Groove Tubes work in progress - some secret stuff ....
|Many folks know Groove Tubes for it's vacuum tubes. Some know about their Custom
Shop series of guitar amps, SFX systems, and cabinets. Some know about their studio
microphones (their MD1 replaced the U47 in many studios), and some know about their
high end studio gear such as the ViPRE.
This is what is on the horizon .... works in progress ... a sneek peak ....
Some of these things will be shown at NAMM 2004.
For more detailed info in many cases, follow the menu bar below.
Click on the photos if you want a larger image.
This is Mr. Bill .... our head scientist. Bill Isenberg has worked in the "electronics
industry" for decades. Chances are, if you have any sort of high end hi-fi gear, radio
gear, or studio gear, this fellow may have had a hand in it's development and creation.
The USA GT-6CA7 ... still a rumor out there, but here are some of the real ones in a
750 volt B+ amp, a Music Man after running some tests. These are not just a big bottle
EL-34 as other folks are selling as 6CA7 tubes today. These are the real deal, off the
original tooling as used by GE. These also have an active beam forming element, a
critical item missing in the "6CA7" / EL-34 big bottle "conversions" sold today by some
other vendors. With any luck, these will be shown at NAMM. You may even see an
NVM (new vintage manufacture) 12AX7!
More of GT mad science in progress in the below photos - click on any of them for a larger image
Maybe you can guess what is in the works .... The Brick is already on the GT website, so that is not
Development and tests below of new GT 12AX7 "M" "NVM" (new vintage manufacture)
NOS Telefunken curves recorded for
reference. Mullard curves are also
recorded, as are 12AX7C, ECC83S,
12AX7R2 (LPS), Sovtek 12AX7R3
(EH), NOS RCA, Sylvania, Tung Sol,
Raytheon, Mazdam Amperex, and RFT
JJ ECC83 recorded curves and data.
This was as close to 12AX7 specs
expected as could be chosen, and was
an standard spec tube, even though
these are rare from the factory.
GT 12AX7M III.
All tubes tested at the same scales,
settings, and voltages from the above
GT 12AX7M I and 12AX7M II on dual
shelf module on Tektronix 570 curve
GT 12AX7MIII tube 3-3 life tests. Please excuse the
fuzzy pix ... no tripod.
This tube is being subjected to 350 volts with the bias
adjusted to provide 4.5 milliamps of current. This is far
in excess of design spec. I am subjecting the tube to
very high stresses, and when this is complete (if the
tube does not fail), the voltage and bias will be reduced
to 250 plate volts and a -2 volt bias to see how far the
tube has dropped from it's original recorded specs and
curves when it was new.
Tom Dunn had one of the Series M3 tubes in hard live
use, and there was a slight drop on the tubes return,
hence more excessive punishment was needed. On
Tom's other two tubes, there was no degrade noted.
Tom was given a new set of tubes to try. On the M1 and
M2 no change was noted. On other M3 tubes, no
change has been noted so far. On this current M3
(Tom's original M3) the A side of the tube is being tested
A: 2.0 / 2130
B: 2.2 / 2150
Upon Tom's return:
A: 1.8 / 2060
B: 2.1 / 2150
12/24/03 1030pst specs:
A: 1.5 / 1860
B: 2.1 / 2120
B side stable
A side dropping at very
high voltages and current
requirements. Other M3
tubes have not had this
area of concern. M1 and
M2 have had no issues
noted at this point.
In testing various tubes, new made and NOS, it is noted that due to tube design, materials, or other factors, one
reason some players prefer NOS tubes such as the Mullards and Telefunkens may be due to their increased
touch sensitivity, touch dynamics, and harmonics. There are differences in just about any two items; a $10,000
car may not have the same level of build quality or components as a $50,000 car. The same goes with tubes.
What is noted in some of the great NOS tubes of the past, are a few factors that at times are missing today. The
cathode materials are more pure in NOS items generally. Bits of junk in the cathode material emit gases down the
life cycle, and are less efficient from the start. These gases may not support the degrading of the internals, but
may lessen the vacuum perhaps. The plate materials may be of a single layer rather than a multi clad material,
and may be of a different metal formulation. Doing these things right will cost more money, but this cost is offset
to many folks in a more articulate tone and in much longer tube life.
