Diode clipping mod, Ge and Si in parallel?

ajm · From: Los Angeles

A question for the electronics experts.

Distortion/clipping diodes: Say I have a circuit that has two diodes in parallel, back to back.
They are both the same type, 1N4148 silicon, 0.7 v drop.
If I don't remove either and PARALLEL one of them with a germanium diode (0.3v drop), will this produce asymmetrical clipping?
In other words, I'd have three diodes in the circuit, all in parallel.
- One silicon pointing in one direction.
- Two other diodes both pointing in the opposite direction from that, a silicon in parallel with a germanium.

Why do I ask?

A lot of solid state amps have diodes used for distortion/clipping, usually in the feedback loop of an op amp and/or on the output to ground.
The diodes are usually the same type, and in this case are 1N4148 silicon (0.7 v drop).
That makes the clipping of the waveform symmetrical.

A lot of tube amps, in their clipping design, have a stage that is biased asymmetrically. (Some tube amps use clipping diodes also, but we'll leave that alone for now.)

The problem is that swapping out parts off of a circuit board can be VERY difficult.
Most of them don't have the circuit traces on the top, so you need to remove the entire board. To do this, that often means that you need to unattach the pots, jacks.....you need to literally take the entire amp apart!!!!!

To avoid this, could you "tack on" a GE diode in parallel with an existing Si diode?
I'm thinking that the GE 0.3v drop would override the Si 0.7v drop and give the asymmetrical clipping.
That way you wouldn't need to remove the circuit board.
This is often done with resistors and capacitors, so why not diodes?

Not that it matters, but the amp is an early 80's Peavey Studio Pro 40 with the Pre-Saturation-Post set of gain controls.

Thanks in advance.

Georg Sørtun · From: Mandal, Agder, Norway

Randy Schneider · From: SW New Mexico, USA

ajm,

Seems logical to me. The Ge diode would essentially short-circuit the Si diode by beginning to conduct when the voltage reached the 0.3 (approx). The Si diode in parallel with it would never conduct in this scenario.

That is the reason why you cannot use similar diodes in parallel to add current capacity to a circuit -- unless the diodes' forward voltage drop were *exactly* the same, even the slightest variance in devices causes one diode to conduct before the other, and all current flows through the diode with the just-slightly-lower forward voltage drop, with the second diode never conducting at all.

Keith Hilton · From: 248 Laurel Road Ozark, Missouri 65721

I have been studying distortion circuits for the last 2 years. I have formed some personal opinions about distortion circuits from my studies. There are literally thousands of distortion boxes for sale, plus most of the newer amps have distortion channels. There are basically two types of distortion, soft and hard. Some of the hard clipping is done with diodes in the output, mostly 1N4148 signal diodes. Some of the soft clipping is done in the feedback loop of an op-amp. There are other distortion arrangements, using JFETS. Of course there are different types of diodes used, even LEDs.
Some people swear by Germanium diodes. Germanium diodes are difficult to get, and are very expensive. Expensive does not mean they have to be better--just my opinion. I have used a lot of different diodes, including Germanium, and decided there was a lot more to the subject of distortion than the choice of diodes.
I do know this, satisfying people with distortion is not easy. People hear something on a recording that impresses them. Then continue to search for that sound, and seeming never find it exactly.
I think the important questions are: #1 How controllable is the distortion? Meaning, can you quickly dial in the distortion you want. Everything from soft clipping to hard clipping. #2 Do you need to alter the tone of the distortion? #3 Does the distortion unit add any sustain? #4 Is there a true bypass switch?
I have a proto-type distortion built that is totally different than any circuit I have ever seen, and I have seen hundreds of distortion circuits. I let a person use the new proto-type at a steel show, and had several people wanting to buy it. I think it beats any distortion on the market, but I may not put it into production. It is all about supply and demand. Meaning there are thousands of distortion units on the market, and everyone already has one. I don't think a manufacturer could make production costs producing a distortion unit, even if it was the best distortion unit in the world.
Probably the two most popular distortion units today are the OCD, and the Tube Screamer. The problem with the Tube Screamer is that there have been about 50 different versions, so you don't know what you are getting.

