Home (amp tone and effects placement)

 


3-stage amp rigs: test results, conclusions, and revised opinions

 

January 2001

Mike D wrote: <!-- MDoyle at pilotonline com -->

>I am considering changing my existing set-up.

 

Current setup:

Minimal fx

Marshall

Cab

 

Proposed setup:

Minimal fx

Marshall head

Spk Out jack

CAE Load Box

Power amp

 

>1.  What is the advantage to doing this?  

 

You can get the exact same amount of power-tube saturation at any SPL (any final-monitoring volume in the room).  Thus you separate SPL from power-tube Tone.

 

You can optionally listen to time fx in the guitar speaker, rather than having to use a separate full-range speaker to hear time fx.  Without the garbling that results from the ordinary, half-bad chain:

preampdist

time fx

power tube saturation

guitar speaker

 

 

 

>2.  How will my tone be different? 

 

Slightly less dynamic depth.  This is the only downside and there are many advantages.

More consistent tone at any volume

More control over the tone and ability to get variations by inserting EQ between power tubes and cab (equivalent to swapping speakers)

Ability to put heavy time effects after heavy power-tube saturation with no garbling of the Basic Tone.

 

 

>3.  Do you recommend this method? 

 

Yes, but speaker smoothing from a single hard-pushed speaker is still good to add.

 

 

>4.  Why do so many professionals choose to do this?

 

So they don't go deaf and get thrown out of home and clubs and recording studios.

So they can play through a 95% great Tone, in 100% of the situations they want to.

So they can run time fx with power-tube saturation without garbling the Basic Tone.

 

For control, in the broadest sense.  When combining many preamps, heads, and cabs, people use a full dummy load and linear power amp to normalize all the heads and cabs to a compatible level.

Cabinet emulation and control:

o  eq before preamp distortion gives you a lot of control of preamp distortion voicing -- same as giving control over your pickup response (still can't control inter-gain-stage voicing)

o  eq before power tubes gives you control of power-tube saturation response

o  eq immediately before cab gives you a lot of control over your speaker frequency response

 

To insert eq in that impossible spot, do this chain:

o  power tubes

o  output xformer

o  dummy load

o  eq

o  time fx

o  linear power amp

o  guitar speakers

o  mics

 

Eddie Van Halen originally did this, and might still do it -- I've seen more nonsensical and garbled diagrams for him than anyone.

 

>- Mike D

 

 

 

Notes:

 

Custom Audio Electronics RS-10 pedalboard. The same used by Eddie van Halen and Steve Vai. $400.00

 

...information regarding a custom made "load box," which was approx. 8"h x 5"w, black and rectanglar in design, with three knobs and a large VU meter.  I believe this load box was designed in the late 1970's and used during the 1980's by David Gilmour.  Gilmour used a MESA/Boogie Mark II amp head to smooth out the sound from his distortion pedals to make them more harmonically pleasing, and to increase sustain, and sent this signal via the load box into some effects and Fender Twin Reverb amps

 

 

 

Contents

6/13/2000 - just finished a good A/B test of 15 watt Kay amp pushing speaker directly (miked in isolation box) vs. Kay amp pushing resistor, then eq, then solid-state amp pushing the speaker in the isolation box. I used 2 Radio Shack Source Selector switchboxes. I tried changing eq before the tube power amp, after the mics, and between the ss amp and iso box - but concluded that the resistor then re-amp lacks some dynamics of directly pushing the speaker. This puts the London Power: Studio or Session amp back into the picture, and the AX84 5-watt EL84 amp kit. The surest bet of all is the large 6x3x4' speaker isolation box containing a conventional 2x12 cab, driven directly by a tube power amp. The next thing to test is whether loading the amp with a raw speaker in a blanket, or with another iso cab, produces all the dynamics of directly driving a guitar speaker with a saturating tube amp. I also need to use a variety of playing styles (EVH tone was used today). Directly driving produced juicy, very present Tone, and especially, when yanking the EQ all over (before the tube power amp), it was hard to find a bad sound -- that setup is ultra-forgiving and seems to present every setting in the best light. But when driving a resistor as load, tapping that, and re-amping finally to the speaker, that EQ stage produced some bad sounds, and overall the R load sounded relatively flat and perpetually mis-equalized (kind of closed, cramped sound), perhaps halfway between a POD and directly driving the speaker with the tube amp. The conventional approach produced a great, definitely acceptable sound, even with a wide range of EQ settings throughout the chain. The alternative, R approach produced an always questionable sound, it always needed a little more adjustment, it would never quite sound right. ___________________________

