How Much is REALLY Lost in SMT Conversion?

[Anchoret]

There are more qualified participants in this forum, so I'll ask here:

When the same stomp circuit is converted to SMT, how much -- if any -- performance, "tone," etc. is lost?

Where?

Why?

Thanks for any qualified input on this. I've seen many arguments and a lot of contradictory assertions (sometimes from the same people ) on other forums, mostly from people who obviously don't know what they're talking about.

[the_utp]

I think it would depend on whether there's a one-to-one match for the components in question. If you can find the exact same component (i.e., same exact specs) in both through-hole and surface-mount versions, I'd say the difference would be minimal and not noticeable. The only difference would be how the pins come out. As I understand it, it's usually pretty easy to find simple components (i.e. resistors, simple transistors, etc) in one-to-one matches for both formats. It's the complicated ICs where things get more tricky.

However, I should preface this with the fact that I can't hear the difference between metal film resistors and carbon film resistors or most brands of transistor. I can hear the difference between silicon vs. germanium, etc, but not between most brands of the transistor with the same specs. I've used both the good stuff and the cheap stuff, component-wise, and not heard the difference. I use the good stuff when I can, for durability, but mostly I save my money for a good switch, good jacks, good pots and a good enclosure because I think that has more end effect on tone and durability.

[wnorcott]

There is not going to be any difference. Most integrated circuits are made both in SMT and DIP8 packages. The parts are identical. For a stompbox we are talking milliamps of power here.

SMT is more the domain of big electronics manufacturers not the hobbyist or boutique builder.

It is a whole lot harder to build using SMT because the parts are so tiny to work with and you need a different type of soldering station. And most of all you need much better lighting and better eyesight.

Assuming you are good with a soldering iron there is no advantage either way.

Bill

[11 Gauge]

Originally Posted by wnorcott

SMT is more the domain of big electronics manufacturers not the hobbyist or boutique builder.

Not true any longer.

Way Huge Electronics and Lovepedal (amongst others) use surface mount on their stuff.

And while it's a gray area between mass produced and bootweak-ish, all of the Tech 21 (SansAmp, etc.) stuff is surface mount as well.

And I guess that since Cusack builds the pedals for Lovepedal that their designs must be surface mount as well.

Folks get so caught up in the componentry that they begin to fail to see the forest for the trees. IMO, components are really over hyped. As long as the value of one is close to that of another, they should sound similar, especially in a low powered device.

[callaway]

You don't need a "different type of soldering station". Maybe you might want a smaller tip for certain parts, but that's it.

[Anchoret]

I'm still not seeing the real answer I'm after here.

It is **constantly** asserted by effects buffs that the unmeasured and unrated secondary characteristics of passive components affect the sound of low-V effects.

If that's so, and there's a big issue between, say, a carbon and foil 50K full-size resistor, or a .047 mFd capacitor's composition, etc., then obviously there's going to be some pretty substantial changes going on tonally with the use of SMT passives.

I had one "expert" on TalkBass insist that there was no difference between through-board and SMT circuits and then in a post a few days later commences to cork-sniff passive components in through-board builds.

That's BS; he can't have it both ways.

So...which position is wrong?

[callaway]

I think this is just one of the things you'll have to decide for yourself. Like you say, it can be a heated issue. I've played several effects with surface mounted components and they sounded great. Of course, as an electrical engineer, I have no problem saying surface mount components are fine... although there are some sectors of electronics that are still largely reliant on through-hole (such as space applications) for their increased reliability of the solder joint compared to SMT. But that is the only reason for them.

The human ear is a remarkably sensitive instrument, capable of detecting very slight differences that are mathematically hard to quantify. Of course the ear is also susceptible to biasing from the brain when it tells the ear that something should sound better or worse, you will often hear it that way.

So, I think you're going to have to decide for yourself. If you don't want to spend the effort and money to experiment yourself, you can go to a shop and play some of Electro-Harmonix's new "nano" effects series. They are all SMT versions of their pedals, so you may even be able to compare with the full size side-by-side.

