Is there any noticeable difference among silicon diodes?

[plexi12000]

i've seen a B.M.P. schematic with 914 diodes...and 4148's.   i think the 4148's were more the 'standard' diode for B.M.P.'s?

is there any noticeable difference....or is it a "not a big deal" thing?  Thanks

[samhay]

4148s are the new 914s. As far as I know there is essentailly no functional difference between them*.

Edit - * for typical stompbox use.

[~arph]

+1

BUT, there are different types of diodes (silicon) regular, schottkeys and zeners..
zeners might not be replacements, but schottkeys are interesting as they have a lower forward voltage compared to the regular ones (0.3V vs 0.7V) they will clip sooner.

[Mark Hammer]

I suppose there will be differences between diodes, else why would there be different numbers.  But there is a distinction to be made between differences related to things well outside what is relevant to pedals (e.g., high-speed switching), and what IS relevant to pedals.

In "our world", most of the time the only thing we're going to care about is forward voltage.  And in that respect, a 914 = a 4148.

[R.G.]

There are micro-differences in every physical object in the universe. We'll no doubt hear from people with curves of voltage versus current for different diode types. I've seen them before.

The real problem with calling a "1N4148" different from a "1N914" or any other silicon diode lies in the philosophical question of what a "1N4148" (or any other part number) really is.

Simply put:

AN ELECTRONIC PART NUMBER IS A DATA SHEET, NOT ANY ONE PHYSICAL PART.

A manufacturer of semiconductors cooks up a batch of diodes. He decides these will be packaged and sold as "1N4148s". He can simply sell them as that, and most people would never notice. Or he can test them. He tests them to the specified limits of what HE publishes as HIS datasheet for 1N4148s.  If they meet all the minimums and maximums on the datasheet, then HE is not lying when he sells them as HIS 1N4148s. They meet HIS promises for the part number. Notice that other makers may or may not use the same datasheets for their "1N4148s".

You see the problem?

Just to make it worse, datasheets almost never specify any number that's useful to us in making pedals sound a certain way. Every single diode coming out of a wafer-fab will have a voltage-current curve that's different from every other one, even from the same fab, even from the same WAFER. And it's that V-I curve that affects the sound. Manufacturers being the clever businessmen that they are, they will NEVER specify any detail that they cannot control well enough to keep them out of legal hot water. So they're not about to specify stuff that doesn't matter to their big customers, the electronics manufacturing giants. Certainly not for the good of the pedal making cottage industry.

So this leaves us with the quandary - 1N4148 verus 1N914? That's a valid question, all right. So is 1N4148 verus 1N4148 from a different manufacturer in a different semiconductor oven on a different date on a different wafer.

[R.G.]

Oh, sure Mark. Say it more simply and directly while I'm busy typing!!

[teemuk]

Like with ALL diodes, I expect there to be a great degree of variation in "knee characteristics" and slight degrees of variation in actual forward voltages. These are of course entirely irrelevant characteristics in bulk of rectifier applications, but quite important characteristics when low voltage audio signals are deliberately clipped to produce "musical distortion". (Unfortunately this is a total niche application in the great scheme of things so you hardly find any relevant graphs from the datasheets or whitepapers discussing research concerning the matter).

But I like this graph.

For example, 1N34A and AA112 are both Germanium diodes. BAT33 and 1N5711 are both Schotkky diodes. Unfortunately, no other "silicon" curves plotted but 1N914. Anyway, other diode types had plenty of "margin". I would expect silicon diodes to be no different.

...So, I wouldn't be surprised that if you listen very, very carefully you hear some differences. At least the high power diodes of integrated bridge rectifiers are said to have somewhat different characteristics from low power high speed switching diodes (e.g. 1N4148).

So yes, there are noticeable differences. But they are not noticeable in a very obvious way. You really have to focus and listen. ...So in the whole it's not all that important. Plus you can shape forward characteristics of any generic diode with carefully thought out series and parallel resistance values.