Glenn Demichele · From: (20mi N of) Chicago Illinois, USA

A few thoughts:
1). Yes, the Ge diode would shunt most of the current around the Si diode, and also have a different clipping shape. Your output level would be lower too (6dB: .3V vs. .7V).
2). You could always snip through the 1n4148 glass body and solder whatever you want to the flying leads as long as you don't get them so hot they desolder from the back side.
3). Years ago I built myself a 9V battery-powered buffer in an aluminum sunglasses case that I strap to the leg by the changer. It sounds great. I also have a switch and a pot that puts in an adjustable amount of soft clipping dirt. I used 8 Schottky diodes (1N5819) for that - four in series antiparallel with another 4 in series. Its symmetrical, but its way softer than a single junction.

Maybe clip your diodes, put in some temporary alligator clips, and try a mess of different things until you're happy. You can always put it back to original condition since 1N4148s cost about $0.005 each - as good a bargain as toothpicks.
_________________
Franklin D10 8&5, Excel D10 8&5. Both amazing guitars! Homemade buffer/overdrive with adjustable 700Hz "Fender" scoop., Moyo pedal, GT-001 effects, 2x TDA7294 80W class AB amps, or 2x BAM200 for stereo. TT12 and BW1501 each in its own closed back wedge. Also NV400 etc. etc...

Glenn Demichele · From: (20mi N of) Chicago Illinois, USA

A few thoughts:
1). Yes, the Ge diode would shunt most of the current around the Si diode, and also have a different clipping shape. Your output level would be lower too (6dB: .3V vs. .7V).
2). You could always snip through the 1n4148 glass body and solder whatever you want to the flying leads as long as you don't get them so hot they desolder from the back side.
3). Years ago I built myself a 9V battery-powered buffer in an aluminum sunglasses case that I strap to the leg by the changer. It sounds great. I also have a switch and a pot that puts in an adjustable amount of soft clipping dirt. I used 8 Schottky diodes (1N5819) for that - four in series antiparallel with another 4 in series. Its symmetrical, but its way softer than a single junction.

Maybe clip your diodes, put in some temporary alligator clips, and try a mess of different things until you're happy. You can always put it back to original condition since 1N4148s cost about $0.005 each - as good a bargain as toothpicks.
_________________
Franklin D10 8&5, Excel D10 8&5. Both amazing guitars! Homemade buffer/overdrive with adjustable 700Hz "Fender" scoop., Moyo pedal, GT-001 effects, 2x TDA7294 80W class AB amps, or 2x BAM200 for stereo. TT12 and BW1501 each in its own closed back wedge. Also NV400 etc. etc...

Paul Arntson · From: Washington, USA

AJM, check out what they call the shaka braddah circuit.
Instead of parallel, you put two different components in series.
http://www.generalguitargadgets.com/how-to-build-it/technical-help/articles/design-distortion/

It's partway down the page.

I have it in two different items and it's very satisfactory.

Also, some distortion boxes sound better to me if I reduce the gain. Less is more.
_________________
Excel D10 8&4, Supro 8, Regal resonator, Peavey Powerslide, homemade lap 12(a work in progress)

Steven Paris · From: Los Angeles

I think that the two anti-parallel diodes are useful ONLY for preventing the type of distortion that you get from a Class-D amplifier clipping. THAT type of distortion is awful, horrible sounding, so the prevention is quite useful; but not really GOOD sounding, either.
_________________
Emmons & Peavey

ajm · From: Los Angeles

Thanks to all for the comments.

I recently bought a heavily used early 80's Peavey Studio Pro 40. The CL seller said that the pots were noisy but otherwise it was fine.
I took it apart for a cleaning and well, there were a couple of other very minor problems. But some contact cleaner on the pots was the #1 thing that it needed.
So now I have a present for a friend of mine.
Total cost? $25. (Don't hate me.) At that price I couldn't pass it up and had to take the chance.

The reason I'm asking about the diodes.......