[Stevil]

I've never A/B'd any of that sort of stuff but my guess would be people arguing over changes in sound due to caps in the signal path. I've heard it stated that large voltage film caps sound "better" than smaller voltage ones for input/output/etc. caps. Large voltage film caps are not tiny things, and if they have managed to make smt versions of those, I'm sure they're very different in composition indeed. I think the only way anyone could truly say one way or the other is to breadboard something with smt caps and regular large film caps set up to be switched back and forth between the two. My guess is few have actually done that and are going by theory alone. I've heard the guys who use theory alone say that different types of transistors don't sound different in a Muff if they have the same gain; I have personally tried 7 different types, measuring the gain of each and can tell you that there is a world of difference between the sound of one built with a 2n5088 transistor and others of the same gain. I'm not an EE, but I could absolutely hear a difference (and was actually one who thought it was all BS to start with) and in the end that trumps anything you'll find on paper.

[wnorcott]

It would be very simple to put this to an empirical A/B test in a DIY pedal. Build an effects pedal designed for a DIP8 and use an IC socket. Now get two of the same identical IC from the same manufacturer, one of them a DIP8 package and the other SMT. Solder the SMT to a DIP8 adapter. Now swap the two IC's back and forth and tell me if you can hear a difference.

Better yet record two identical clips and post it as a blindfold test.

The world is waiting.

Bill

[Anchoret]

Originally Posted by callaway

I think this is just one of the things you'll have to decide for yourself.

No. No. No.

This is a technological question with a real, physical answer.

It's just that consumer-grade fora are so informationally poor that it's apparently impossible to get it, and the gear industry is jealously tighter with their data and try harder to keep the rubes in the dark about EVERYTHING than even the government.

Instead of real information, we get marketing puffery to manipulate our perception of relative awesomeness.

Like you say, it can be a heated issue.

Among ignorant people who are merely guessing in the dark, yes.

I've played several effects with surface mounted components and they sounded great.

Sure, there is no doubt about this, but what and how much compensation for SMT was factored into the design of, say, an adaptation of a through-board circuit?

When SansAmp went SMT, was the board an exact copy of the through-board circuit? If not, what changed? If so, did the A/B comparison of the prototypes show any measurable performance difference? Where? Why?

There are hard answers to these questions. I'd love to see them.

The human ear is a remarkably sensitive instrument, capable of detecting very slight differences that are mathematically hard to quantify. Of course the ear is also susceptible to biasing from the brain when it tells the ear that something should sound better or worse, you will often hear it that way.

I trust nobody's ears, certainly not my own.

Over the decades, I've seen so many demonstrations of the utter unreliability of people's perception of sound that I'm totally dismissive of it except in extended blind testing with solid experimental protocols.

you can go to a shop and play some of Electro-Harmonix's new "nano" effects series. They are all SMT versions of their pedals, so you may even be able to compare with the full size side-by-side.

But -- again -- are they exact copies of the previous circuits, or are they adaptations? Unless the former, the comparison is meaningless.

Then we get to the question of component consistency in secondary, unrated properties that reflect in the sonic performance of a circuit, which matters hugely in effects (notoriously in old OpAmps and germanium components). Is that better/worse with SMT passive and active components? If there's a difference in sound, is it due to the innate differences in SMT circuitry or merely cumulative error?

Sure, there are a lot of questions there, but they actually mean something.

Answering them once and for all would give a more meaningful understanding of the situation than the typical gristlehead stoner's bong-influenced Perception of Relative Awesomeness.

AND save a lot of bandwidth.

[hou35male]

What's the big deal? I think that YOU should try and test a schematic with thru-hole components and SMT components and YOU TELL US THE DIFFERENCE(if any)!!!!!

[Anchoret]

Originally Posted by hou35male

What's the big deal? I think that YOU should try and test a schematic with thru-hole components and SMT components and YOU TELL US THE DIFFERENCE(if any)!!!!!

It's not my job.

There are a lot of people whose job it is.

Those are the people to whose information on the subject I want access.

[callaway]

Originally Posted by Anchoret

It's not my job.

There are a lot of people whose job it is.

Those are the people to whose information on the subject I want access.

