The Evolution of the 100W Circuit: From JTMs to JMP Superleads
By Roe
Fremstedal ©2011-2012
This
article tries to give an overview over the early Marshall 100 watt amplifiers
by piecing together available information and shedding new light on
transitional models from 1967. The first 100 watt amps – known today as
“JTM45/100s” – used “JTM45” plexi faceplates and white “Super 100 Amplifier”
back-plates. However, the 100 watt PA amps used “JTM100” faceplates. In 1967 several
changes were made. First, the plexi “Superlead” and “Superbass” backplates were
introduced, then the so-called “Black Flag” “J.T.M.” plexi faceplate was used
in a transitional period. Finally, the “JMP” plexi faceplate replaced the earlier
faceplates. The main stages in the evolution of these early 100 watt amps are:
1.
Prototypes (1965)
2. Amps with dual output transformers (1965)
3. First amps with single output transformer (1966)
4. First EL34 amps (1967)
5. Second series of EL34 amps with dual rectifiers (1967)
6. JMP faceplate and new power supply (1967)
7. New Superlead circuit (1968)
8. New chassis and higher filtering (1969)
9. Yet another chassis, last plexis (1969)
10. Aluminium panel amps (1969-)
Famously,
the first 100 watt amps were made on request by The
Who’s Pete Townsend. The first prototype is reported to have used four 6V6
output tubes and a single GZ34 rectifier tube. It probably used only one Radiospares
Deluxe Output transformer, reportedly giving an output of 60 watt. The second
prototype reportedly used two output transformers, two GZ34 rectifiers, and
four 6L6 output tubes, whereas the third prototype used KT66 output tubes. Ken
Bran and Dudley Craven at Marshall eventually dropped the GZ34 tube rectifier, increasing
power and reliability. The output of the third prototype is supposed to have
exceeded 100 watt (Doyle, History of Marshall, pp. 33-34). However, it is uncertain whether this refers to power at
the point of breakup or maximum output.
The first
production amps used a
big Hammertone Radiospares power transformer (with taps for both 350VAC and 425VAC).
According to tech reports, these early amps have around 470VDC on the plates of
the KT66s (indicating that the 350VAC tap was used). These amps run at much
lower voltages than the later KT66 amps, producing less power while being
easier on the tubes. Also, it should be kept in mind that the voltages drop or
“sag” significantly under load, since the power transformer was only rated for
200mA (by comparison, a Twin transformer is 450mA). This power supply sag gives
a softer, more dynamic response with more distortion when running flat out.
As opposed to the prototypes, the production
amps used Drake output transformers. The early ones had ohms switches mounted
to the transformers, whereas the later ones had ohms switches on the backplates
like later amps had. The transformer was the same Drake 784-103 used in the 1965-66
JTM45s, although two transformers was used for a quad of KT66s. Like the
JTM45s, the early 100 watt amps ran the KT66s inefficiently by using a
high load. The Drake transformers have very high primary impedance (8K),
something that reduces power and bandwidth significantly. These 100W amps are
basically like a dual JTM45 but with a solid state rectifier, giving slightly
higher voltages. Whereas a JTM45 puts out 30-35 watt at the point of breakup,
these early amps put out around 70 watt at the point of breakup (and more than
100 watt at full breakup).
The choke (inductor) on these early amps
appears to have been the Radiospares 20 henry choke often seen on the early
Marshall schematics. The Radiospares choke is different from the Drake and
Dagnall chokes used later. The Radiospares choke is rated for 20 henry and
70mA, whereas the Drake and Dagnall are rated for 3 henry and 100mA. It also
has a have a high resistance (690 ohms) compared to the later chokes (110 ohms).
This high resistance gives more voltage drop across the choke than with the later
chokes. When running the amp hard, screen voltages (and preamp voltages) will
sag more, giving a softer response with more breakup.
A key element to element to the sound and
response of these amps lies in the power supply, especially the filtering
(smoothing). The preamp used a single 16uf capacitor (for V1 and V2) like the
JTM45s; whereas the phase inverter (V3) used 32uf (a JTM45 typically has 16uf).
