Powering a guitar pedal

[Hugh Robjohns]

I'm surprised this is even worth discussing.

The answer is yes, of course different batteries and PSUs can make a difference — but only in very poorly designed FX pedals which allow the audio to degrade as the power rail voltage declines and with higher voltage source impedances.

But it's not a musical feature, it's a mark of shoddy design and bad engineering!

It's very easy to design a pedal's audio circuitry to perform perfectly over a wide range of input power voltages and supply source impedances. However, it requires better engineering and a few more (and more expensive) components. Sadly, few guitar FX pedals are designed properly in my experience....

The problem comes if the user likes the sound quality/character achieved with a half-dead battery in a naff FX pedal design, because maintaining that depleted voltage and source impedance consistently is very difficult with real batteries, and potentially expensive with an external 'tunable' PSU.

[Martin Walker]

If you were in a session would you take that chance that your battery will completely die in the middle of your session, and/or will the lower voltage harm the pedal's circuitry? I know that too much voltage is a problem, but is not enough voltage also an issue?

No, running an analogue guitar pedal on a lower than specified voltage won't do any harm at all. There is I suppose a possibility that pedals incorporating digital circuitry might take exception to this, but I still doubt that any damage would ensue.

[Drew Stephenson]

If I had a pedal that I thought sounded best at a lower voltage I'd just buy a power supply that had a lower voltage setting.

[Martin Walker]

...or buy (or DIY) a sag function that mimics the effects of a dying battery (it's not just the lower voltage that results in the most desired 'sag' effect, but developing a higher internal resistance).

https://jhspedals.info/products/volture

[merlyn]

... But it's not a musical feature, it's a mark of shoddy design and bad engineering! ...

Just as well Jimi Hendrix didn't know too much about electronic engineering, or he might never have used a Fuzz Face.

A Fuzz Face is a pedal that would be affected by a dropping battery voltage. Introduced in 1966, it's a very simple circuit. Simple doesn't necessarily mean bad, and hundreds of records have been made with a Fuzz Face, including Eric Johnson records.

I think what Hugh is complaining about is that a Fuzz Face doesn't have a voltage regulator. The pedal Martin linked above turns that bug into a feature.

[Hugh Robjohns]

Just as well Jimi Hendrix didn't know too much about electronic engineering, or he might never have used a Fuzz Face.

To be fair, it wasn't his job to know about electronic engineering... but it was the job of its designer, and accommodating battery voltage variations is a foreseeable requirement... hence a poor design.

[ef37a]

... But it's not a musical feature, it's a mark of shoddy design and bad engineering! ...

Just as well Jimi Hendrix didn't know too much about electronic engineering, or he might never have used a Fuzz Face.

A Fuzz Face is a pedal that would be affected by a dropping battery voltage. Introduced in 1966, it's a very simple circuit. Simple doesn't necessarily mean bad, and hundreds of records have been made with a Fuzz Face, including Eric Johnson records.

I think what Hugh is complaining about is that a Fuzz Face doesn't have a voltage regulator. The pedal Martin linked above turns that bug into a feature.

These pedals were, as you say very simple. The early ones usede Germanium transistors, notoriously variable in parameters, they were also "True Bypass" i.e. there was no buffer stage so when not "effecting" the guitar was connected thru'.
I have no quarrel with that, way things were done and if it got them the effect they wanted, fair enough. What I am very cheesed off about is the fact that True Bypass has been elevated to a desirable even must have quality of a guitar effect. THAT is bad engineering. For consistent results with various pickup and guitar wirings a high impedance input stage is needed to isolate the "effecting bits" from the guitar. Then the output resistance needs to be decently low to avoid cable capacitance shunting high frequencies.

When pedals were being knocked up in sheds as one offs and tailored to a particular guitar and player's requirements, no harm but once you put something into production you need a specification and consistency.

Dave.

[merlyn]

I thought it was all about the market and selling products. True bypass sells, and engineers are perfectly capable of delivering that.

I don't think what musicians hear should be written off. The engineering of music gear serves musicians. If a musician says they hear something, they probably do. There's no point in pulling out sophisticated test gear and telling them it's not there.

Perhaps no surprise that Eric Johnson was fussy about pedals changing his tone when bypassed. He has said that he chooses pedals based on how much they change the tone when bypassed.

Is 'bad engineering' the new phrase? It's just old engineering, and guitarists often want (i.e. buy) how it was done in the old days. The source of all the problems with toan suck, of course, is the electric guitar pickup. Bang up to date technology from the 1950s.

[Hugh Robjohns]

Is 'bad engineering' the new phrase?

No, it's a statement of fact.

