GuitarPCB.com

Guitar Pedal Projects. Top 5 in DIY With Forum Support!

Forums

Post Reply
Forum Home > Tips, Tricks and Tutorials! - Builder's Guides for GuitarPCB Boards! > Sticky: Choosing Clipping Diodes

David
Member
Posts: 122

I just finished another Capt. Munch.  The circuit calls for 2 LED's.  I wanted to "tame the Beast" some by using lower gain diodes. My problem is that there are so many options to choose.  I have a large # if diodes which to choose from, ie. 1n60, BAT46, 1N34A, MPSA 18 (low noise tranny) etc.  Is there a resource available that lists the relative gain, in a chart form, for comparison? Again I am looking for lower gain diodes. Thanks experts.

September 27, 2011 at 5:04 AM Flag Quote & Reply

TNblueshawk
Member
Posts: 238

David, I assume you saw the answers over at BYOC? I learned something off them for sure.

September 28, 2011 at 8:48 AM Flag Quote & Reply

Barry
Site Owner
Posts: 8793

There are other options but not using LEDs are not going to Tame the beast per say.

You can try Germanium and see what happens though.

Different trannies will yeild different results.

You can learn more by Downloading the Datasheets and reading what the average Gains are.

This is suopposed to be a High Gain pedal so if you really want to tame the beast I believe Tonmann might have a suggestion for you but will be back in a few days. Maybe bump it then.


October 2, 2011 at 10:27 PM Flag Quote & Reply

tonmann
Moderator
Posts: 1239

Before I tackle this one, there are two points to make:


1. Diodes don't  have any gain.  The main characteristic that we are interested in is the forward voltage of the diode.


2. "Tame The Beast".  I love the name so much that if it becomes a switchable mod I will be naming the mod TTB.  The main problem here is that you will have to define TTB a little bit more precisely - the same sound with less ouput level, more signal clipping, softer signal clipping and so on.


Once we have TTB defined, I'll do an extensive  post, diagrams and all.

October 4, 2011 at 9:48 PM Flag Quote & Reply

TNblueshawk
Member
Posts: 238

I'm not speaking for David but I thought I would jump in. When David first posted this, and me being fairly new to this world of electrons, I thought I was thinking the same thing as him. In my mind I thought yeah, I wonder what softer clipping and hard clipping rules of thumb there are if such exists. Not sure if that is where David was going. Maybe a softer clip results in less volume, gain and or 'harshness"  ,if you will, although not technically correct. Noob language I call it :) .

October 5, 2011 at 3:40 PM Flag Quote & Reply

patrickbrose
Member
Posts: 297

I would suggest that the difference would be in relation to diode forward voltage. For example, in the Aeon Drive circuit, when you switch to of the 1n914's to LEDs you get a much more aggressive overal overdrive. So for the TTB concept, the 1n914 (or 1n4148s) would be tame compared to LEDs.

-P

--


October 5, 2011 at 9:36 PM Flag Quote & Reply

David
Member
Posts: 122

Thanks Tonmann,  Good question. What do I mean by TTB (I love it). I guess as you pointed out, a slightly less output volume as well as slightly softer signal clipping.  Additionally,  what would be your recommendation as far as a replacement for the OP amps to reduce the gain some.  I would get more use out of the pedal with a little less gain. That is just my playing style.  Thanks

October 6, 2011 at 3:34 PM Flag Quote & Reply

Barry
Site Owner
Posts: 8793

Tonmann I believe David would like to lower the amount of available "Distortion" as opposed to the actual Gain or better yet Volume of the pedal. In the US we often use the term Gain and we really mean Distortion so I am assuming that David is only wanting to reduce the amount of available Distortion without sacrificing Volume.


David feel free to correct me if I am wrong.


If I am am indeed correct then....

David if you turn the "Distortion" Knob down to say 10:00 are you saying that it is still too much?

Point is it may help to know where the happy point is so that a reasonable medium can be found.


Lastly if you even cannot get a satisfactory sound with the "Distortion" knob at 10:00 this may not be the right pedal for you?? I am not saying for sure but merely speaking hypothetically.

