The dreaded diode debate

I'd like to see someone take a noninverting op amp gain stage, test various clipping diodes in it at various amounts of gain, and use an oscilloscope to view the waveform. Would a red LED have a similar waveform to a set of 1n4148s (to match the same forward voltage)? Same thing with 1n4148 and germanium diodes like 1n34a. Plus take into account that different diodes start to conduct at different voltages and hard clip at others. I think there would probably be at least a slight difference in the wave form.

Also with op amps I bet the difference mostly be the input impedance of the op amp when you use diodes in the feedback loop. If you're using diodes at the output, you're going to hear the op amp itself distort before it hits the diodes and gets clamped. This is going to be at least somewhat if not very noticeably different between certain op amps. Try a TL071, UA741, and OP07 in a Rat or a DOD250 style pedal to see what I mean.
 
I'd like to see someone take a noninverting op amp gain stage, test various clipping diodes in it at various amounts of gain, and use an oscilloscope to view the waveform. Would a red LED have a similar waveform to a set of 1n4148s (to match the same forward voltage)? Same thing with 1n4148 and germanium diodes like 1n34a. Plus take into account that different diodes start to conduct at different voltages and hard clip at others. I think there would probably be at least a slight difference in the wave form.


I"m not sure if this is exactly what you're looking for but if you search DIYS and FS there are a *LOT* of threads about diode clipping.
 
As was mentioned, a TS is not a great test bed for op amp comparisons.
I spent a lot of time trying various op-amp on this circuit :
Screenshot 2023-03-16 at 22-05-22 Microsoft Word - GOD CITY INSTRUMENTS – Bass Brutalist VB1.2...png
Screenshot 2023-03-16 at 22-19-23 Microsoft Word - GOD CITY INSTRUMENTS – Bass Brutalist VB1.2...png
I still try some op-amps combinations with it from time to time, just to see how good it will be.

It's a variant of a variant of the Providence Stampede SDT-1.

In the first half of the gain control rotation, the single op-amp is more audible, the second half is where the dual op-amp kicks in... or so it sounds.

For me, it is impossible to really feel the various op-amp flavors if the sound samples are separated by even the smallest amount of time. I have to really listen to the two different sounds I am comparing, by putting them side by side. In other words, I won't bet on my memory to notice the changes, it's easier if I can "zoom in" on the contrast, and feel it as directly as possible :

I record a first sample on my RC-3 loop station, I stop the sound sample, and leave the RC-3 recording a bit of silence. Then I restart the loop, wait for the first sample to end, and start recording the second sound sample right when the first sample stops.

That way, I am able to really hear it. Otherwise, if I start joking around or just wait a bit between the two sound samples, contrasts are getting much more difficult to notice.

I do the same thing when testing various diodes settings, when the changes are subtles.

I don't care if "most people" wouldn't notice anything. I mean, before l learned how to play guitar, I didn't care how a guitar player would move his fingers. I mostly looked at the hand playing the chords rather than the hand picking the strings. Nevertheless, I enjoyed the sound, without being interested by how it was created, much less noticing the picking techniques. Then, I didn't enjoyed music as much as I do, once I started to play myself.
My tastes in music changed also accordingly, to the point where Iooking back today, I'd say I didn't *really* enjoyed music before playing some myself. Same goes for electronics, I guess ?
 
Last edited:
That's how I need comparisons, too.

People in yooboob vids that compare stuff while playing different licks/different guitars etc... or worse, talk for two minutes between the two clips to be compared...

Looper, and change between the two pedals/settings or whatever at say... 3-second intervals, and making that change back and forth a few times.



As others have mentioned, it might not make much of a difference in a live-band context — but where do you draw the line?
Badly intonated cheap guitar with cheap pickups and dead strings feeding into a poor quality cable that has loads of impedance into an anemic amp through low-quality speakers BUT I've got those magical Ge Diodes in my real-deal Klon Centaur.

No, I want all my gear to be as good as possible because — also as others have said — all those fine details sum up.
My playing is bad enough, I've got to compensate somehow somewhere...


Whatever, I've posted in this thread at least three times too many, if not four.
 
