i'm building a distortion pedal but i've hit a design snag...

[methodofcontrol]

my pedal works and sounds pretty good. similar to a RAT in tone but w/ less harshness. i want the pedal to sound the same but w/ increased high-end and low end. i copied a rare Japanese pedal for the basic design, the biggest differences being that my pedal omits some unnecessary features as well as omitting the on-board transformer and AC line cord and over built power supply.
the original pedal starts w/ a 36v transformer, TONS of filtering and a couple of regulators until you have a +15v rail and a -15v rail feeding the circuit.
my pedal starts w/ a standard 9v power supply into a MAX1044 to get +9v and -9v. i've since swapped the MAX1044 for a LT1054 and bumped the input voltage up to 15v giving me +15v and -15v. increasing the voltage on my pedal didn't produce a noticeable difference

the two pedals have the same timbre, but the original sounds hi-fi w/ extreme low end and high-end capabilities compared to mine. someone suggested to me that perhaps the overbuilt power supply on the original provides that hi-fi quality and that, in fact, overbuilt power supplies are very common in hi-fi audio applications.

thoughts?

[Mick Bailey]

Difficult without comparing schematics. What is the original pedal?

Have you measured your power supply output while the pedal is working and with a good input signal? It could be that the LT1054 isn't delivering enough peak current to cope with the lower frequencies. The original pedal may have more capacitance and higher peak current delivery (though a distortion circuit doesn't generally need much current). Numerous other factors can affect the pedal's tonal range and it can be very difficult to reproduce a pedal exactly.

[Mark Hammer]

Mick's right. We need a schematic, or at least some point of reference (e.g., the original "rare Japanese pedal"). And if you're trying to develop a commercial product and "borrow" advice, you may want to rethink that.

A few preliminary considerations, though:

- Why the need for such a power supply? Is it a matter of current draw? A requirement to use a bipolar supply? Just the supply voltage? If the original needed enough current to require a transformer, then a charge pump may not deliver the goods in terms of current, even if it can give you the requisite bipolar supply at higher voltage.

- Higher supply voltages do tend to raise the headroom, although headroom is not necessarily what one wants in a distortion/overdrive.

- Plenty of pedals have "overbuilt" supplies and do not really benefit from them; the manufacturer just went a little overboard. The original EHX Hot Tubes pedal had a power cord and onboard transformer, but if you make one and power it with a 9v battery, you'll find it runs just fine. The battery won't last a LONG time, but it won't die in an hour either.

- Big bottoms and sizzling tops are generally a byproduct of midscoop filters. Keep in mind that since what we call distortion is really multiples of the fundamental, and since guitars don't have fundamentals THAT low, scooping out the mids will leave you with a bottom that seems clean, simply because the harmonics are not as audible.

[methodofcontrol]

here's some hand drawn schematics of my pedal (power supply not drawn, but it's a charge pump IC. the negative rail goes to pin 4 of the op amps and the positive rail goes to pin 8.), this schematic is exact to the original pedal minus some features that have nothing to do w/ the issue i'm trying to solve. the other drawing is the power supply for the original pedal. I did it on the fly some time ago when i had the japanese pedal in hand. let me know if anything doesn't make sense.

edit: to clarify, other than omitted features, the ONLY difference between my pedal and the original is the power supply. i really want my pedal to operate on standard boss adapter so the biggest help would be in figuring out how to maintain the extreme high end and low end of the original into my clone w/out losing standard operating voltage. an explanation as to why a beefed up power supply would help provide extreme highs and lows would also be appreciated.
and as always, thanks so much for the help!

[Mick Bailey]

Voltage is voltage - if your PSU can maintain +/- 15v into your circuit under all conditions then it won't matter how over-built it is. It could be capable of supplying tens of amps at that voltage and won't make a bit of difference. But if it's lacking then the current draw could pull one or both rails down under heavy bass transients. You'd have to measure that.

As for the HF side, I'm thinking there are cumulative differences. You say that the only change to the original pedal is the PSU, but did you measure every component in the original circuit? the tolerances and subsequent drift over time can make a considerable difference - the ESR of caps also changes. Also, did you use the same spec for the components or substitute say, MF for CF, film for ceramic? Is the construction to the same layout? It all adds up.

[g1]

You must have access to the other pedal or you would not know firsthand of the difference in frequency response. So try your circuit with the other supply, and try the other circuit with your supply.
Then you will know if the supply is the issue.

[J M Fahey]

Sorry but you are comparing actual sound to a dim memory
And you are not using the same guitar / amp / cable / strings / etc. nor are you playing it in the same place nor are you the same person.
It all adds up.

EDIT: and no, it's not the power supply.
And your circuit will be happy with +/-9V rails.