I was trained to design devices that worked consistently under all conditions, and to prevent or warn of any conditions outside normal parameters. That's what I consider 'good engineering'.

We all know that battery voltage declines with use, so a battery-powered device should operate consistently over as wide a range as practicable and then cease to work. It shouldn't exhibit dramatic changes of tone as the battery is drained because that would be 'bad engineering'.

If, perhaps during development, someone discovered a cool sounding effect when the circuit operated at low Volts, then design the FX pedL to work consistently that way, by regulating the battery voltage as necessary!

[merlyn]

The three legged voltage regulator that we're familiar with didn't exist until 1969.

[ef37a]

The three legged voltage regulator that we're familiar with didn't exist until 1969.

No but we had transistors and I am not sure when zeners became readily available but they could have used the BE breakdown of a transistor for a reference (and before that valves and neons!)

I agree with Hugh, you design and make things to perform to a specc' as best you can for the cost and over as wide a supply voltage as possible. If something sounds (to someone else) better at a lower voltage then fine, put a flippin' switch on it!

Neither I, nor I am sure Hugh wishes to belittle the sensibilities of musicians. They would not be so excellent in their art if they were nor....sorry! like us ordinary folk. Has anyone here seen the guy working on piano hammers? Some get more compacted (brighter) and some pricked with a triple needle (less bright) I can't hear the bloody difference and I doubt many can but Steinway would not employ him if he was BSing them!

My beef with the TB pedals is that they are often a rip off for a very simple circuit. There is a review of a high voltage valved pedal in the current issue. A lot of circuitry and controls for £250 but others are charging a similar amount for two transistors and couple of jacks!

And don't knock "scientific testing". I know of at least three situations where a musician told me there was a fault which I could not hear* but once we got scopes and stuff in there we found the culprit. I would go so far as to say, IF someone can consistently call out a sound problem, that problem can almost always be diagnosed. Humans, on the whole are not nearly as good as an AP test set!

*Yes, I am pretty deaf and I still was 20yrs ago but at least one of the problems only occurred when a 100W EL34 amp was driven close to clipping into a 4XV30 cab....even I can hear THAT ****!

Dave.

[Hugh Robjohns]

The three legged voltage regulator that we're familiar with didn't exist until 1969.

True... but discrete regulator circuits did, some very simple and inexpensive, too.

[resistorman]

The most dissected fuzz tone has to be the riff from Norman Greenbaum's Spirit in the Sky. Allegedly it came from a half dead battery powering a DIY fuzz circuit built into his guitar.

[merlyn]

For those not too familiar with electronics this is what we're talking about; this is the circuit in a Fuzz Face:

Putting the board into circuit board. It's a board with a circuit on it. Not a lot of room for any more components to make a voltage regulator. I'm sure this was designed to a spec, and one of the specs was the cost.

A well engineered distortion pedal, eh? The first thing to do to make a distortion pedal well engineered would be to get rid of the distortion.

[ajay_m]

It's presumably because the electrons from some brands of battery have a different quantum spin as a result of differences in battery chemistry. Due to the collapse of the wavefunction and its associated eigenvector, this will result in subtly different temporal interference effects, which depend critically on the coherence of the electron source powering the circuitry.

Electrons from mains PSUs have taken a long and tortuous journey through the power distribution system before being rudely thrashed around inside a transformer and teleported across the magnetic barrier to be reincarnated on the other side. This gives them very poor temporal coherence, resulting in a degradation of the sound.

Got this far and were mentally drafting your response to my silly post? Of course I'm just making up total bollocks. But if you are non-technical, who knows, perhaps you really DO feel that battery brand affects sound. A scientific A:B test of course would sort that out.

[Kwackman]

I’m pretty sure that’s the same circuit board in my iPhone…

[ajay_m]

actually it reminds me of the circuit board you see in an episode of The Avengers where an amulet is supposed to have some fiendish power to drive domestic cats mad, and claw the victim to death!

[Hugh Robjohns]

Not a lot of room for any more components to make a voltage regulator.

I can see bags of space for a simple discrete voltage regulator... and since they designed from scratch they could have made it any size necessary to do it properly, anyway!

The simple truth is that the original designer didn't give battery voltage variation a moment's thought...

I'm sure this was designed to a spec, and one of the specs was the cost.

Maybe... but cutting out a critical function spec (consistent performance) purely to meet a non-critical spec (low cost) is, by definition, bad engineering!

I've seen it called 'value engineering'... but that's just a woke way of saying "we left out as much as we could before it stopped working completely."

The first thing to do to make a distortion pedal well engineered would be to get rid of the distortion.