There are basically three different types of clipping circuits used in this amp. All use back to back diodes, all Si 1N4148.
1) Soft clipping in one of the op amp feedback loops. Adjustable with the PRE control. Similar to the Tube Screamer and a zillion other clones.
2) Hard clipping on the output of one of the op amp stages. Hard wired in, not adjustable. Similar to the Boss DS1, MXR Distortion +, and a zillion others.
3) Back to back soft clipping in an op amp feedback loop (or maybe the output of) for the SATURATION circuit.

It's #3 that I'm looking at modding.
To be honest, the Saturation circuit for my tastes leaves a lot to be desired.
For country or classic rock (not metal) I've read that the secret is to turn the Saturation off. Then use the Pre and Post controls just like a regular amp to get any overdrive/distortion. Then, if you want more, turn the Saturation up. I agree that this is the secret set up for this amp.

Using a footswitch, this is essentially then a two channel amp. But the Saturation circuit doesn't really do much for me personally, so I'm hoping that doing the GE diode mod will change the OD/dist character a little. And I agree with another comment: For my tastes most amps have more than enough gain/OD/dist.

If I decide to try the GE diode in parallel in the Saturation loop I'll post back. My gut tells me that it may not do the trick, though, so it'll probably end up being stock.
I don't want to get too heavy into mods on this amp. It sounds good enough as is. And since the traces are on the underside of the board, doing any heavy soldering would require removing the board which would be a pain. That is why I mentioned the "tacking on to" the existing leads on the top of the board. If I decide to replace the main filter caps that will be the method.

In addition, I also have a couple of OPA2604s that I should try. If I do I'll report back on that as well.

Even though this is a 40 year old amp, when I fired it up I was blown away. I expected some solid state, sterile, honky, fake sounding box. However, IMO this is not the case at all. This amp would do well in any southern rock type band. It really likes a Strat. The bright switch and the mid shift add a lot of tonal control. I was also surprised by the effects loop; I tried a Boss BF2 flanger and it seemed to work fine, although at extreme volumes it might overdrive the pedal. And speaking of volume, this thing is loud.

How impressed was I, you ask?

The next day I was stalking/prowling/cruising CL and ran across almost the exact same amp, with almost the exact same problems, a Peavey Studio Pro 50. However, this one was more expensive. $30. It is being cleaned up as we speak. The best one will go to my friend, and the other will stay with me as a grab and go amp.

There is one odd quirk that I noticed with both of these amps. On both of them, the:
- Reverb send from the amp was connected to the OUTPUT on the reverb pan, and
- Reverb return on the amp was connected to the INPUT on the reverb pan.
The DC in/out resistances (not impedances) on both of the reverb pan were pretty close to being the same (100 ohms versus 300 ohms) so I don't know if it matters much.
I corrected it on the SP40, and when I got the SP50 later I noticed that it was backwards as well. Strange.
I have not tried reversing them to see if it matters one way or the other.

Keith Hilton · From: 248 Laurel Road Ozark, Missouri 65721

Good luck ajm. Remember one thing; The output signal of a op-amp can be changed by the change in resistance, or capacitance, in the feedback loop of a op amp.

Ian Rae · From: Redditch, England

What Keith says about hearing a recording you like is important. That sound will have been further processed in the mixdown and is lost forever.
_________________
Make sleeping dogs tell the truth!
Homebuilt keyless U12 7x5, Excel keyless U12 8x8, Williams keyless U12 7x8, Telonics rack and 15" cabs

Georg Sørtun · From: Mandal, Agder, Norway

Keith Hilton · From: 248 Laurel Road Ozark, Missouri 65721

I did not make it clear what I meant by resistors, capacitors, and other things in the feedback loop of a op-amp. What I was getting at was the output signal strength of the op-amp. Adding or taking away stuff in the feedback loop of a op-amp will either increase, or decrease, the output signal strength of the op-amp. This is important, because the output of the op-amp feeds other things.

ajm · From: Los Angeles

Update on the second amp, the Studio Pro 50: It's up and running, and sounds as good as the SP 40.

