, maybe even going as high as .0047uf or .01uf.

In contrast, if you use the feedback diodes for your clipping, the necessary "anti-fizz" filtering is already done by the suggested feedback cap.  I certainly wouldn't make my decision based on the cost of using one treble-taming cap vs two, since that's a 10-cent factor.  Try lifting one pair of diodes, and then the other, and see which pair meets your needs.

Teemuk is also correct in asserting that taming the low end  can "smooth out" clipping, by essentially attenuating that part of the signal that is closest to the clipping threshold.  That is, in fact, the strategy used by the Tube Screamer and similar.  In the case of the FX-50, and its predecessors, the DOD250 and Dist+, my sense is that that choice to trim the bottom as gain is increased was really more chance, or motivated by other reasons.  In particular, tapering off the bass as gain is increased would make the distortion seem more "piercing" at higher gain settings, akin to an additional treble boost without the cost of adding an additional control to achieve it.  It also has the interesting side effect of making any hum coming from single coil pickups not quite as apparent as the gain is increased - a very real consideration at the time the 250 and Dist+ were developed.

At this time in guitar history, though, many more guitarists have hum-rejecting or hum-reducing options available to them than 30 years ago, and they can now place more emphasis on tonal goals than on noise-control.  At least that's MY take on it.  The perk of being able to do that means that one can use the additional bottom end that more readily results in clipping, and leverage it into being able to use a higher clipping threshold, which, in turn means you can rely on the hotter output level of the pedal itself to push the amp into overdrive, using a signal that the pedal has only mildly overdriven.  The combined grind of a pedal AND amp is often the preferred tone of many players.  So, if you don't have to rely on the pedal alone, that's a favourable situation.  Why not take advantage of it?

[Mark Hammer]

(deleted - duplicate post)

[kristopher612]

just wanted to add, that i really like Mark's idea to use a DPDT toggle to remove a single set of clippers in on position and the other when flipped.  its like getting a tubescreamer and a D+ in a single box.  this pedal may also benefit from something like the SWTC or something like it.
would changin that 470K resistor that's going to the voltage divider have any noticable effect on the tone?  i know its not there for the tone necessarily, but in the TMK fuzz designed by tim escobedo, there is a pot set up as a variable resistor there, so i don't know if that would do anything noticable here.  that value is 1M in the D+, 510K in the TS and 470K in the DOD 250. 
I also agree that changing the .047 cap to something bigger would probably be better.  i was never happy with the bass in a stock D+, and .22 is just about right IMO.

[teemuk]

would changin that 470K resistor that's going to the voltage divider have any noticable effect on the tone?

It's part of a RC circuit with the 10nF input cap and also has an effect to the impedance bridging with the guitar+coord circuit so... yes. Probably won't make a lot of difference if the value remains moderately high but once you start to go significantly lower than 470K you will definitely start to hear effects.

Lower values will cut more lows but simultaneously also kill the resonant peak of the pickup/guitar circuit, which you will hear as loss of highs and "characteristic tone" of the pickup.

also agree that changing the .047 cap to something bigger would probably be better.  i was never happy with the bass in a stock D+...

I still hold on to my point of view that if the thing sounds too fuzzy then you do not want to add any more bass to the response (before clipping). You want to remove more of it (and then possibly boost some back post distortion).

[kristopher612]

I still hold on to my point of view that if the thing sounds too fuzzy then you do not want to add any more bass to the response (before clipping). You want to remove more of it (and then possibly boost some back post distortion).

this is true.  i hadn't thought about it, but the ibanez sd9 has a cutoff at about 1500 Hz, but with the tone control aafterwards, plenty of bass is there to be had, and it sounds very natural. 

[Mark Hammer]

Depends on what you have in mind for the purpose of the pedal.  If the idea is to have it on all the time - the way some folks use a TS-9 - then the loss of bass is not problematic, and the user will likely adjust their tone at the amp.  If the intent is to use it as "an effect" that gets turned on and off, then the missing bass content will be noticed. 