However, screens filtering is only 16uf for 4 KT66s (a JTM45 has 32uf for two
KT66s) and mains filtering is 32uf (like a JTM45). This low filtering gives a
loose feeling and an open tone, great for blues and classic rock. Unlike the
JTM45, this amplifier use two capacitors in series at the screens and four
capacitors (series and parallel) at the mains. Putting capacitors in series
like this will increase the voltage handling but decrease capacitance (two 32uf,
450V capacitors in series gives 16uf, 900V). Capacitors were Radiospares 32uf,
450V. Today RIFA and BC Vishay 33uf are popular replacements.
The head cabinet on these amps use two narrow
Vox type vents (used until late 1966). The first amps used the rare block logo,
the later ones a standard “Marshall” logo. The indicator was the round yellow
Radiospares lamp used on the JTM45s, and speaker cabinets had Celestion silver
alnico speakers. Serial numbers are in the 6.500-6.600 range. Needless to say, this
amplifier is quite rare. It was recreated by Marshall as the 40th
anniversary stack (and Wallace amps did a version on JTM45 chassis). The only
known user of the original amps is The Who (http://www.thewho.net/whotabs/gear/).
In September 1965 a new output transformer, the Drake 1202-84, was introduced (2” stack; 4K
primary; 8 and 16ohms secondaries as well as a 100V secondary for PAs). This
transformer was designed for a quad of KT66s, placing the two 784-103 used
earlier. The power transformer was still the big Radiospares one, a transformer
mostly seen on the PA amps. And the choke appears to still be the Radiospares
20H. Some of these amps have the same power board layout as the previous amps
(#2), whereas others had the later layout, although filtering was kept the same.
Some use the round yellow indicator, although some use the square red indicator
seen on later plexis. The headcabs still used Vox vents. Serial numbers covers
the range from 6.600 through 6.800.
New power
board:
This
version used the Drake 1204-43 power transformer (3” stack) instead of the
Radiospares transformer. These amps run very higher voltages, something that is
hard on the tubes. Two original factory schematics show 560V at the plates (and
559V at the screen). Attempts to reverse engineer these amps, including the
transformers, have given similar voltages, although users of the originals amps
report everything from 530V to 625V. Compared to the earlier amps, these amps
have more power, headroom, and punch. With 560V on the plates, you get around
80W at the point of breakup. However, the design maximum voltage of GEC KT66s
is 500V, while absolute maximum is 550V (both screens and plates). In reality,
it is the screens voltage that is a problem. Nevertheless, the voltages sag under
load and the GEC KT66s usually work well in these amps. But new production
KT66s, both Chinese and Russian, has serious problems with high screens
voltages (anything above 520-530V is risky). New production tubes can even have
problems with the high preamp voltages in these amps (400V in the cathode
follower). Although this amp still uses the JTM45 preamp, voltages are
significantly higher, resulting in a clearer tone from the preamp and a punchier
response from the phase inverter.
These amps went from using the big Radiospares
choke to using the Drake 352-114 choke. The 559V shown on the original
schematics indicate that Marshall used the Drake (110 ohms) rather than the
Radiospares choke (690 ohms). However, this schematic, like many later
schematics, actually show a 20H (Radiospares) choke. Nevertheless, the relatively
high screens voltage indicates the use of a choke with low resistance (the Drake
352-114). As mentioned, the high resistance of the Radiospares choke leads to
lower screens voltages (also, it reduces current drawn by the screens). The Radiospares
choke is easier on the power tubes, giving more screen sag and a softer
response. Also, the bass will be a little tighter, and the tone somewhat
clearer, since the Radiospares choke forms a more effective filter than the
later chokes (being a 20H choke, whereas the later ones are 3H). However, the
Drake choke gives a rawer tone with more power and less screen sag. It simply
sounds a little more like the later plexis.