Looking at some of the tubes of the past such as the Telefunken, Raytheon, Mullard, and Tung Sol, it is noted
that these tubes do share aspects of tight specs, long life, and beautiful curves and rise time.
The Telefunken of the past is not the same tube as the Ei produced long smooth plate item, although there are
some aspects shared. The materials are different, and the level of quality differs. This is one reason the Ei has a
higher incidence of physical microphonics. This is an example of trying to copy an original design in some
What has been difficult on the GT "Mullard" is the reproduction of the compression cathode plates and fittings.
This is one aspect of this tube that cannot be overlooked. This is also one reason these tubes were so resistant
to physical microphonics. This design aspect is not shared by other tubes.
The sensitivity of the Mullard, Telefunken, Raytheon long plate, and some other NOS greats is one BIG area that
separates the old tubes and the current ones. Sensitivity = touch dynamics, and other things. Basically, a
sensitive tube requires less input change in voltage (your guitar pickup) to make a change in output of the tube.
A sensitive tube will follow subtle player dynamics much closer than a less sensitive tube. This is one aspect of
the main are of "magic" in these tubes.
In testing, the new "GT-12AX7M" has the same sensitivity as the originals. As far as today's sensitivity goes, the
worst tube out there may be the 12AX7LPS, but these may be great as phase inverters for some folks or for 2nd
gain stage functions. It is all personal taste. Maybe you want to cover your mistakes? The JJ ECC83 is about
medium in the sensitivity area, but will degrade over a somewhat short period compared to many NOS items. The
current 12AX7EH will degrade even more quickly.
I will be doing a life test on the current ECC83 JJ with graphs at a later date below.
In any case, these "NVM" 12AX7M tubes are performing as the originals as of my testing as of 12/24/03.
GT-ECC83S (JJ ECC83S)
under life and stress test at
340 plate volts at start of
test after normal spec burn
in at the left:
A: 1.3 / 1500
B: 1.1 / 1490
1/7/2004 GT GE 6CA7 Current
Ref at RCA specs 100 mA expected in Triode
13.5 bias Triode 400 / 36 current
1 97.3 28.3
2 105 34.3
3 99.2 30.7
4 106.1 33.8
5 91.2 30.0
6 101.4 34.5
7 120.4 56.4
8 116.4 40.6
9 91.2 26.7
10 99.8 31.5
Average 102.8 34.7
E34LS #5 99.4 27.0
National 89.1 27.9
Tests against Sylvania NOS 6CA7, Philips 6CA7, and original 6CA7 showed the NVM 6CA7
to have curves most close to the GE and Sylvania in triode mode. In Pentode mode,
the average tube (#6 above used) had a slightly faster rise time than NOS GE
Due to the higher vacuum of a 6CA7 over a soft vacuum EL34, these tubes should have a
much longer life in guitar amp usage than an EL34. In life tests so far, at least a
40% increase could be easily expected. These tubes may have a very strong appeal to
players wanting more clean headroom in Marshall style amplifiers and may be sonically
preferred over the 6550 where folks use these. Listening tests are in progress.
Tests in the Musicman 750 plate volt amp are in progress.
Tubes in triode mode test as expected, in pentode mode they show the characteristics
of an active beam forming element. This is NOT the case on today's new so called
6CA7 tubes by other vendors. These "new" copy 6CA7 tubes are nothing more than the
companies EL-34 in a big bottle without the active beam forming element, and
generally show performance inferior to the same companies EL34.
These NVM GE 6CA7 tubes have a very different bias than any other EL34, including the
E34LS. The E34LS generally needs an additional 1.5 to 2 volts of increased bias over
and EL34. These will require even higher bias voltage. Bias info should be included
in tube packaging as is currently done with the GT designed E34LS tube.
These NVM GT GE6CA7 tubes are even stronger than the current GT E34LS. In something
like a 100 watt Marshall, I would not be surprised to see an average amp with a good
power transformer bench 170 watts.
Click on photos for a much