Dude, I hope you were being sarcastic, because that was pretty insane. You make it sound like some big conspiracy or cover-up or something. Um, it's not! You can look up the datasheets to every component in every pedal. The bottom line is that there are a lot of intangibles that are impossible to describe in a datasheet. You wouldn't believe some of the things I've seen happen in my field (practicing electrical engineer in military/aerospace field with a master's) due to changes in semiconductor fabrication that impact characteristics no one would expect to be in a datasheet.

I had a ton more written here, but it doesn't matter. When you look up a datasheet and it shows different packaging options (DIP-8, SO-8, uSOP-8), they all use the exact same semiconductor die. If you're to the point of obsessing about what difference the package makes, I think you're not just playing guitar enough. I could mathematically prove that parasitic capacitance, inductance, coupling, ESR, etc., are different. But it's on such a miniscule scale that it doesn't matter.

I guess I'll draw a line if you really want me to. I guarantee you that if you buy identical parts in through-hole and surface-mount packages and do an intelligent layout for each board, you will not tell the difference in a blind test. (By the way, surface-mount electrolytics are not very common. Many applications still use through-hole for electrolytics.) Any differences in sound that someone has attributed to through-hole vs. surface-mount are almost certainly due to comparing a vintage effect (or even something 10 or 20 years old) with a modern version that happens to use surface-mount components. In that case, I guarantee you that semiconductor die are vastly different between the parts, thus accounting for the difference.

Either way, I guarantee you will get better tone by practicing instead of fretting about this issue.

[Anchoret]

Originally Posted by callaway

Dude, I hope you were being sarcastic, because that was pretty insane. You make it sound like some big conspiracy or cover-up or something. Um, it's not! You can look up the datasheets to every component in every pedal. The bottom line is that there are a lot of intangibles that are impossible to describe in a datasheet.

You've just pointed out the problem in the same paragraph.

The datasheets are useless for determining the unrated secondary audio characteristics that determine how these products will sound, or the consistency of those unrated characteristics (though for specialized hi-end audio components like Burr-Brown OpAmps it may be).

And there is a "conspiracy" (as you call it) with manufacturers compulsively imposing total non-disclosure restrictions on their employees to NEVER discuss the technical details of their products (ironically, it was once my job to police and enforce this at Intel Corporation).

I'd love to talk to someone at Harman and ask, "Hey, when you guys converted the [cough!] fabulous DOD Grunge over to DigiTech SMT, how was that? Did you have to materially change the circuit? What problems were there in the changeover and how did you address them? Why do the new ones sound different? What's in that new speaker simulation? Can I see a schematic?"

You'll get dead silence. You can absolutely hear the sphincters tightening.

I went through this with some guys at Korg not long back.

So, don't tell me that the information I seek is not being withheld. It is, which is doubly ridiculous when you consider that their actual competition knows exactly what they're doing.

I guess I'll draw a line if you really want me to. I guarantee you that if you buy identical parts in through-hole and surface-mount packages and do an intelligent layout for each board, you will not tell the difference in a blind test. (By the way, surface-mount electrolytics are not very common.)

Maybe not in analog stomps, but I see a lot of them in bigger digital effects these days.

Any differences in sound that someone has attributed to through-hole vs. surface-mount are almost certainly due to comparing a vintage effect (or even something 10 or 20 years old) with a modern version that happens to use surface-mount components.

OK, so that makes you also on-record as saying all the angst and cork-sniffing over low-V through-board passives is also hogwash?

Hi-V audio applications are of course another matter.

In that case, I guarantee you that semiconductor die are vastly different between the parts, thus accounting for the difference.

I know -- classic case being the totally inconsistent JRC4558D v. the current retro-labeled JRC4558D which is constantly being pimped as "the same as the Tubescreamer's!!!" which of course it absolutely is not, the tooling for the original (crappy) chip having been scrapped by NJR a couple of decades ago.

A significant assertion I have also not been able to confirm is whether current DIP or SM OpAmp chips are now merely the same wafers in different packages.

Either way, I guarantee you will get better tone by practicing instead of fretting about this issue.

Hey, well thanks for that advice, huh?