[karbomusic]

The real problem with calling a "1N4148" different from a "1N914"

I'm getting a difference of 3234 here. 

[Mark Hammer]

I think it also bears noting that even where such diode differences in what happens to a wave on a scope are real and replicable, once you consider that diodes will generally be used by us in a context where amplifiers, speakers, and basilar membranes will be pushed hard, whatever differences are a result of diodes will be lost in an ocean of other artifacts.

I mean, once you make a sauce with ghost peppers, does it really matter whether you used table salt, kosher salt, sea salt, fleur de sel, or himalayan salt?  No one will be physiologically able to tell the difference.

[stallik]

I mean, once you make a sauce with ghost peppers, does it really matter whether you used table salt, kosher salt, sea salt, fleur de sel, or himalayan salt?  No one will be physiologically able to tell the difference.

Unless the salt was the thing that made the peppers hot in the first place. If you're running into an already overdriven amp you may not hear any difference but if the diodes were creating the clipping and that signal was boosted into the amp then perhaps you would?

[R.G.]

The music world seems to be a three-way race between the princess and the pea, the blind men and the elephant, and the emperor's new clothes.

[Groovenut]

I did a quick setup up of diode sets for comparison. I used 7.5 VAC RMS for the signal and snapped some pics of the waveform.

1N4148 Tayda



1N914 Tayda



1N4002 Tayda



1N4937 Tayda



1N914 All Electronics



The setup

[Mark Hammer]

The music world seems to be a three-way race between the princess and the pea, the blind men and the elephant, and the emperor's new clothes.

I guess the winner depends on who you'd rather see naked: the emperor, the blind men, the elephant, or the princess. 

[Resynthesis]

Remind me, which diodes are in the Klon? 

[mremic01]

4148s are the new 914s. As far as I know there is essentailly no functional difference between them*.

Edit - * for typical stompbox use.

My understanding is that the 1N914 replaced the 1N4148, and current production 1N4148s are actually just 1N914s that are 'mislabeled' to fill demand for both parts.

Remind me, which diodes are in the Klon? 

It's a secret.

[mac]

I did a quick setup up of diode sets for comparison. I used 7.5 VAC RMS for the signal and snapped some pics of the waveform.

Do you have some BD237/175 or 2N3055 that have high internal capacitance to compare to low output capacitance like MPSA42 or 2N2369 of similar gain, wired as diodes?
As transistor they sound different, maybe as diodes you can hear a difference...

mac

[PRR]

> My understanding is that the 1N914 replaced the 1N4148

No.

First the "1N" number suggests the '4148 was registered later (though this is not conclusive).

FWIW, Wikipedia sez "1N4148 replaced the older 1N914,which had a 200 times greater leakage current".
https://en.wikipedia.org/wiki/1N4148

Finally, my dim memory is that "the" diode was '914, the '4148 snuck in much later.

[R.G.]

The 1N914 was the end product of using gold doping of silicon junctions to speed up recombination and shut off of the junction when voltage reversed. It got its high turn off speed this way. Also it's much higher leakage than ordinary silicon signal diodes. This was a matter of some high importance in the day of the 914, as it was used as a component of discrete-component logic circuits.

Better silicon processing speeded up turn off in junctions without the leakage (and expense!) of gold doping. The end result of that line of evolution was the 1N4148 and its ilk.

I researched this back when I was developing the Millenium Bypass, which depends on the near constant reverse leakage of diodes like the 914.

[antonis]

I did a quick setup up of diode sets for comparison. I used 7.5 VAC RMS for the signal and snapped some pics of the waveform.

1N4148 Tayda



1N4002 Tayda

I thought that 1N4148 was "faster" than 1N400X series...
(unless you've altered measurement frequency..)

[plexi12000]

haha--- you guys are a riot!   WOW-- thanks so much.  Great info and commentts.  i like how R.G. said it.....all physical objects have differences.  i can understand those "broad" references better!  haha

Man.....wish i could knew how to work an o-scope.  anyone want to give some lessons?? LOL