Er... judging from that bonkers nonsense I'm not sure you understand what engineering is at all. You appear to have lost the plot, sir.

[ef37a]

For those not too familiar with electronics this is what we're talking about; this is the circuit in a Fuzz Face:

Putting the board into circuit board. It's a board with a circuit on it. Not a lot of room for any more components to make a voltage regulator. I'm sure this was designed to a spec, and one of the specs was the cost.

A well engineered distortion pedal, eh? The first thing to do to make a distortion pedal well engineered would be to get rid of the distortion.

Hmm, I can see Hugh chopping this thread off at the ankles before too long!
However..PLENTY of room on that PCB for a 100mA regulator. Room even for an OC81 and a 5v6 Zener. But really, if we are talking about a supply going crap then a simple resistor will suffice.

"Engineer out the distortion"? Don't you read SoS? There are scores of reviews of pre amps which deliver pristine audio quality but also have the ability to generate distortion, "mojo", whatever you want to call it.

I shall try one last time. A properly engineered DISTORTION pedal would have a high Z input buffer (the humble TL-072 say) That feeds the distortion producing stage, yer geranium transistors if you like, and that stage is then further buffered so that whatever is connected to the output will, within reason, not compromise the desired effect. This isolation means the pedal will give consistent results with practically any guitar and any sensible length of cable into almost any guitar amp. Such consistent performance is vital when you are mass producing things world wide.

I see Andertons want nearly £200 for a Fuzz face. That is a lot of money for two transistors, a switch, a case, couple of pots and two jacks IMHO.
It seems you can buy a basic kit for 30 quid?

Dave.

[Wonks]

First fuzz face: 1966
TL072 hit the market: 1978

Given the variability in germanium transistor performance in 1966, introducing voltage regulation from a pretty stable voltage source i.e. a battery, was probably not the top priority of the Fuzz Face designer.

You could argue that after the change to more stable silicon transistors, then voltage regulation could have been considered, but it probably wasn't a priority in the race to get more and more product out the door. And at the time, variation in tone between different units probably wasn't considered important. in pre-internet/YouTube times, people were lucky to have two or three pedals in their setup. Certainly not the 15-20 pedals that are common on pedalboards these days and the luxury of comparing the sounds of the same pedal with different batteries.

[ajay_m]

The traditional discrete voltage regulator uses an NPN transistor essentially as an emitter follower and with a zener providing a voltage reference to the base. However early transistor designs were based around PNP germanium devices and complementary NPN devices were initially not available. Additionally you need to allow for a fairly significant forward drop across the transistor so for a 9V source you probably would be losing a couple of volts at least across the regulator. Then you are also wasting power from the battery as well.

Finally, any active components, even diodes and certainly zener diodes, were not cheap in the sense that they individually had a more significant impact on the total BOM cost compared to nowadays.

Even in the mid 70s, Hewlett Packard calculators came with a charger that relied on the battery to act as a voltage regulator because the cost of linear regulators would have been significant. This had the unfortunate effect that even a momentary disconnection of the battery could destroy the calculator. [but, nerds being nerds, a couple of very dedicated people have created replacement circuit boards to repair these machines, which are now very collectible]

[ef37a]

The thread has morphed into a discussion about voltage regulators but that is not the effect a depleted battery would exhibit. As someone else said, the voltage would drop of course but the internal resistance will also rise, this will modify the behaviour of the simple transistor circuit because although a linear circuit will be "class A" and supply current demand virtually independent of signal level, once things get very distorted, all such bets are off.
Feeding the circuit from a very low source resistance regulator (likely lower than even a good PP3) will not produce the same audio effect AFAICS?

And regulators do not always have to 'rob' much of the supply. I found a complimentary two transistor plus zener circuit a very long time ago in Wireless World's "Circuit Ideas", it has a low differential and is one of the few regulator circuits I have tried that is truly short circuit proof. I built a box for a chap who needed to power the glow plugs in model engines from a car battery and he never managed to blow it up! The downside of the circuit is that it is not a brilliant load regulator.

Dave.

[Martin Walker]

As someone else said, the voltage would drop of course but the internal resistance will also rise...

That was me m'lud.

[merlyn]

... Maybe... but cutting out a critical function spec (consistent performance) purely to meet a non-critical spec (low cost) is, by definition, bad engineering!

Seems like a lot of people like the sound of bad engineering.

[ef37a]

... Maybe... but cutting out a critical function spec (consistent performance) purely to meet a non-critical spec (low cost) is, by definition, bad engineering!

Seems like a lot of people like the sound of bad engineering.

And they are very welcome to do so. I just don't like people getting ripped off for "two transistor trash."

Dave.