The power board of these amps looks like the
last amps with the Radiospares power transformer, and filtering is kept the
same. Schematics show the use of 270pf treble capacitor in the tone stack
rather than the 250pf usually seen, giving slightly more upper mids. Like the
later Bass and PA amps, the early amps (## 2-6) all have a warm tone with
plenty of bass. This appears to reflect the fact that these amps are derived
from a bass amp, namely the Fender tweed Bassman. On all these amps, it is
usual to turn down the bass when playing at high volumes (although some prefer
to decrease coupling capacitors on V1 and V3 to get tighter bass).
Like their predecessors, these amps had aluminium chassis, JTM45
faceplate, and the white Super 100 amplifier back panel. However, on the last
of these amps the aluminium chassis was strengthened by block ends cast in aluminium
(also found on late JTM45s). The head cabinets went from using narrow Vox vents
to using a single narrow vent (used until 1968), offering better ventilation.
Some used yellow, round indicators, others square red indicators. Serial
numbers go from the 6600 range through the 7100 range. Speaker cabinets used
Celestion G12M 20W at first. Later the G12H 25W was also offered (probably 55hz speakers).
Pictures indicate that both Cream and the Jimi
Hendrix Experience used Super 100 amps extensively in 1966-67. These amps (especially
#3B) are commonly thought to have been used on Fresh Cream and Are You Experienced – something that seems probable if we judge by ear (however, Disraeli Gears might have been ##4 or 5A). The Who is reported to have used all the early 100 watters
(##2-4) live on guitar, bass, and PA, and some of these amps may have been used
by Free as well. Also, Malcolm Young of AC/DC is reported (by Marshall Law
magazine) to prefer a 100W plexi with white back panel (alongside a Superbass).
According to Ritchie Flieger, the amp Malcolm used on Ballbreaker had 625V on the plates of the KT66s (suggesting it is a
#3B or possibly a #4A). However, the choke is rumoured to have been replaced
with a big resistor in order to reduce screens voltage and increase tube life. Finally,
Eric Johnson is supposed to have used several recreations of this amp over the
last years.
Partly as a result of this there has been a
renewed interest in these amps lately. The so-called JTM45/100 has gained
reputation for having great clean and distorted tones without being overly loud
or bright. Dimed, it can sound like Clapton on the Normal channel and Hendrix
on the Brilliant channel. It has been recreated by Marshall (100JH), Germino
amps (Monterey), Metropoulos Amplification (45/100), Wallace amps (BKW 45/100),
and others. It should be noted, however, that these recreations typically
reduce voltages so that new production KT66s can be used.
The 100JH is based on the Dickinson amp, an amp
used by Hendrix. This amp represents an anomaly since it has a 33K/500pf
tonestack, 32+32uf preamp filtering (with 10K resistor between the nodes), a big
10K resistor (not 8.2K) between screens and phase inverter, and 470R screen
resistors (instead of 1Ks). Basically, this is the tonestack introduced in 1968
with 1967-68 preamp filtering, and JTM45 screen resistors. Authentic parts are
used, except the big 10K resistor that seems to be a modern replacement. However,
it is not clear who made these changes and when they were introduced. In any
case, these specs will make for a slightly more aggressive amp.
Schematic: http://forum.metroamp.com/download/file.php?id=3106
In February 1967, the Drake 1202-119 output
transformer was introduced together with EL34s. When changing from KT66s to
EL34s, the primary impedance of the output transformer was reduced greatly
(from 4K to 1.75K). However, the new transformer still has the same stack
thickness and the same secondaries as its predecessor. It is not known whether
the impedance was changed in order to accommodate EL34s or in order to produce more
power. Although EL34 load lines suggest that a 1.7K load is appropriate for a
quad (and 3.4K for a pair), this presupposes voltages around 420V (as shown by
Randall Aiken). However, Marshall used the 1204-43 transformer,
meaning that the voltages are significantly higher. The combination of low
primary impedance and high voltages results in a very powerful amplifier that runs
the EL34s very hard (leading to redplating). However, good tubes (Mullard xf2
EL34s) and power supply sag alleviated this problem somewhat. Theoretically,
the change of output transformer would double the output, making it a powerful
amp indeed. In reality, the power is less than doubled although it exceeds the
115-120 watts produced by later EL34 amps. The new transformer results not only
in a louder amp but also changes tone and feel somewhat, making it more dirty
and aggressive. The harmonic overtones contain more 3rds, 5ths, and 7ths; it is
not as good for playing complex cords (e.g., E7+9) as its forerunners. This amp
excels at crunchy rock and light breakup. The clean sound is very good, but not
as great as the KT66s amps. Overall, the sound is very similar to the later
1967 amps, although slightly cleaner sounding.