October 6, 2011 at 5:31 PM Flag Quote & Reply

patrickbrose
Member
Posts: 297

I was thinking the same as Barry, Maybe what you need is a pedal with less overal distortion. The munch is pretty high-gain distortion. You may want to look at a lower gain overdrive. Maybe an Aeon drive or a MKC1...

--


October 7, 2011 at 1:51 PM Flag Quote & Reply

tonmann
Moderator
Posts: 1239

OK a look at diode clipping - first off I'll be looking at what happens and not how or why it happens as that would involve a lot of techie explanations.  Lots of diagrams to accompany the explanations so you wount get bored.

The circuit is the Capt Munch and the frequencies used for measuring are 150mV peak signal at 84Hz (it should have been 82Hz but I didn't notice that my tone generator was slightly off untill after a few measurements, so I left it) which is somewhere between the low E and F on the sixth string and 20mV peak at 500Hz (just because it is a nice round number)  which is between the B and C (7th / 8th fret on the first string) - these are not unreasonable values for a guitar.


I removed the diodes, used a 500Hz signal and set the gain pot from 20% to 80% in 20% increments and looked at the signal where the diodes should be connected (between R6 and C8:



A few things to notice

The signal is not completely symetrical - the negative going part (the part that moves from zero volts downwards) of the signal is slightly larger than the positive going signal, this is caused by the bias voltage of the op amp being slightly off (I measured it at 4.8V) so the positive going part of the signal clips earlier than the negative part.

Once the signal clips (at 60%) the clipped portion isn't completely flat - there is a slight slant caused by capacitors connected to this stage of the circuit.

Neither of the above will have any noticeable affect on the circuit.


Looking at the above the signal is just starting to clip at 60% rotation and looks more or less like a square wave at 80%.  The signal at 60% could be discribed as soft clipping - the sides of the signal aren't nearly as vertical as the 80% signal and only a small amount of the top of the signal is clipped.


Inserting two LEDs for D1 and D2:



The signal is clipped at 1.5V (peak value) - we usually measure from zero volts upwards and assume that the negative part of the signal is at the same level as the positive part of the signal.


Also notice that the signal at 40% rotation has a very slight flattening at the apex of the signal - very soft clipping indeed.  With the inclusion of the LEDs, the signal at 60% is clipped harder than without the LEDs i.e. op amp clipping.


Soft and hard clipping has nothing to do with the amplitude (level) of the signal but with the shape.

For those who are still scratching their heads, here is a Sound Sample of a 110 Hz signal going from a sine wave (like the light blue wave in the above diagram) to a square wave. The signal level is constant but as the shape of the signal changes to a square wave you can hear more high frequencies making the signal sound harsher. That is the difference between hard and soft clipping.



October 7, 2011 at 5:41 PM Flag Quote & Reply

tonmann
Moderator
Posts: 1239

Just so I dont get timed out, I'll split the post.

The sound sample in the last post tends to stick on the first play; let it run and play it again.


Looking at using different types of diodes:



Using various types of diodes, including the base / emiiter of an MPSA18 and the drain /source of an IRF730 and two Shottky diodes you can see that the signal levels going into the tone control stage vary quite dramatically.  As Patrick pointed out this is to do with the forward voltage of the diodes, the lower the forward voltage - the sooner the signal clips, the sooner the signal clips - the more "square wavish" the signal looks.

Other factors that I will mention briefly are the time it takes a diode to turn on and off and the linearity when turning on and off - these also affect the shape of the apex of the signal to some degree.


Using the same diodes and applying the 84Hz 150mV signal:



Due to the high input signal and the high gain, all signals are looking more like square waves albeit with a pronounced slope.

In order to get rid of the slope and reduce the signal level slightly you would have to look at R6:



In this circuit I've used 1N914s as the clipping diodes. By increasing the value of R6, I've flattened the slope and reduced the signal level slightly; for the 500Hz signal:



The signal level is reduced and the clipping becomes softer.  I would suggest not going above 4k7 for R6.


Next post.......

October 7, 2011 at 6:53 PM Flag Quote & Reply

tonmann
Moderator
Posts: 1239

In order to "Tame The Beast" in the Capt Munch circuit the problem lies in the fact that low frequencies are boosted too much - we need high gain to boost the higher frequencies so that they clip but this is boosting lower frequencies so much that they are hard clipping.


