SMT vs. Through hole soldered components

[Microtone]

Hi,

I've posted a couple of times about SMT boards and have a new puzzle....

I recently ordered a prototype board made with SMT (naturally using SMT mini components), based on a simple overdrive circuit that I have previously made using a through-hole PCB and soldering the individual (normal-sized) components onto it.

The SMT board arrived and works and sounds just like the through-hole home-soldered version. However, its output is way lower and I'm not even sure that it achieves unity gain. (Not sure how to measure this accurately apart from by ear through an amp).

I realise I am again jumping ahead of myself and dabbling in voodoo that I don't fully understand.

However, I wanted to ask if there is maybe an obvious reason for the lower volume coming out of the SMT-based pedal that I am not realising. The components and component values are identical. i.e. 2N3904 transistor, same-name schottky diodes, resistor and capacitors with identical values etc. I haven't really paid attention to the stated wattage and voltage values of the SMT components, but I think they're all over 0.25 W and at least over 9V, as this is the power supply I'm using.

Any ideas?

[Mark Hammer]

Tell us what the circuit is, and that might prompt some usable suggestions.

[Microtone]

Tell us what the circuit is, and that might prompt some usable suggestions.

Thanks for the reply. Well, like I said, it's a very basic single transistor overdrive with two soft clipping diodes and a bunch of resistors and capacitors.
However, my question is more a meta one:
I have produced two pedals containing the same identical circuit. I.e. all component order and values are identical.The only difference is that one is normal size components soldered on a through hole PCB and the other is mini SMT components surface mounted. The SMT version is significantly quieter at full volume than the through hole PCB version.
I am wondering whether there could be a reason for this that is unconnected to the specifics of the circuit itself.

[blackieNYC]

Do you have a voltmeter?  You could compare the two units point by point, no?
There is no reason they should be different.  Surface mount should be just as loud as thru-hole😃

[R.G.]

I am wondering whether there could be a reason for this that is unconnected to the specifics of the circuit itself.

No.

If it's much quieter, and it's really, truly, honestly and in fact the same part values, connected the same way, and the transistor (which is the big hitter in this mystery circuit) is similar to the through hole ones, it'll be a similar volume.  It is very, very likely that there is a component value off or an unexpected short or open, something like that. Electrically, SMD/SMT parts do work like their bigger brothers. Otherwise, the industry would have abandoned this technology a long time ago.  Tone gourmets will hear tiny, tiny differences in EVERYTHING, but pretty much a 1K is a 1K.

[amptramp]

There is also the possibility that if you are using older transistors for the through-hole version that you may have counterfeit parts.  I don't think counterfeit surface-mount parts are as common although it would not be difficult to pass off a fake surface-mount part as few people can read any markings on them.  Or the label on the SMD reel may say something different from what the devices say.

I once amazed a manufacturing manager when I found a roll of surface-mount resistors had different resistor values from what is on the label during an inventory.  He said it was impossible for me to read the resistors (which clearly said 6141 on them or 6140 ohms.)  The label on the reel had a different number.  He put a resistor under the microscope.  He looked up at me.  He looked back under the microscope.  He wondered just what kind of cyborg I was.  He eventually corrected the label.

[Phend]

^ mom always said eat your carrots .
I had the opposite problem.
A homemade circuit board using smd.
It is a simple Muff Fuzz. Very simple that is.
And it was very loud.
Until I added a resistor to tame it down.
3904 transistors. Silver epoxy traces.

[PRR]

He said it was impossible for me to read the resistors

All the books/magazines said you could not "read" a phonograph stylus. It was as plain as the nose on my face (which I could also focus). I was near-sighted enough to get over "10 power" just by holding close in a bright light. Working without a magnifier made me more nimble at turning the point this way or that to study the glint. Maintaining recordplayers for radio and classrooms, this was valuable.

As I got past 25 those needles got fuzzy, even tho I could read a (tiny LED) calculator past 40 specless. Advancing age and toxic lifestyle led to optic overhaul, and now anything within 10 feet needs readers (I just bought a box of 10 generics, as well as several Rx readers for my residual astigmatism.)

[kenjib]

Josh Scott/JHS apparently tested this out and came to the conclusion that the SMD versions sounded identical and maybe even better/less noisy.

Myth #5:

https://youtu.be/yEZ0yzswOt4?si=gi4nHPF_Gbzvs7kC

He doesn't have the test there -- just talks about it. Maybe you have a faulty part?

[R.G.]

No test will convince a True Believer that tiny, tiny differences are not present. The hifi audio world is full of "Golden Ear" experts that concocted a whole mythos of micro differences in things like the audibility of the sound stage and the separations between the notes, attributing this to - well, remarkably nearly everything that ever changed. Like all True Believers, they only got quiet when embarrassed by inabilities to discriminate like they thought they could. Double blind tests were so effective at debunking Golden Ears that they retreated into saying that the very act of testing obscured the subtle differences they heard, then started fading away.

So yeah, some people will accept decent tone from SMDs, some will never accept that. It's still the same old story ...

[mzy12]

I have yet to see credible (read: any) evidence to suggest that SMD parts are worse sounding than their through hole counterparts. Through hole parts are great for ease of soldering, repairability, support for larger sizes (I mean large, I don't see anyone making surface mount components for an 80V 10,000uf filter cap :p) and, if it is necessary, better heat dissipation characteristics (due to their larger thermal mass, surface area and the metal legs don't hurt either). SMD parts are often cheaper, allow for more compact/optimal board design and generally have much higher availability, especially with items like JFETs that are almost impossible to find in through hole versions.

I am, of course, open to having my opinion changed with peer reviewed scientific papers on the matter =P

Don't forget part to part tolerances can play a massive part in your problem, Microtone. If you could drop your schematic/transistor used, members on this board could point you in the right direction!

[Ice-9]

The most likely issue will be an incorrectly identified resistor or capacitor value, SMD resistors will have a code on them that you can use to easily work out the value but capacitors usually have no code so the only way to be sure you have received the correct value caps is to measure them with a capacitance meter.  The other problem could be an accidental short on the PCB or a mistake.

When it comes to SMD v Through hole, given identical circuits they should sound the same. The only thig that would make a slight difference in sound is the tolerance of the components, but this applies to BOTH SMD and TH.

[PRR]

If you are doing ULTRA-low hiss circuits, ultra-small parts may hiss more.

In thousands/millions of posts I have read, only one guy quantified this.

[amptramp]

There is a practical consideration that if the board flexes for any reason, through-hole parts take it in stride whereas SMD parts will definitely suffer from solder cracking except for some gullwing parts that have a natural flexibility.  Large items are worse off than small items.  An even worse problem is flexprint with surface-mount parts on them.  We had a number of Analogic digital panel meters become erratic or quit entirely after a short time from this.

SMD solved a problem for manufacturers - the drills used to drill circuit board holes go dull rapidly when epoxy-glass boards are used and anything that can reduce the number of holes makes the board cheaper.  Boards are quoted with a large portion of the cost determined by the number of holes of various sizes.