(3-stage = Guytron amp and Van Halen's rig: preamp, saturating power amp, final amp)

Main conclusion: for home studios for Rock guitar, based on these A/B tests, you could use a conventional tube power amp with conventional power tubes, driving a load -- THD Hot Plate seems well-conceived, Weber coneless speaker will be interesting -- with line-level tap pot, then eq and time fx, then nondistorting amp, then guitar speaker, particularly a conventional cab with all but 1 speaker disconnected, in an isolation box large enough to contain a 4x12 cab, mikes, and space out front (6x3x4). This rig enables pushing the speaker (with nondistorting amp) hard enough for speaker rounding, but at a low enough volume to enable using a condenser mic -- and at a high enough volume to rise above the mic noise and amp's background noise.

Revised hypothesis: I often said power attenuators sound worse the more you attenuate. I now suspect that they could sound perfect until you attenuate down to the level at which each speaker is being pushed at less than 1 watt (1 watt per speaker in 4 speakers is so loud, you would tend to attenuate this much). A new concept in attenuation is that as you attenuate the wattage, you must disconnect speakers so that each speaker is pushed by at least about 1.5 watts (which is still fairly loud). The attenuators were not, after all, to blame, but rather, the lack of understanding of wattage and log scaling for perceived volume. People were radically attenuating (measuring by the great reduction of wattage) so that each speaker was being pushed by less than half a watt. *That* is why people thought that attenuation degraded the signal. Extreme attenuation doesn't degrade the signal; it just fails to push each speaker hard enough to develop rounding. The solution is to remove speakers as you attenuate, and use sound isolation materials (such as gobos) to keep the speaker above 1 watt without too much SPL.

Next test pass: I have to test the lowest acceptable volume (to give minimum leakage) but I can already conclude that the speaker does not sound markedly better when pushed to the upper limit of 15 watts/ condenser clipping level, compared to the lower limit needed to get above the noise floor of the mic and amp. The bottom line for speaker leakage SPL is that you can get the authentic Rock sound by pushing the speaker to about 1.5 watts, which, with a single-layer isolation box containing the conventional 2x12 cab with 1 speaker, will be just a little audible in the next room. When convenient, you can turn up to the condenser mic clipping level, about 15 watts, but the speaker won't really sound better, just louder (test this again, specifically).

Next test pass:
A.
tube power amp
speaker
mic

B.
tube power amp
resistor load
normalization eq
amp
speaker
mic

I already concluded that the Tone is preserved except with an incidental EQ shift [next day, though, this a/b test indicated clearly that A is better, see top of this page] -- but I need to specifically confirm that the distortion character and distortion dynamics are preserved (now that I have learned how to probe these).

Technique: to assess the distortion character of a power amp, set up two or more amps with pre-eq pedal and a/b boxes for their signal In's and spk Out's. Yank the eq levels all over the place, to check how quickly the amp saturates, and whether it does so gracefully or instantly produces a garbled, strangled, unforgiving type of distortion response. Can you pinpoint a place at which it starts to suddenly, radically distort, in an unpleasant way?