Although the circuit is designed to boost higher frequencies much more than lower frequencies, we need less gain at the lower frequencies.


Probably the best place to do this is at C5:




Although you still have 50% gain pot rotation to play with, changing C5 from 220n to 33n might just be a bit  too tame but it is a good starting place (probably change R6 to 1k).

October 7, 2011 at 7:26 PM Flag Quote & Reply

draehn
Moderator
Posts: 579

Pictures really help me out on this! Thanks, Tonmann for putting this together. This thread should be in the reference library for advanced users.

October 7, 2011 at 7:42 PM Flag Quote & Reply

patrickbrose
Member
Posts: 297

draehn at October 7, 2011 at 7:42 PM

Pictures really help me out on this! Thanks, Tonmann for putting this together. This thread should be in the reference library for advanced users.

It has been done.

--


October 7, 2011 at 9:27 PM Flag Quote & Reply

Bruce R
Moderator
Posts: 1680

Wow, Tonmann, you've outdone yourself, and that's sayin' somethin'

--

_________________________

Please visit PedalPartsAndKits.com if you want a GuitarPCB kit in the USA!

October 7, 2011 at 10:38 PM Flag Quote & Reply

draehn
Moderator
Posts: 579

I'm not known for doing things the easy way. With this in mind I was looking at the grouping of different diodes and had an idea...

Tonmann, what would a MOV do for clipping in place of a diode pair? How would the forward voltage characteristics work out in that scenario? I have a COR500 I might try it out on if you think it wouldn't be a waste of time.

October 7, 2011 at 11:17 PM Flag Quote & Reply

David
Member
Posts: 122

Tonmann, Thanks for your intuition and knowledge.  I just wanted to make clear that the pedal as is is fantastic for what it was designed for. That is a high gain unit. I love the tonal quality of the pedal, the active tone etc,. As such and for my style of playing, if I could "tame the beast" by lowering the distortion a little and soften the clipping, this might well be my go to pedal for highter gain and lead stuff. As it stands now I accomplish this by running the LDO and MKC back to back. I use other combinations as well (zenith, other dirt, etc.).  I am going to look at playing with the values of C6 and R5 as suggested  :P

October 11, 2011 at 9:57 AM Flag Quote & Reply

tonmann
Moderator
Posts: 1239

I don't think an MOV (Metal Oxide Varistor) will work.  These are normally used in high voltage applications to prevent transient (sudden)  voltage spikes from reaching circuits - a typical example is an MOV across the secondary winding of a power transformer.

A diode is basically high resistance at a low voltage (no current flows through it) and  low resistance at higher voltages (current flowing through the diode).




At some arbitary point the diode is said to be conducting i.e. gone from a high resistance state to a low resistance state; this voltage is called the forward voltage (or sometimes clamping voltage).


Any diode used in a clipping circuit must have a lower forward voltage than the signal feeding it otherwise the diode would not turn on and the only clipping that would occur would be the op amp or transistor clipping itself.


All of the diodes we use for clipping are going to have a forward voltage of about 0.2V to 1.5V:



MOVs have a very high foward voltage running from tens of volts to hundreds of volts - so they would never be turned on by any signal that a pedal circuit can produce.



October 11, 2011 at 6:24 PM Flag Quote & Reply

draehn
Moderator
Posts: 579

Thanks, Tonmann. There wasn't enough info on the interwebs for me to discern the low voltage application. I only knew that they were mechanically a diode pair and thought that maybe they could impart some character to a the clipping section of an effect.

I second the comment that you have outdone yourself on this thread and thank you for giving us a way to visualise some of the terms that have been used so commonly.

October 11, 2011 at 9:38 PM Flag Quote & Reply

TNblueshawk
Member
Posts: 238

+1 to draehn. It is an artform to explain these things to those of us who know next to nothing (speaking of myself here) and in turn have us understand what you mean. Some of the worst teachers I've had in life, are some of the smartest people I knew, but they couldn't explain 1+1 to me if I didn't understand it.

October 13, 2011 at 4:05 PM Flag Quote & Reply

You must login to post.