With the exception of EL34s and the
1202-119, this amp is identical to its immediate predecessor. However, the Vox
vents are no longer used; all headcabs use a long narrow went, although its
thickness varies. It is uncertain whether all amps used 16uf filtering in the
preamp or whether dual preamp filtering was introduced. Although a few
pictures, reports, and recordings exist, this amp is very rare and has not been
reissued or recreated (except by the present author). Serial numbers seem to be
confined to the 7200 range.
These amps
are identical to its immediate predecessors with two exceptions. First, the
power board was changed to include 8 filter caps (like later 1967 amps), instead of the 7 caps on earlier
models. Instead of using a single 32uf capacitor at the phase inverter, two
32uf caps were used in series for a total of 16uf. This will lead to slightly more squish and compression but the amp will handle higher phase
inverter voltages without blowing up filter caps (this is a problem when amp is
turned on without power tubes). Second, the last of these amps – but not all –
used two fuses on the back panel like the later Superlead and Superbass amps,
making it easier to change the HT fuse. Finally, it might be mentioned that one
of these amps have 16+16uf preamp filtering, whereas another seems to have
32+32uf, something that anticipates the later 1967 amps. The higher filtering
will reduce voltages in the first gainstage (V1) slightly (by using a 10K
dropping resistor between the nodes) and it will give slightly tighter bass in
the preamp. These amps are typically Super PAs with serial numbers from the
high 6.600 series through the 8.400 series.
These amps were the very first Superleads and
Superbasses, yet the very last JTM45s. They have JTM45 faceplates, gold
Superlead or Superbass backplates (with two fuses) and block end aluminium
chasses. On the Superlead (1959) model, a 5000pf bright cap was introduced, giving
a brighter and more aggressive tone. With the exception of the bright cap, the
bass model (1992) used the same circuit as the lead amp – something that was
also the case earlier with 100W, 50W and JTM45 amps. The Super PA and Super
tremolo amps also used the same circuit, except the use of 470K mixer resistors
instead of 270Ks on the PAs. PA amps also left out the 500pf cap at the mixer
resistor (a cap that makes one channel brighter and the other darker).
The Drake 1202-119 output
transformer was still used on these amps, but a different power supply was
introduced with the Drake 1203-80 power transformer (2.5” stack). The 1203-80
has two 95VAC secondaries that are feed to two different voltage doubler (bridge)
rectifiers. After rectification each rectifier delivers around 250VDC for a
total of ca. 500V. Although this voltage is high, good EL34s can usually handle
it. The power board on these models have dual rectifiers and 8 32uf capacitor.
Also, the single 16uf preamp filter cap is replaced by a dual 32uf cap
(occasionally a dual 20uf).
These amps are the second generation
of EL34 amps with serial numbers in the low 10.000 series. Headcabs used both
the narrow vent found on earlier amps and the wide vent used on later amps. It
is uncertain is any of these amps had steel chassis like the later amps did. It
is not perfectly clear why Marshall changed the power supply significantly on
the 100W amps (but not the 50W amps). Although the new power transformer
(1203-80) has a smaller stack than its predecessor (1204-43), it still delivers
sufficient power while reducing the total voltages slightly (something that is
good for tube life).
Aluminum
chassis with new back-plate: http://www.amparchives.com/album/Marshall/100%20Watt%20Series%201965-1981/1992%20JTM%2045%20Bass%20100W/1992%20JTM%2045%20Super%20Bass%20100W%20SN%20SB10017/slides/9cb5_3.html
The
so-called Black Flag amps are easily recognizable by having a reverse logo “J.T.M.”
faceplate. These models are sometimes referred to as “JTM100s” as opposed to “JTM45/100s”.
