Clipping diodes - Marshall JCM 800

[BeëlzeM]

Hello guys, I'm new here... I'm interested in modding amps and guitar effects. Done some in the past, mostly pedals. But I don't have much experience with modding amps.

I own a Marshall JCM 800 2205 head. I found a mod through someone, he strongly recommended it, to reduce the fizzy, cold, sterile gain this amp has. It uses 1N914 clipping diodes, as seen in this image...

http://img397.imageshack.us/img397/1...5diodesjz6.gif

It's D4 and D5. The mod was to replace the 1N914 diodes with 1N34A's and 1N4001's in series. This would increase the warmth and depth of the tone and gain, and take out the sterile and cold buzz which seems like a great idea.

My question was, does this make sense to you guys? Since I don't know the specifications and characteristics of different types of diodes. Also, does it make sense to solder those particular two in series?

Any other clipping combo's that would work well for this amp? Another guy also recommended me 1N4148's with a LED on one side and just one 1N4148 on the other side.

I'm looking for a deep raw tone like this, if it helps:
http://www.youtube.com/watch?v=UdSdgOYBH88

Thanks in advance

[jrfrond]

Makes NO sense at all! D4/5 are steering diodes for the footswitching circuit.

?????

[Robert M. Martinelli]

Telling the truth, none of the highlighted diodes has a clipping function, two diodes (D2 and D3) are used as rectifiers, D1 is part of the V1B biasing network. I can't read well the component to which D6 is connected, the letters are blurred, but that doesn't look look a clipping stage either.

Back to the "clipping diodes" question and their effect, everything lies in the different threshold voltages different diodes have.

Clipping diodes are usually found "reverse-paralleled", this configuration allows the two halves of the signal to pass freely until the signal amplitude reaches the diodes' threshold voltage, at that point the diodes start to conduct and shunt the signal exceeding the diodes' threshold voltage to GND.

Putting diodes in series adds threshold voltages and this reduces clipping.

Germanium diodes have a lower threshold voltage ( 0,.2 - 0.3 V ) so they tend to clip hard and squash the signal to 0.4 - 0.6 VPP, the resulting waveform, especially if the input signal amplitude is high, is rather "squared" so the corresponding sound is rather "fuzzy".

Silicon diodes' threshold voltage is around 0.5 - 0.6 V, so they clip less and thus the resulting waveform is more "gentle".

Red LEDs have a threshold around 1.6 V so they give a good "crunch" without much waveform "squaring".

A combination of different diodes yields asymmetric clipping, which some find rather pleasing to the ear.

Hope this helps

Best regards

Bob

[teemuk]

Telling the truth, none of the highlighted diodes has a clipping function... I can't read well the component to which D6 is connected, the letters are blurred, but that doesn't look look a clipping stage either.

Actually, two of them have. D6 is connected to a rectifier bridge, which effectively creates a three-diode shunting string per each halfwave. It's a similar setup to that used in the JCM900's, just replaces the discrete diodes with an integrated rectifer.

D1 clips the positive halfwave of the signal in the cathode. When the gain stage inverts the signal, effectively the negative halfwave ends up clipped. If the tube is "traditionally" overdriven it clips the positive halfwave as well. I think it will also push the grid into negative DC offset during clipping. I'm pretty sure this setup was discussed here in a better detail at some point. Can't recall in which thread, though.

[BeëlzeM]

Oh well. I'm confused about the diodes. So it's obviously not D4 and D5.

I'm sure there is a diode bridge in here. Would that be B05 where D6 is connected to?

I've spoken to guys before that changed the diode bridge. What can I do?

[Robert M. Martinelli]

Actually, two of them have. D6 is connected to a rectifier bridge, which effectively creates a three-diode shunting string per each halfwave. It's a similar setup to that used in the JCM900's, just replaces the discrete diodes with an integrated rectifer.

@ Teemuk : As I said in my previous post, I couldn't read what the component tied to D6 was, the letters were blurred, so I was unable to state its purpose.

@ Beelzem : Being that a rectifier bridge as teemuk says, ( maybe my sight is getting worse ) it's most likely a silicon one, so the sound will be much alike using 4x1N4007 and the like. It could be worth to try Germanium diodes in its place, but this would mean the following stage's gain will have to be doubled, as the Germanium diodes will clip the signal harder reducing the signal amplitude going to the following stage.