I can definitely definitely tell a difference between bat41, 1n34, and LEDs, and schottkeys (if I know which one I’m putting in lol) and A/B-ing them, but in a band setting and through different amps and tone adjustments it just sounds like distortion to me. Recorded in a studio is a different situation.
 
so how about that diode arrangement on the mesa boogie grid slammer (ts9 clone but with different clippers):
43E66D03-D687-4C64-A530-B25CDB2EFCA2.jpeg
seen and done the ol pair of x2 diodes-in-series (x4 total) plenty, but not this style…(apparently the timmy does this too).

so i gave it a go… made a clipping switch exactly as above.

not sure how it really differs from pairs of x2 in series.. 🤔

and tbh i couldn't hear much difference between GS vs TS9 either ..
if anything, GS had more 'punch', 'sustain' (gain?) at the cost of a very slightly different EQ response (flatter? idk)
 
I can't see how that would be different than two pairs of diodes in series. On one side of the wave, only D4 and D5 would be on, while on the other side, only D6 and D3.
ahhh well that explains it, thanks!

so really it’s kinda pointless.
makes me wonder why mesa or the timmy designer would bother with this added complexity for no real benefit.
or maybe it’s a pcb layout thing?
 
ahhh well that explains it, thanks!

so really it’s kinda pointless.
makes me wonder why mesa or the timmy designer would bother with this added complexity for no real benefit.
or maybe it’s a pcb layout thing?
I keep wondering if it has something to do with diode reverse leakage or something like that but I’m not versed in these devices enough to have a good understanding. Maybe others on the forum know better.
 
I just built an MDMA that has a toggle for opamp clipping, then soft clip diodes with the opamp, then to hard clip diodes with the opamp.
I modified the circuit so that the op amp only clipping was only slighty more volume and gain than the soft clipping. Also the hard clipping was made louder than both of them.
When playing, you have to play in one setting for a while, then flip the toggle, to hear the subtle changes.
What I notice is that the soft clipping sounds tighter and smoother. Perhaps more midrange. The opamp only setting is seemingly more like hard clipping but the dynamics and harmonics are more noticeable. The hardclipping setting is not as harsh or raw as the opamp only clipping. It is similar but sounds more controlled. All of this changes depending on how much "gain" there is.
I can hear a difference, it is subtle though.
I doubt anyone at a gig would hear it but me and it would be more about harmonics and dynamics than anything else.
I know this is not a "hard fact" helpful post, just my observation.
If you can't here a difference, not sure what to tell you. I guess at that point it doesn't matter. Just crank it up and play.
 
so how about that diode arrangement on the mesa boogie grid slammer (ts9 clone but with different clippers):
View attachment 44233
seen and done the ol pair of x2 diodes-in-series (x4 total) plenty, but not this style…(apparently the timmy does this too).

so i gave it a go… made a clipping switch exactly as above.

not sure how it really differs from pairs of x2 in series.. 🤔

and tbh i couldn't hear much difference between GS vs TS9 either ..
if anything, GS had more 'punch', 'sustain' (gain?) at the cost of a very slightly different EQ response (flatter? idk)
the difference is that there's never really 2 diodes in series with matching antiparallel diodes in this arrangement.
let's follow the signal from sw1 ... signal passes to D6 and then passes in more or less equal measure to D3 and D4 and gets blocked by D5 ... because this is alternating current, there's also the signal from sw2 that passes through D5 then D4 and D3 equally and gets blocked by D6
 
the difference is that there's never really 2 diodes in series with matching antiparallel diodes in this arrangement.
let's follow the signal from sw1 ... signal passes to D6 and then passes in more or less equal measure to D3 and D4 and gets blocked by D5 ... because this is alternating current, there's also the signal from sw2 that passes through D5 then D4 and D3 equally and gets blocked by D6
thinking about and visualizing your description here, it almost seems as if this could be pseudo-asymmetrical
 
the difference is that there's never really 2 diodes in series with matching antiparallel diodes in this arrangement.
let's follow the signal from sw1 ... signal passes to D6 and then passes in more or less equal measure to D3 and D4 and gets blocked by D5 ... because this is alternating current, there's also the signal from sw2 that passes through D5 then D4 and D3 equally and gets blocked by D6
In your first example, if current flows through D6 and then D3, it can never flow through D4 because D4 would not have a positive potential across so it’s by definition in cutoff. Or am I missing something?
 
Back
Top