These are the first 100W amps known to feature a steel chassis. The new chassis
made it possible to reduce cost while increasing strength. However, the
transformers were placed very close together, something that can lead to noise.
Except for the chassis and faceplate these amps are identical to the last
JTM45s. The output transformer was typically a 1202-119, but the very Black
Flags appear to use the Drake 1202-132 (introduced June 1967). The 1202-132 has
a 4 ohms taps instead of the 1202-119’s 100V tap. These transformers are very
similar but the 132 has slightly less bass than the 119. The first Black Flag
amps used 270K mixer resistors (like the JTM45s), whereas the last ones used
470K mixer resistors (like the later amps), resulting in slightly more gain as
well as more low midrange on the bright channel. Both Piher and Iskra carbon
film resistors are used, although Iskras became predominant in the late 1960s. Carbon
composite resistors are used less frequently than on the early amps (but some
Black Flags have a 1W carbon composite 10k tail resistor). The voltages on
these amps are recorded to be around 480V, slightly lower than earlier amps. Headcabs
used both the narrow vent found on earlier amps and the wide vent used on later
amps. All these amps had serial numbers in the 10.000 range.
Pictures show Hendrix with a Black Flag the
fall of 1967 and the spring of 1968. This amp can be seen the first time June 18
1967 at Monterey and appears to have been used extensively for dirty tones on Axis:
Bold as Love. It can probably be heard on BBC Sessions from the fall of 1967 and on “Catfish Blues” on Blues. Tonal similarities suggest Hendrix
used it also on “Voodoo Child Blues” (Blues),
“Voodoo Child (Outtake)” (Electric
Ladyland & Beyond), and “Voodoo Child” (Electric Ladyland). Another Black Flag user is Paul Kossoff, as
shown by pictures from Free
album sessions (found in the sleeve of Tons
of Sobs). Finally, recent pictures and videos show Eric Johnson with a Black
Flag, probably used for Hendrix inspired dirty rhythm tone.
Partly as a result of this, there has been
renewed interest in Black Flags amps lately. The circuit has been recreated by builders
on the Metroamp forum and by Germino amps (the early Fillmore Classic, no
longer made). Some users of the original amps (as well as users of recreations)
have reported problems with ghost notes at loud volumes. Some originals and
clones are modified with increased filtering that minimizes ghosting, but this
gives a harder tone and feel to the amps. Others have changed the grounding
scheme (especially the ground loop at the potentiometers) in order to deal with
ghosting. How these amps differ from the earlier and later EL34 amps tonally is
still not perfectly clear. Some find these amps to sound like the later 1967
amps, while others find them to have a unique dynamic response due to the dual
rectifier power supply. Needless to say, these amps are capable of great blues
and rock tones when properly maintained and set up.
New
faceplate and chassis: http://www.amparchives.com/album/Marshall/100%20Watt%20Series%201965-1981/1992%20Super%20Bass%20100W/67%201992%20JTM%20Super%20Bass%20100W%20SN%20SB10214/slides/amp13.html
Basically identical to the Black Flags, these
amps are easily recognizable by having a “JMP” faceplate and sometimes also a
polarity switch on the front (especially on US models). Like the last Black
Flags, these models used 470K mixer resistors. Both 1202-119 and 1202-132
output transformers are used. Serial numbers are still in the 10.000 range.
Still dual
rectifiers: http://www.amparchives.com/album/Marshall/100%20Watt%20Series%201965-1981/1959%20Super%20Lead%20100W/67%201959%20JMP%20Super%20Lead%20100W%20SN%20SL10228%20-%20Flemming%20Hansen/slides/14.html
As shown by
schematics, Marshall made changes again to the power supply: Instead of using two
voltage doubler (bridge) rectifiers, Marshall used one voltage doubler (bridge)
rectifier like the later amps have. So instead of feeding two rectifiers with
95VAC each, Marshall now feed one rectifier with ca. 190VAC, getting basically
the same voltage after rectification as before. This was initially done by
bridging two secondaries of the 1203-80 power transformer
(however, later transformers have only one 190V secondary with centre tap). A
1967 factory schematic indicates that Marshall also lowered the plate voltage
to 460VDC, something that increases tube life, reduces power while giving more
compression and distortion (Doyle, History of Marshall, p. 204). Power
transformers went from the Drake 1203-80 to the Dagnall T2562 laydown power transformer. The new power transformer was similar to the
Drake 1203-80 (delivering 180-190VAC) but had a thinner 1.8” stack and a shield
for reducing noise. Both the Drakes and the Dagnalls sag significantly under
load. Typically, the plate voltages drop 80-90V at full distortion.