My double-layer speaker isolation cabinet seems to be working great now, especially since I discovered how to put EQ between the saturating power tube and the guitar speaker. I am pushing the speaker with 2-15 watts, and usually use the ancient Rola speaker with small magnet, but I plan to use a Weber speaker instead. Currently there's a Shure SM57 dynamic mic facing the speaker, 1" away, 1 1/4" from the edge, and on the other side of the speaker is an AKG C1000S condenser mic in the same position. At first I thought this position was far too dark, but I seem to have gotten used to it - I haven't even been using any EQ after the mics, for most experimentation. I put the mics there while emulating the amp tone on Metallica's album _Ride the Lightning_ - a very dark miking tone. Their album _Kill 'Em All_ is much more crunchy like a distortion pedal played through a 6550 tube power amp below its saturation region, and without pushing the speakers hard, but the leads are smoothed out, not crunchy/crispy.

I have a sample ready for MP3 conversion, of a Ride the Lightning passage. I hope to post some samples in a week. I was able to emulate the Tone very closely, using the 1.5 watt Tiny Tone (1 6AQ5) and 3 eq stages, and the preamp distortion of a Marshall Valvestate. But I want to redo this with the tube power amp from the ancient Kay (point-to-point wired, 2 6V6) or the Fender Blues Jr. (2 EL84).

Conclusion: a speaker isolation box doesn't have to be huge, but I don't know if a small symmetrical cab like Demeter Silent Speaker Chamber is pushing it too far, too compact. However, other tests indicate all the dynamics and EQ can be preserved by inserting an EQ between the tube power amp and speaker, helping a lot to compensate for undersized cabs -- as follows: tube power amp, load (raw speaker, resistor, coneless speaker), line-level pot, eq, final amp (ss ok), guitar speaker

Conclusion: the 1.5 watt Tiny Tone tube amp's power amp distortion is halfway between the breakup response of a conventional tube amp and a solid-state amp. The solid-state amp I compared was just a 0.5 watt Radio Shack experimenter's $13 test amp with input and output jacks as well as internal speaker.

Conclusion: when set up in a research modular system with full EQ'ing and gain controls, and both a miked speaker or a power resistor, tap, eq, amp, then miked speaker, a solid-state amp has a rattier, more alarmed breakup than a convention tube power amp. You can always easily tell where a solid-state amp's breakup point is -- the point at which it screams suddenly if you turn it up higher. ss amps have a very limited range -- no range of semi-breakup, compared to a conventional tube power amp. You cannot tell where a tube amp breaks up - it has a wide region between clean and slammed (completely overwhelmed and mangled-sounding).

Technique: this modular A/B/C test system is awesome -- you can compare amps and setups and loads pretty fairly. My conclusions aren't 100% certain, but this setup gives enough information for a confident decision.

Conclusion: for line-level power-tube saturation, it's better to use a conventional tube power amp -- perhaps one of the 5-watt EL84 circuits is ideal -- rather than a 1-watt tube power amp. But this conclusion is based on just 1 example of such a low-wattage amp.

Conclusion: Preamp distortion is absolutely essential for most Rock sounds. Power-tube squash is also essential, to provide a broad region of very gradual, innocuous, forgiving breakup. Speaker rounding, from a speaker driven over 1 watt, is also essential. For an authentic Tone, you *must* have these 3 elements; these are the important elements. You can do some tricks (iso cab, resistive load then re-amp) and still obtain perfectly authentic classic Tone, but you *must* have a conventional power tube, and preamp distortion, and a guitar speaker pushed past 1 watt. These are the three key elements which work together -- that is, they cascade to form the resulting Rock sound. The power tube and speaker do *not*, as I thought before, have to interact directly with each other. Instead, it should be pictured as a standard 3-stage additive or cumulative process:

o First, voiced preamp distortion (2 or 4 stage, many pedals and circuits)
o Then, fattened greatly by squashy power-tube saturation (key seems to be the rate of onset of distortion - Aspen Pittman's Groove Tubes classification of "hard" (sudden) vs "soft" (gradual).
o Finally, rounded and smoothed by guitar speaker, pushed over 1 watt per speaker
Also, mic placement and mixing/phasing affects the recorded sound.