Again, the tone-shaping is neither good/correct or bad/incorrect, in and of itself, but compatible or incompatible with the way one intends to use it.  Myself, I turn these things on and off, so the tonal compatibility between my bypass and effect tone is something that I'm conscious off.  I want the tone introduced by a distortion/overdrive to be more agressive-sounding than my bypass tone, but I don't want it to be thinner, and regardless of how much bass is present or absent, I don't want it to be the sort of piercing treble that sends me running back to the amp to turn the treble down.  Ideally, for me anyway, a 2-knobber (drive + output level) should not require any tone adjustments at the amp.  And if it does, then there needs to be some form of tone control on the pedal itself so that amp adjustments are not required.

[soupbone]

Depends on what you have in mind for the purpose of the pedal.  If the idea is to have it on all the time - the way some folks use a TS-9 - then the loss of bass is not problematic, and the user will likely adjust their tone at the amp.  If the intent is to use it as "an effect" that gets turned on and off, then the missing bass content will be noticed. 

Again, the tone-shaping is neither good/correct or bad/incorrect, in and of itself, but compatible or incompatible with the way one intends to use it.  Myself, I turn these things on and off, so the tonal compatibility between my bypass and effect tone is something that I'm conscious off.  I want the tone introduced by a distortion/overdrive to be more agressive-sounding than my bypass tone, but I don't want it to be thinner, and regardless of how much bass is present or absent, I don't want it to be the sort of piercing treble that sends me running back to the amp to turn the treble down.  Ideally, for me anyway, a 2-knobber (drive + output level) should not require any tone adjustments at the amp.  And if it does, then there needs to be some form of tone control on the pedal itself so that amp adjustments are not required.

The pedal actually has a little bit of a bass boost to it.I was going to use it sort of a like a tube screamer,but more as a booster,with the drive down and the level up.It does have a lack mids which i didn't mention.More mids,and no fizz is basically what i was going for.Thanks Mark,and to everyone's responses!

[ashcat_lt]

It's my understanding that diodes in the feedback loop cause the thing to apply its set gain up until it would cause the output to overcome the diode drop, and then goes to basically unity plus the diode drop. 

That is:
if
  in x gain < diode drop
then
  out = in x gain
else
  out = in + diode drop

If that's actually true, then the diodes to ground (assuming they have the same forward voltage as those in the feedback loop) will chop off the rounded parts that the feedback loop clipper left on top of the nearly vertical sides.  I don't really see how this is any different from the sound of a straight diode-to-ground clipper like the Dist+ or the Rat with the same gain.

[kristopher612]

The pedal actually has a little bit of a bass boost to it.I was going to use it sort of a like a tube screamer,but more as a booster,with the drive down and the level up.It does have a lack mids which i didn't mention.More mids,and no fizz is basically what i was going for.Thanks Mark,and to everyone's responses!

then i would look at lowering that .047 cap for sure.  between .047 and .022 should probably get some of the bass boost out, and adding the cap in the feedback loop should help too. 

[Mark Hammer]

Again, depends on how you're gonna use it.  The bass rolloff frequency is given by F = 1/[2 * pi * R * C].  With the gain control set to 100k resistance, the gain is around 10.5*, and the bass rolloff begins around 32hz.  With the gain pot down to 10k, the gain is now 69* and the rolloff begins around 230hz.  Drop pot resistance to 5k and gain is now 104X with rolloff starting around 350hz.  At max gain, the rolloff begins around 720hz.

The lesson here is that if you use it with the gain below max, then there will be little or no audible bass loss.  If your preference is to use it near max gain, then the bass loss WILL be noticeable.  My suggestion to increase the value of the .047uf cap was for those whose preference is to run the unit fairly near max gain most of the time.  Keep in mind that the use of germanium clipping diodes will keep the output level fairly modest.  Use of silicon 1N4148 will allow for a noticeably hotter output level, such that decent boost-n-grind can be achieved at less than max gain (in comparison to the Distortion+).  On the other hand, use of those same silicon diodes will also result in less clipping, increasing the tendency to dime the drive/gain control.

Again, no set orthodoxy here.  If your uinclination is to use it for just a touch of hair, then the values shown will do just fine for bass.  If your inclination is to crank the gain up and back off on the volume, then the way to recover that lost bass is by increasing the .047uf cap value.

Regardless of how you address your bass end, unwanted fizzies and harshness are cured by the suggested feedback cap addition, and by increasing the value of the cap paralleling the clipping diodes from .001uf to something roughly 2.5-4 times the existing value.