The 1967 factory schematic shows four
32uf 450v mains filter caps like the earlier amps had. However, this
configuration is rarely seen on EL34 amps. Most amps with the new power supply used
only two 100uf main caps in series for a total of 50uf (although two 32+32uf caps
in series for a total of 32uf may have been used). These changes to the power
supply reduced the number of parts, and probably also reduced costs and
ghosting. The screens filtering was typically 16uf (two 32uf caps in series),
but some late 1967 amps have 50uf (two 100uf caps in series). 50uf gives a
tighter and faster amp, whereas 16uf gives a softer and smoother amp. 50uf is
great for heavy riffs, the 16uf for bluesy solos. Phase inverter filtering was
32uf on some amps and 48uf (16+32uf) on others. However, the phase inverter
filter cap is no longer place on the power board inside the chassis, but put on
top of the chassis, near the power transformer like the later plexis had (the
hole in the chassis was unused on ##5B-C).
On the last of these amps the preamp
voltages were reduced by increasing the voltage drop resistor between the
screens and the phase inverter (from 8.2K to 8.2K+10K). These amplifiers
typically used the Drake 1202-132 output transformer, although a few used the
Drake 1202-119 and the new Dagnall C1998 (1.5” stack; introduced August 1967).
The Dagnall has a crisper and tighter sound with a broader bandwidth than any
of the early Drakes. Since the all these early amps are bass heavy, I prefer
the 1202-132 for these EL34 amps. Chokes were typically the Drake 352-114,
although a few may have used the Dagnall C1999 (same specs but nylon
insulation) or even the 20H Radiospares choke. The negative feedback loop in
the power amp was often connected to the speakers jacks (instead of the 16ohms
output like the earlier amps), giving a rawer tone on lower ohms settings. Both
narrow and wide vents headcabs are used in this period. These amps were the
last to use serial numbers in the 10.000 range.
Pictures from 1968 and very late 1967 show Hendrix
with a JMP100 amp (##5C or 6). And pictures from October 1967 and later show
Cream with JMP100 amps with polarity switches (cf. Doyle, History of
Marshall, p. 36). Cream acquired new amps (##5C or 6) when returning to the
UK in October 1967 after a US tour. These amps can be heard on Wheels of
Fire and on BBC Sessions recorded
October 24 1967 and January 9 1968. The late 1967 amp (#6) has been recreated
by Metropoulos Amplification (10.000 series) and Germino Amps (Fillmore
Classic,). The Fillmore classic actually went from double to single rectifier (from
#5C to #6), suggesting that Germino prefers the later circuit.
Schematic: http://forum.metroamp.com/download/file.php?id=3115
In 1968 Marshall
changed the preamp towards the now famous Superlead circuit. The voicing of the
first gain stage was changed when the so-called split cathode preamp was
introduced. The normal channel cathode was now 820R/330uf, giving a very dark
and gainy sound. The brilliant channel cathode was 820R/0.68uf, giving less
gain in the lows than the normal channel. The voicing of the brilliant channel
was also affected by reducing the coupling cap from 22nf to 2.2nf, giving a
faster and tighter sound with clearer bass. The second gainstage (common to
both channels) was changed by adding a 0.68uf bypass capacitor to V2A’s
cathode. This results in more gain in the mids and highs, but reduces the
ability to clean up somewhat, giving a less open tone than the earlier early
amps. Famously, the tonestack was changed from 56K/250pf to 33K/500pf (560pf), offering
yet more gain. Coupling caps in the phase inverter were changed from 100nf to
22nf, giving a tighter and faster tone with less low bass. Some amps use 27K
negative feedback (like the late 1967 amps), whereas others use 47K (typically connected
to 8 ohms output). The latter increases distortion and volume while reducing
bandwidth and clean headroom, giving a rawer tone that is slightly looser in
the lows. The response of the power amp will be slightly slower and more
dependent on which speakers are used.