Conclusion: the type of load for the saturating tube power amp affects the EQ at the amp's Spk Out jack, but does not affect the dynamics in any signficant or limiting way.

I did not recognize how essential preamp distortion is, before. Now I know it's pretty meaningless to talk about power-tube saturation as though it can stand by itself. Almost all good Rock sounds are necessarily some combination of preamp distortion, power-tube saturation, and speaker rounding. Imagine: o preamp distortion without power-tube saturation or speaker rounding (worst-case DI)
o power-tube saturation without preamp distortion or speaker rounding (high-power speakers)
o speaker rounding without preamp distortion or power-tube saturation (200 watt Marshall playing clean into a 25 watt speaker at about 20 watts)

Conclusion (from previous test) -- a miked speaker (the air in front of the cone) contains essential dynamics that are not present at the speaker terminals (that is, the Spk Out jack of the amp, or a Red Box speaker tap's Line Out jack. Speakers do not merely filter -- they add dynamics.

Technique: it works very well for emulation, to capture an isolated passage from an album, and record yourself playing the passage direct/dry/digital. Then route these to an A/B switch and set two digital players (MD deck, MD portable, or CD-Rs in CD players) to repeat track. Adjust controls and route connections while the decks play the passage (original and direct/dry cover) nonstop.

Design: I will probably build the large ~6x3x4' cabinet isolation box, because it will have nice side access doors to remove the cab (or combo amp) and position or remove the mics. You could even put two cabs in it for two instruments simultaneously. This will probably have a 2x12 cab, though it's sized for 4x12. My current box muffles well but two layers are very non-ergonomic and hard to design a door for.

Technique: I used the a/b/c Source Selector switches from Radio Shack ($15) to do interesting comparisons of amps as signal processors. One Selector box routes the input signal. Today's input signal was heavy rock style playing, bridge pickup humbucker, direct/dry/digital playing back from a Sony MiniDisc deck (sometimes from the portable MD player - both are practically perfect). To make this recording, I was careful to plug the guitar into the DOD EQ pedal, bypassed, to preamp the guitar to preserve high treble, unlike plugging into the MD deck's RCA In jack.

There are many little pitfalls such as that. You have to get good at making sure you are getting clean, flat, linear, 0 db response from all the gear.

The other Selector box routes the amp output to a load resistor. Earlier, I did A/B tests of a tube amp into a load resistor/amp/eq/guitarspeaker, vs. tube amp directly driving a guitar speaker, and put normalization EQ on the A branch, then double-checking by putting it on the B branch. I definitely concluded that the only difference was EQ - but I want to run this again since I have gotten better at measuring the range of squash for amps, related to dynamics. The resistor does produce different frequency response -- that's probably why Torres charges $45 for the Line Out option for the Tiny Tone amp (one of my test amps) -- so R/L/C might be a more accurate copy of signal that results at the amp's Spk Out jack when driving a speaker as load.

Conclusion: Tentative but strong conclusion: the amp doesn't care whether you drive a speaker or resistor - just a plain difference of eq results, not a reduction of tube squash character. All the character is the same, with a simple EQ change.

Conclusion: a raw speaker in blankets does not make much noise at all. You could create a load box with a muffled raw speaker. Weber is making a coneless speaker as load.

Conclusion: I cannot recommend the 6AQ5 power tube for a power-tube saturation pedal or for a saturation stage in the middle of an amp. I'm getting great classic power tube squash using an EL84 and a 6V6 combo, driving any load (iso cab, raw speaker in blankets, resistor) -- but the 6AQ5 sounds closer to a solid-state power amp in its alarmed fairly sudden breakup. If you design a power-tube saturation pedal, or use a head into a dummy load/power attenuator as a power-tube saturation pedal, or if you design a power-tube saturation stage in an amp, I would suggest a better result using a conventional power tube. I would guess that a preamp tube used as a power tube sounds halfway between a conventional power tube and a solid-state device, in its distortion character.