Keith: I read your post a couple of times and get what you are saying.
I know that by changing resistors and caps in the feedback section you can change the gain and frequency response. However, for the time being, that is not the issue (or at least I don't think that it is).

The Saturation feature is still not exactly floating my boat. I did some casual browsing on the internet and didn't find any mods for these amps. I might try searching for Bandit mods from that era, as they are probably close to the same amp.

I'm starting to think that, at least for my tastes, the key to the Saturation knob is to use it sparingly.

The experiments continue.......

Steve Sycamore · From: Sweden

I can agree that distortion based on even non-simple diode clipping is very crude and horrible sounding. Also I wouldn't want to make any changes to the components on my amp's circuit board unless I was sure that the circuit and its effects were already proven.

So it would seem to be a better idea to prototype the circuit before you ever touch the amp. I'd also look into JFET's used to generate a much better sounding distortion. And also VCA circuits that can generate a variable amount of gain based on whatever characteristic you choose. Extremely high gain, to my ears, if it is limited largely to the 700 Hz to 1800 Hz frequencies can produce a really magical distortion sound with superb dynamics.

Georg Sørtun · From: Mandal, Agder, Norway

Those who really, really want "tailored distortion" can build their own with a boatload of 1N4148, resistors and capacitors, set up as output clippers, gain-modifiers in feedback loops, or for zero-cross distortion – or a combination of all three basic distortion models.

I'm partial to zero-cross distortion myself, as such circuits can be set/adjusted to distort more and more as the signal weakens, instead of relying on overdrive as the other two wariants do Very Happy

To have mentioned it: the old trick for building soft-clippers, is to use common/GND offsets and resistors in series with the diodes. Three or more clip-stages can provide as hard or as soft clipping as one wants – add more stages for softer clipping, and with enough well-tuned stages one gets close to the sound of tube-clipping.

Keith Hilton · From: 248 Laurel Road Ozark, Missouri 65721

Georg, I agree with what you said about resistors in series with the diodes. It should really be a pots wired as a variable resistors.
You can also put one, two or more diodes in series to reduce the clipping. The more diodes and resistors in series, the less or softer the distortion.
As Steve said, the JFETS work well. The JFETs are wired like a variable diode, with the voltage to the base determining how they conduct. I previously stated I have built a totally different arrangement. In some ways it is like the JFETs, except it uses variable transistors wired as diodes. I think the desirable thing is being able to control the distortion as much as you want. With control still have the sustain at low levels. I always liked the distortion Carlos Santana uses. Very soft clipping with seemingly endless note sustain.

Steve Sycamore · From: Sweden

I just started using a Mesa Lonestar and love it for electric guitar. The large cathode capacitors for each preamp stage give the tone a real heft. Channel 2 has a switch to add either 250 pF or 500 pF capacitors to bypass the tone adjustments. The 500 pF setting gives a lot of extra gain in the upper mids that gets you, you guessed it, that sweet, super thick Carlos Santana endless sustain sound.

ajm · From: Los Angeles

Well, I never thought my simple question would generate such an in depth discussion. Glad to see it. Who knows what a couple of you guys might dream up as a result?

Back to the original intent/mission, I tried a GE diode in parallel with the three other back to back pairs in the circuit.

I'll cut right to the chase. The mods didn't do what I had hoped. Therefore, they are not worth doing IMO.
I did not try adding the GE diodes in more than one place. That may be an option for the future. However, doing so may mess with the non Saturation settings, which is something that I don't want to happen.
Also, as I stated earlier, I DO NOT want to get into heavy mods that involve removing the circuit board. I also don't want to cut any existing leads to add a part in series with another. I only want to "tack on" parts from the top side. I want to keep this as simple and non-invasive as possible.
Doing this, I could also mess with adding parallel capacitors and resistors in various parts of the circuit. That is an option for the future.

Adding a single GE diode in parallel with the following and the results.
C3/C4: To my non-canine semi-unsensitive ears, it may have added a very slight amount of top end sparkle and/or cut the bass. This was only noticeable with the Saturation control maxed out, which is something I don't see most people doing.
C5/C6: Once again, fairly subtle and only a the max settings. But this one seemed to cut the gain somewhat.
C7/C8: This one produced an audible hum with no diode added. I rechecked my connections, found nothing wrong, and abandoned it. Who knows what it could have been, I assumed it was due to extra flying leads being attached.