This isn't about my own personal tone-quest, but about confirming or dismissing once and for all all the yammering tavern-chat among gristlehead gear users about SMT effects being innately inferior to the old stuff in terms of tone.

To the extent there's anything to it, I want to know what it is, where it happens and how manufacturers compensate for it. If it's total nonsense, I want to be able to say so so they'll shut up and sit down so we can move on to more edifying subjects.

This is not unknown information, it's just not accessible to consumers.

[Stevil]

Originally Posted by Anchoret

This is not unknown information, it's just not accessible to consumers.

A circuits growing pains might not be but this originally started as curiosity about smt vs. through-hole...and that could be tested with minimal effort in a day, by ear. If you want to get fancy you could mic it and analyze it, but if you or someone with a good ear can't hear a difference, analysis via computer is useless.

[Ben Harmless]

I really don't understand what we're hoping to achieve here if the foundation of the question is that documentation of the characteristics of interest are unavailable.

Are we hoping that the manufacturers are going to pop in and tell us that they have documented all the minor, tertiary characteristics of their components which absolutely none of their customers have ever questioned?

And why? So we can clone something, and sound exactly like a guitarist who's been dead for 30 years?

Personally, I'm not worried about it. When I built my Marshall clone, I didn't bother figuring out the composition of each component, or worrying about the undocumented characteristics. Why? Because Jim Marshall didn't do too much worrying about that either. A lot of vintage amps of the same model sound different from one another. That's the nature of the beast, and we embrace it.

If I select a component that doesn't sound good, I'll change it.

[Montana_Dawg]

SMT vs PTH

Simple, footprint is smaller, components are cheaper, quality of most SMT parts is on par with the better PTH stuff, and it does cut down on noise issues.

With smaller and shorter traces, and the fact that the components are not standing on end to conserve space (antennae anyone?), the issue of spurious noise is significantly negated, and there is more space in the box.

I am currently working on a layout that is SMT, with Molex connectors that get rid of wire to board soldering. Also, I still prefer 4-layer boards with the power and ground planes internal. Makes for a compact, neat, quiet, and very efficient device.

Summary: If you can do it, then go for it. SMT is every bit as good, if not better than, the PTH stuff. It is a challenge for novices, but a good one!

[Montana_Dawg]

Originally Posted by Anchoret

This isn't about my own personal tone-quest, but about confirming or dismissing once and for all all the yammering tavern-chat among gristlehead gear users about SMT effects being innately inferior to the old stuff in terms of tone.

To the extent there's anything to it, I want to know what it is, where it happens and how manufacturers compensate for it. If it's total nonsense, I want to be able to say so so they'll shut up and sit down so we can move on to more edifying subjects.

This is not unknown information, it's just not accessible to consumers.

Doesn't matter what the facts are. People will believe what they will, no matter the truth.

There is no conspiracy. The information is out there. For audio related info considering stomp pedals, well, the lack of such info is not due to a conspiracy, but due to the fact that the main components weren't designed for those types of uses.

As a consumer, I can get whatever information I need. Being in the industry for 30 years just makes it easier to know where to go for the info.


But to answer your question: SMT is NOT inferior to PTH, or vice versa. The whole question of tone, once and for all, is TOTALLY SUBJECTIVE. Whatever works for the individual is the best choice. Period, full stop, end of the line.

I just can't understand why people can't get that through their heads. Of course, manufacturers and boutique builders depend on the ignorance of the masses to make their coin.

It amazes me to no end how much money people will throw away just because of something someone said or heard. Like PT Barnum once said," There's a sucker born every minute."

[Stevil]

Originally Posted by Montana_Dawg

components are not standing on end to conserve space (antennae anyone?), the issue of spurious noise is significantly negated

What exactly are you tuning in using a standing resistor as an antennae vs. a resistor that isn't standing?

[Montana_Dawg]

Originally Posted by Stevil

What exactly are you tuning in using a standing resistor as an antennae vs. a resistor that isn't standing?

Resistors with capacitors form an RC circuit, which sensitive to frequency changes, and noise.

[Stevil]

I realize you'll find RC circuits (filters, etc.) in almost every stompbox, I was just wondering what you were tuning in when you stand a resistor on end to save space vs. having it laying down.