[methodofcontrol]

ahhh... to clear up some of the confusion. YES- i have compared my pedal to the original recently. (same guitar, amp, room, ears, etc...) i borrowed the pedal from a friend to clone it. i have since returned it. putting MY pedal's bass and treble knobs all the way up gets me the same sound as the original w/ the tone knobs up to say, 6 or 7. the original just has more OOMPH to give. specifically in the tone controls, i can match gain and volume no problem but my clone can't put out the extreme highs and lows as the original. as far as cumulative differences... i totally understand the suggestion, but this is not a subtle difference i'm trying to describe. something is definitely amiss.

as to the voltage... that's what i thought. if i can provide the same voltage w/ a different supply, why wouldn't the pedal sound the same? the current draw affecting bass transients must be the point that a hi-fi enthusiast was trying to make to me, (which is what got me thinking about robust power supplies affecting tone). thanks for jogging my memory there!

there is one thing i forgot to mention about the original pedal... on each pin 4 and pin 8 of the signal op amps there is a 100 uf electrolytic to ground in parallel w/ a .1 uf tantalum. (since pins 4 and 8 of the op amps are referencing the power supply, i omitted this from my clone.) my understanding is that caps in parallel are simply adding capacitance, so why would a 100uf cap be paralleled w/ a .1 uf? what does that achieve?

g-one's suggestion is so simple and staring me in the face, and i can't believe i hadn't already thought of that. i'll see if i can re-borrow the pedal and do that w/ out cutting traces on the original! if nothing else, i will at least confirm or deny the power supply as a culprit f the differences i'm hearing.

[Enzo]

WHAT bass current transients? This is class A circuitry so the current draw should be pretty much flat.


SO you removed all the bypass caps from your circuit? You might like to return them. If you look in any big PA mixer, you will find each IC has bypass caps from the power pins to ground. Unless your circuit is tiny, there wants to be bypass right at the ICs. That is separate from the main power supply filters. Why is a 100uf in parallel with a 0.1uf? Just to mess with your head... Nah. That 100uf cap is real good at filtering out low frequency stuff, like hum and audio. But when it gets up to higher frequency things, like transients (I knew we'd work them back in) those high value caps are a bit slow.

And if you look at data sheets for things like voltage regulators you will see notes to the effect, this bypass cap needs to be close as possible to the IC.


This may not be involved, but worth checking.

And a tip: When you see "extra" parts in a circuit like that, your first reaction should be "What am I not understanding that these are needed?" As opposed to "Hey, what are these for? I'll just toss them out."

[Mick Bailey]

The schematic for the 4558 shows a class AB output stage.

[silverfox]

Not to change the subject but

I've posted an unrelated question regarding the schematic for this pedal. It is here: Tube Version of OP Amp Circuit

[methodofcontrol]

enzo- always so much trusted information! thanks for chiming in!
to address your concerns about the IC bypass caps, when i first cloned my friend's pedal i didn't understand the bypass caps other than as standard power filtering on the + and - rails. since MY power supply came off a charge pump, i had a hunch that the electrolytic i put on my DC jack was doing the same thing. so i listened to the first build then i listened again after removing the "extra" caps and i heard no difference in noise or in response. then i tried increasing my DC jack electrolytic from my standard 100 uf to a few different values up to and including 1000 uf and still heard no difference. i went back to 100 uf to keep it simple and decided those "extra" caps were a necessity only for the original power supply. most pedals i've seen do not have power filtering beyond the DC jack, now that my pedal operates from a standard 9v supply i figured i was safe in discarding the "extra" caps. please let me know if you still think that was a mistake on my part. thanks!

[pdf64]

' i listened to the first build then i listened again after removing the "extra" caps and i heard no difference in noise or in response'
Did you check over all combinations of control settings and other relevant variables? How about after the electrolytics have aged a few years?
Bear in mind that instability at super audio frequency may not be immediately apparent.
Pete

[J M Fahey]

The schematic for the 4558 shows a class AB output stage.

True, which implies it's biased into Class A while idle, and pulling more current (getting into the Class B area) when load demands go beyond that.
That's the definition of Class AB.

Now how much current is that?

An RC4558/TL072/LF353, etc. (the classic workhorse Op Amps) are biased at idle at around 3.5mA per chip, so , say, 1.8mA per individual Op Amp .

If supplying less than 1.8 mA peak into the load, they are still inside the Class A area.

Let's see, and remember these are preamps, here they are not driving speakers, headphones, reverb coils, or 600 ohm loads.

Out of those 4 Op Amps the most heavily loaded is the first one, which "sees" basically the 4K7 feedback resistor.

Suppose it's heavily overdriven by an active guitar or some boost pedal, in any case it's output won't swing beyond +/- 4.5 V peak, and in practice not beyond +/- 3 V peak (because this is not a rail to rail Op Amp).

So Peak current into the load will not go beyond 3000/4700=0.63 mA , solidly within the 1.8mA Class A area.

So although, yes, you are right, it does have a Class AB output stage, yet it works as Class A almost always, except in the 3 special cases I mentioned above.

Or to put it another way, for 99% of preamps (a distortion pedal is just one particular case of that) current consumption does not vary with signal.

I guess that's the meaning of what Enzo posted.

[Enzo]

Thanks. That was my basic intent, though I had a couple things wrong internally.



If it sounds OK, then fine, move on.

But if you care still.... when you removed or installed your other caps, did you parallel them on the others? Or did you connect them right at the IC power pins? Where you put them matters. The caps I'd have the most concern over would be the small ones, the 0.1uf or whatever. No, I don;t expect 1000uf filters to act any different from 100uf ones.

When you were trying it all out, did you use a variety of amplifiers? Did you feed the thing from a guitar, and from a Wah, and from some other effect? Sometimes combinations matter.