The one mod that I did try, that had somewhat mixed results, but I'll probably keep: Changing the first two dual op amps from the stock RC4558 to OPA2604. This made the most difference with the mid shift control pulled out. The hiss went down a noticeable amount. However, some of the settings for this test were maxed out. That means that their practical value may be somewhat debatable. In actual practice, it's probably not really worth it. Since I had a couple of OPA2604s laying around, I'll probably keep it that way.
Note that for this last test/mod I have a Studio Pro 40 and a 50, so I was able to compare a modded and unmodded amp side by side. I gave these two schematics a quick glance, and there is almost no difference between them.

That is all........for now.

Keith Hilton · From: 248 Laurel Road Ozark, Missouri 65721

Steve, you talking about cathode capacitors, means you are talking about a tube amp. ajm is dealing with solid state amps, but I guess you knew that. When people talk, the perfect distortion seems to always come from a tube amp.
I am trying to think of a famous player who got near perfect distortion using a solid state amp, with a stomp box?
Wonder what amp Carlos Santana uses?

Steve Sycamore · From: Sweden

These links seem to tell some of the essentials of Carlos' equipment:

https://www.groundguitar.com/carlos-santana-guitars-and-gear/
https://www.premierguitar.com/articles/20782-rig-rundown-carlos-santana

Some other material you can find at the Mesa Boogie site. The Lonestar is more or less an updated, with many more options, version of a Fender Black Face amp or Mesa Mark I. I'm making just a few circuit modifications and speaker change to get exactly the tone I want.

Keith Hilton · From: 248 Laurel Road Ozark, Missouri 65721

George said the following:
"I'm partial to zero-cross distortion myself, as such circuits can be set/adjusted to distort more and more as the signal weakens, instead of relying on overdrive as the other two wariants do Very Happy
To have mentioned it: the old trick for building soft-clippers, is to use common/GND offsets and resistors in series with the diodes. Three or more clip-stages can provide as hard or as soft clipping as one wants â€“ add more stages for softer clipping, and with enough well-tuned stages one gets close to the sound of tube-clipping."
George would you explain in electronic terms what zero-cross distortion is? Also explain in electronic terms what common/GND offsets are? There is one part of what you said that I do not agree with. Using resistors in series with the diodes does not work. I have been experimenting with that idea all day today, and the idea does not work. Here is why: 4148 Signal Diodes break over and conduct somewhere between .6 and .7 volts. If you hook up a pot as a variable resistor, the change in voltage either allows the diode to conduct or not conduct. So there no variable range of distortion, the diode either conducts or does not conduct. You either get distortion or no distortion.
Hooking up two of the 4148 diode makes the required brake-over voltage .12 volts and .14 volts.

Georg Sørtun · From: Mandal, Agder, Norway

Keith Hilton · From: 248 Laurel Road Ozark, Missouri 65721

Thanks Georg. I understand the crossover distortion you described. Georg, I am still not clear on using the resistors with the clipping diodes. If you put the resistors in series "before" the diodes, the resistors cause a voltage drop. So the resistors before the diodes have a direct effect on the voltage required for the diode to conduct. I can't hear any change in distortion--in relation to current--when the resistors are in series, in front of the diodes.
If you put the resistors in series "after" the diodes, then the resistors do not cause the voltage drop which effects when the diode turns on. Even with the resistors following the diodes, I don't think the change in current would do much to change the distortion. At lease I don't hear it.
Georg, explain with a diagram how the change in current through a diode using a resistor to ground would cause more or less audible distortion? As I see it, more or less current would cause more or less volume, but the distortion would be the same. I am not saying you are wrong. I just need some circuit examples to understand.

Randy Schneider · From: SW New Mexico, USA

Forum Index > Electronics	Goto page 1, 2, 3 Next	Next oldest topic :: Next newest topic