Conclusion/recommendation: I recommend a chain such as the following, for flexible, authentic, variable-SPL Rock guitar Tone, particularly for home studios:

compressor
eq 1 (for distortion/pickup voicing)
overdrive
distortion (need several models, for several internal inter-stage voicings)
eq 2 (for voicing between the preamp dist and power-tube saturation)
tube power amp (recommend conventional such as 15 watt, maybe 5 or 50 watts)
load (resistor, R/L/C, raw speaker, unmiked isolation cab, face-down cab)
line-level tap pot
eq 3 (powerful - eq's the cab itself)
time fx, post-processing (comp)
non-distorting amp (ss here is reliable, stereo, and flexible)
1 guitar speaker, muffled by gobos, isolation booth, or isolation box
mics (condenser possible since SPL is variable separately from Tone)

Ideally, this would be available as a software-programmable/customizable head, with 2 2x12 cabs. 2x12 cabs are great because you have the eq stage 3 to control the cab response to add bass, and they are stereo and portable. Consider buying cabs marketed as "stereo" -- flexible: you can drive just 1 guitar speaker in one of the cabs, or 1 speaker in the left cab and 1 in the right, to produce minimal volume while still pushing each speaker which is currently connected.

Design/technique: think of gig space and SPL in terms of "how many speakers should I hook up for this room". With two "stereo" 2x12 cabs you can choose per gig:

1 speaker in 1 cab (push at 2-15 watts)
2 speakers in 1 cab (push with 4-30 watts)
1 speaker in left cab, 1 in right (push with 4-30 watts)
2 speakers in left cab, 1 in right (push with 6-45 watts)
2 speakers in left cab, 2 in right (push with 8-60 watts)

Conclusion: there is no great need for 1 watt tube amps, for home studios. Instead, use a dummy load or muffled raw speaker, extract the Tone at line-level, then re-amplify to just the right SPL to push 1 guitar speaker above the noise floor and above the speaker's sterile zone. Use gobos or some form of isolation to block this speaker from being heard too much.

Conventional combo amps are a cheap package deal with little flexibility: just a 1x10 or 1x12 cab, insufficient EQ, only one preamp distortion voicing. They are not usually modular and you may have to wrestle with a combo to isolate its tube power amp, or preamp, or reverb -- separates, on the other hand, are inherently modular. For recording guitarists who want to do modelling and master Tone, it's better to use a modular setup, and racks remain a good idea for detailed control, though pedals can be a nice mix of flexibility and ease of use. It's a good idea to have 3 EQ pedals, comp (or two), several distortion pedals, and some time-fx processors and pedals, and many different-length guitar/patch cables with 1/4" jacks, and an assortment of connectors designed to interface 1/4" mono connectors with 1/8" and RCA.

Conclusion: artists use a variety of sounds. It's ambiguous what "the" Metallica or SRV sound is; each is actually a family of sounds. You can't simply emulate "the" Metallica sound; you can emulate the sound of a particular passage of an album.

Conclusion: when you use multi-stage eq including inserting eq between saturating power tube and guitar speaker, you can get mic sounds good enough to not need eq, even in an isolation cabinet.

So:
thumbs up for large iso cabs, and dummy loads (will re-investigate THD Hot Plate).
Thumbs down on 1-watt tube amps (need more test cases)
Thumbs up on power-tube saturation pedals and in-amp circuits which use conventional tubes and produce classic very-gradual onset of power tube saturation - not the "sudden garble" effect like solid-state power amps produce.
Thumbs down on having a single preamp-distortion voicing in an amp -- need a selection of interstage voicings, for emulation.

Need to A/B:
A: 4x12
B: 2x12 with eq between power tube and guitar speaker.


Amptone.com ultra gear-search page


Home (amp tone and effects placement)

Contact