Hope this helps

Best regards

Bob

[BeëlzeM]

Thanks Bob,

I guess it wouldn't be an option to use diodes that clip harder, since I'm pretty much using all the gain I can get at the moment for my style. I already increased my gain a bit by placing a 10 kohm resistor across the 10 kohm R13 on V2. I also replaced R16 with a lower value resistor for a bit more response and a slight gain increase.

Oh, and about the thing where D6 is connected to, I guess that would be W005.

I'm looking for a tone somewhat like this, recorded in 1989 with 800 heads:
http://www.youtube.com/watch?v=fD0PJ-fKeHA

Any other tips or things I can do? Other combinations that might alter my tone somewhat without reducing the gain for example?

My head isn't as aggressive as that YouTube vid, I'd like to spice it up some more.

[Steve Conner]

You can replace D6 for a single 1N34, or put a 1N34 in series with it, or swap things out for a LED (try different colours for fun, they have different clipping levels) or even yank it out altogether, disabling the diode clipper, to see what happens.

For your style you probably want a Boss Metal Zone and a faithful bass player doubling your root notes. The bass player is where all the so-called "warmth" and "depth" of recorded metal tones come from: if you could get the multitrack and listen to the guitars alone, they'd sound like bees in a can.

[BeëlzeM]

Nah. I'm not into the Metal Zone much. I have a SD-1 to boost my signal in front of the amp and I'm pretty happy with it. And I'm sure those won't sound like bees in a can, guitaramps sound good on their own and these are Marshalls as well, not really bees . I have a bassplayer in my band though, and I agree, you can't do without one.

I'm now thinking what would sound good in my amp, what bridges or diodes would make sense to use, keeping in mind what sound I'm after. Which would clip more or earlier and have more output, etcetera...

[Robert M. Martinelli]

Hi again,
I agree with Steve and I still think experimenting with different diodes it's still worth.As to LEDs, I only mentioned red ones in my first post, so I thank Steve for pointing out that it's also worth to try other colors as well ( green LEDs' threshold voltage should be in the 1.8 to 2 V range, so they would give lesser crunch than red ones ).

I see you're pretty much concerned about gain, saying you' re using as much as you can, but you surely understand that one thing is getting distortion as a result of high gain ( overdrive ) one completely different thing is getting distortion as a result of clipping; If you' re after the latter then you probably don't need excessive gain before the clipping stage, especially if you use Germanium diodes, as their threshold is so low that they clip pretty hard anyway. All other things being equal, the higher the amplitude at their input, the squarer the waveshape ( and the higher the number of bees inside your amp ).

I think too much gain used together with a clipping stage will compromise definition, making the sound blur. I like to get a good distortion, but not at the cost of losing character and definition.

Hope this helps

Best regards

Bob

[Steve Conner]

Well, when you ask for more gain and aggression, I hear you asking for more bees and more cold buzz. In my mental tone vocabulary, warmth and aggression are opposite ends of the same stick.

At the end of the day, fiddling with the diode clipper will probably just make minor changes, compared to the drastic effect of, say, an EQ in your effects loop ala Dimebag, or changing your speakers for Vintage 30s.

Like I said previously, D6 is a good place to explore changes. Whatever you put in there will affect the clipping level and characteristic. Different devices have different characteristics, from a germanium diode at one extreme, to a blue LED at the other. I don't know what the effect of these will be, so why don't you get a 1N34 and a blue LED, and try both.

Nothing you do to D6 will affect the symmetry of the clipping. To change that, you have to yank out the bridge rectifier W005 and replace it with combinations of diodes.

The diode clipper is a trick that Marshall used to get more sustain and distortion out of a given number of tube stages without having them fart out of control, which is what it'll probably do at full gain with the clipper removed. The price paid is a bit less complexity in the tone, that some people might call sterile, when you compare it with the Mesa or Soldano approach to gain.

[Gregg]

I've built this amp in the past and made all kind of experiments with it. Removing the bridge rectifier changes the sound in a negative to my personal taste direction. The same applies to D1.
The best way is to perform these and other mods, listen to the results and decide for yourself instead of guessing and relying on someone else's taste and opinion.