Transformers were typically Dagnall C1998
output transformer, Dagnall T2562 laydown power transformer and Drake 352-114
or Dagnall C1999 choke. The new Dagnall C1999 choke had the same specs as the
Drake 352-114 (3H, 100mA, 110 ohms) but the insulation was nylon instead of
paper. The last reported use of the Radiospares 20H choke is in this period.
In my mind, the change to the Dagnall output
transformer fits this amp, giving it a tighter and crisper tone under
distortion suited for heavier playing styles. These amps are louder, tighter
and more aggressive than its forerunners. Although the tremolo amps also got
the new lead circuit, the Bass and PA amps continued to use the old JTM type preamp
(although voltages were lower, since the dropping string used 8.2K+10K, not
8.2K). However, the power amp of the bass and PA amps followed the lead amps by
using Dagnall transformers and 47K negative feedback (while the phase inverter
coupling caps were kept at 100nf). These late bass and PA amps – both 100W and
50W – are similar to the earlier JTMs, although they sound rawer and feel
stiffer.
The power supply on the new
Superleads typically used 50uf at the mains (100uf caps in series), although
32uf is also seen (two 32+32uf caps in series). Filtering on the screens was
typically 16uf (32uf caps in series), but 32uf is also seen (two 32+32uf caps
in series) as is 50uf (100uf caps in series). Phase inverter filtering went
from 32uf or 48uf (32+16uf) on the first amps to 80uf (40+40uf) and 100uf (50+50uf),
giving less squish and compression at higher volumes. Plate voltages are
reported to vary from 460V to 490V. The headcabs have a wide vent like the
later amps, giving better ventilation. Serial numbers are in the 12.000 range. Finally, speaker cabinets changed in
this period. Grill-cloth went from pinstripe to basket-weave. And power
handling was increased from 20W to 25W on the G12M and from 25W to 30W on the
G12H. The old cabs seem to sound slightly woodier and brighter than the new
ones. But the new cabs fit the bright, aggressive tone of Superleads.
The most famous user of the 1968 Superlead is
Edward Van Halen (especially on Van Halen).
Techs servicing the amp report a stock circuit with reduced negative feedback
(100K at 4ohms), Sylvania 6CA7 output tubes biased hot and a Variac (used to
drop volt voltages). This setup produces a more distorted tone than a standard
Superlead. The negative feedback is modified to 1970s specs, giving a rawer and
looser tone that fits the hard sounding Sylvania 6CA7s. The 1968 Superlead is
also likely to have been used by Hendrix in 1968 and later. Today, this circuit
has become very popular and several builders offer recreations of it, notably Metropoulos Amplification (12.000
series) and Germino Amps (Headroom 100 lead).
Schematic –
preamp:
This
version had the preamp cap on top of the chassis (like the later amps) but retained
the cap board inside (like the earlier amps). Probably
made from July to Sept 1968.
Preamp
filtering on top of chassis
In late
1968 a new chassis was introduced: The output transformer was moved away from
the power transformer and rotated in order to decrease noise (especially hum).
The preamp filter cap was removed from the preamp board and mounted on the top
of the chassis near the choke. Preamp filtering was increased to 50+50uf, giving
a stiffer feel. Filtering was also increased to 50uf on the screens, giving a
tighter bass. The power board was removed, and filter caps were moved to the
top of the chassis between the output and power transformers. Also, 56K bleeder
resistors were added to the screens filter caps. This ensures equal voltage
drop over the caps, preventing them from blowing up and preventing you from
getting a shock when opening the amp. However, these resistors get hot and can wear
out after many years. If you need to replace them, try to increase the value to
ca. 220K in order to increase resistor life and to minimize ghosting. Amps from
this period often have serial numbers in the 10.000 and 12.000 ranges, but
these serial numbers should not be confused with the 1967 and 1968 amps using
the same range of serial numbers. The 1969 Superleads typically use the
abbreviation “SL/A,” whereas the earlier amps had “SL”. Finally, the voicing of
the brilliant channel was changed, probably during 1969 (although the exact
timing is unknown). The cathode resistor went from 820R to 2.7K, giving a
crisper tone that cleans up better. These amps are associated with Jimmy Page,
Hendrix, and Eric Johnson.
