This year, I set out to build my own guitar amplifier with the help of too many people to list here. While it proved top be an extremely entertaining and rewarding experience, there were many headaches that came along with designing an amp. This article was originally written by me and published on the Usenet group alt.guitar.amps, in an attempt to share a few of the lessons I learned in the process. I hope you will find it interesting, and that this information may save you some time and frustration!
Gil Ayan
Los Angeles, October 1998
Since I know some of you do engage in amp customization, I thought I'd pass on to you
some of the things I have learned about lead dress in an amplifier, which is probably one
of the hardest things to do correctly. Although most typically a problem when high-gain
stages are present, even Fender had its share of trouble with lead dress starting with the
second generation Silverface amps, and these amps can be hardly considered high gain. Most
Fender fans will say that Fenders were always less prone to oscillations that Marhalls,
making it sound as if there is something superior about the design. While it's true that
Marshalls almost always behaved worse, it warrants pointing out that Marshalls typically
run at higher levels of gain, so it's the old apples and oranges stuff.
Most people read about this subject, and know a few things, such that grid leads
"should be" short, etc. But, the shortest path between two given points is not
necessarily the best for tone's sake. In the specific case of Fender amps, there is a grid lead that goes from the rightmost volume control to the 2nd stage of the "vibrato" input. This is the single longest lead in the hole amp, and the one that can give you a tremendous headache if you're building additional gain stages within the amp. No pain, no "gain" though (yes, pun intended). [Also, even if you're not considering customizing any amp, you can always improve the stability of your later Silverface amp by tweaking the lead dress and getting rid of the tone-eating caps at the power tubes]
I found that moving that wire around could do one of three things: give a good sound, make
the amp break into oscillation (whether you can HEAR the oscillation or just
"feel" it -- the old parasitic type), or result in a very nasal sound. You can do the lead dress by trial and error, and maybe you'll get there, but there is a way a found to be more effective and it does work indeed.
Use a long wire, temporarily of course, to make the connection whose lead dress you want
to optimize. If you pull that wire away from the rest of the circuit, you will hear the
tone is unfettered, even across all frequencies and without any unwanted side effects --
well, you will most likely find this anyway. Get a cable with alligator clips on both
ends; wrap a portion of that wire around the grid lead. Leave one end of the alligator wire not touching anything (you can clip it to the insulation of the cable so that is doesn't move around), and with the other end, start probing the entire circuitry. When touching plate connections, remember you're dealing with 200+ volts, so... be cautious.
What's going on here is that now you're capacitance coupling your grid lead wire to all
other points of the circuit via that alligator clip patch wire, which is similar to what
would happen if you lay your wired back into the circuit. As expected, plate leads from stages which are in phase with the grid wire in question will make your amp squeal like a pig when you touch them with the alligator clip, so you'll want to stay far from those wires.
Conversely, plate leads out of phase with the grid lead will KILL TONE big time,
again keep your distance. In some extreme cases, you may lose almost all your output, but even if that doesn't happen, you will think there is a wah-wah stuck on the bass side of the throw because of the degenerative
feedback that's taking place. Additionally, there are several points within the circuit
that will also affect the sound, although maybe not as severely. Also, bear in mind there
is some stuff going on underneath the chassis as well, so if you're probing around open areas of the chassis and find that, although it should "all be ground," there is a still change in tone, well... it could be your reverb transformer, the filter capacitor can, etc.
What to do? Map out the areas of the circuit where it is "safe" to run your lead
dress, based upon your experimentation as described above, and you will end up with an
optimum path for the grid wire. In most cases, this will not be a straight line -- which
is obviously the shortest distance, belying the rule suggesting to use the shortest
possible lead for grid connections. If you HAVE to go across another wire which you've determined to be an offender with the probing
exercise, cross it perpendicularly; this will minimize capacitance coupling and the tone will be most likely preserved.
What about using shielded cable? Yes, absolutely, that will help provided you be sure to
ground one end of the shield; other people recommend you not ground either end and, and
instead connect one end of the shield to the corresponding stage's plate; I have found
this to be another tone killer, and there is the potential hazard of having a hot shield.
It's been my experience that shielded wire is not bullet proof and might be subject to the same effects I mentioned earlier, although to a lesser extend if you do things correctly. Using generic microphone cable is bad, probably even worse than using no
shielding at all if you have cheap wire. It doesn't provide good protection against cross talk and the copper mesh will be an antenna
attracting all kinds of stray signals to your lead wire. Even if you ground the shield,
you'll be stuck with a signal riding on top of your grid wire, and there will be some
interaction and or modulation. Foil type shielded cable, although much more expensive, is
the only way to go.
If you've seen MESA/Boogie schematics, you will notice that some models (the Marks II and
above, for instance), have all kinds of capacitors in their overdrive sections that either
shunt to ground or bypass the plate load resistors. Some writers have said that it helps
control an otherwise extremely buzzy sound, and that it helps to stabilize the design.
Yes, it stabilizes the gain stages, but the midrange heavy tone which results may not be
what you're shooting for. Regarding the buzzy sound, if you can allow your highs to go
through all the stages unaltered, they will not sound so buzzy anymore: the will be sweet because they break up easily and give lots of harmonics, resulting in your amp being extremely touch sensitive -- one of the keys to the sound of Dumble amplifiers, for example. If you think you can "control" the treble response
throughout the preamp overdrive stages, a la MESA/Boogie, and then use a presence control to bring the amp back to life, it won't work even close. The presence control will sound stiff and flat because the treble it will let through doesn't have as much harmonic richness since the highs were stopped at the overdrive stages. I do not like the presence control on the Mark II and IIIs for this very reason; implement the same feedback network on a more harmonically rich amp and the difference is striking. [Since I have had a chance to see the guts of some very old Boogies, it seems to be that the Boogie midrange heavy
tone may have evolved out of necessity, because the chassis were very crowded and the amps
would have been unstable without dumping a lot oh highs to ground. However, there is no
denying that Boogie came up with a legendary sound, a great sound in many people's
opinion.]
Hope this helps some people out there to save some of the time I had to spend lead
dressing my amp. Some books address various anti-microphonic techniques (long plate
connections, short grid connections, shielding, etc.), but that alone will sometimes not be enough. I encourage everyone form staying away from using small capacitors to get rid of oscillations if you can... yes, that method does work -- and so does turning off your amp, for that matter -- but it simply nukes tone. Common tech tips, such as wrapping a loop of cable around a lead wire and connecting one end of the loop to whatever quiets down the amp are the same as using capacitors to combat oscillations, so use the as your last resort if you have to.
Cheers,
Gil
©2008 Gil Ayan. All Rights Reserved