In 1969,
yet another chassis was introduced with a standup Dagnall power transformer.
However, the earlier chassis (#8) was still used frequently.
The plexi
era ends in the middle of 1969 when the aluminum panel amps are introduced.
Around the same time the front-lip
on the headcabs went from a flat top to round top. The aluminium amps are very similar to the plexis, although some things
changed: Around 1971 the C2668 Dagnall output transformer was introduced. This transformer
has nylon insulation instead of paper insulation, sounding brighter and more
aggressive. At the same time mustard caps were often replaced with Philip
“chicklet” caps, giving a brighter and more aggressive sound again. The
negative feed was decreased to 100k at 4 ohms, giving a very raw and distorted
tone that is a little loose in the lows. This change also made the 0.68uf
bypass cap on the second gainstage (V2A) virtually unnecessary; since the amp
already had plenty gain (still a few amps had both the 0.68uf and the 100k). Also,
the voltages are very high on the early 1970s 100W amps (often around 520v),
something that is very hard on the EL34’s screens, although it makes the power
amp run a little cleaner. One easy way to improve tube life on these amps – and
the later 2203s – is to decrease the 1M resistor on V3A to 390K like Steve
Grindrod did on the Park 1210. This mod will improve the balance of the driver
under heavy clipping, often preventing redplating on V4 and V5 when running
flat out. It should also be mentioned that print circuit boards (PCBs) are
introduced in 1973. Finally, speaker cabs change from basket-weave to checkerboard
and speakers evolve from Greenbacks to Creambacks and Blackbacks, gradually
sounding brighter and more aggressive. These amps are associated with AC/DC,
Thin Lizzy, and many others.
The present
text represents an attempt to reconstruct the evolution of the 100 watt amps by
using schematics, layouts, pictures, databases, and different tech reports.
Information is based on examining original amps, on circuit analysis, and
attempts to recreate the various circuits. Although some of these amps are
relatively well-known (##2, 3B, 5B, 6-10), a comprehensive overview of
technical details and how the different amps sound have been lacking. The
present text represents an attempt to piece together existing information in a
way that is useable for guitarist and historians, repairmen, collectors and amp-builders.
The evolution sketched above represents a typology
and chronology, but there is likely to be some overlap between the different
models since different variants could be made simultaneously for shorter periods of time. The dating
of the different models represents estimates based on data-codes (notably on
mustard caps) and when transformers were introduced.
Like the 50W amps, the 100W amps sounded
progressively brighter and more aggressive over the years, making it easier to
cut through a dense mix. As is the case with 50W amps, the 100W amps can be
grouped into two main categories. On the one hand we have the post 1968 lead
circuit with an aggressive tone, on the other we have the JTM and Bass circuit
with a warmer and cleaner tone. The former is good for heavy rock, the latter for
blues and classic rock. I
feel that the latter circuit works the best with the JTM power amp and Drake
transformers. The 1966
Super amp (#3B in particular) stands out as a great achievement, although 1967
amps are also great. For more aggressive tones I prefer the 1969 Superlead,
while others prefer the response of the 1968 circuit. Needless to say, no
single amp can reproduce the tone or response of these different circuits
without throwing out the baby with the bathwater.
ACKNOWLEDGEMENTS
Writing
this article would be impossible if it were not for input from others. Thorleif
S. Hoff, Larry Grohmann, and Randall Aiken deserve special mentioning. I have
depended on databases at Amp Archives and Marstran as well as the Metro,
Marstran, Ampgarage and Plexipalace forums.