Mullard ECC83 (12AX7) Reissue vs Original – A Physical Comparison
by Matt Lachesky
Editors note. What follows is a reproduction of Matt Lachesky’s (Lachesky Amplifiers) excellent post on Marshall Amplifiers discussion forum in August 2011 describing his in-depth comparison of the physical characteristics or the Russian ‘New Sensor’ reissue 12AX7 (ECC83) with the original Mullard ECC83 manufactured in Great Britain. During their history Mullard produced several variants of their ECC83 tube including long and short plate types. The ECC83 tubes dissected here are of the long plate variety.
I recently had a 1962 I61 Blackburn Mullard quite literally blow itself up in my amp. Not wanting to miss a great opportunity for an experiment, I looked for and acquired one of the new ‘Reissue’ New Sensor “Mullard” 12AX7′s (Thanks again Joe!). I’m sure the tonal comparisons have been done a million times over, but I doubt many people have dissected both tubes to compare, side-by-side, the actual construction of the tubes. The lengthy post is my notebook from doing just that (and it’s got pictures!)
After gathering my tools (I ended up using a lot more than I planned on, though) I took a starting picture of both victims. As you can see, the Mullard (which is what I’ll call the real one) is shot, with the glass shattered around the bottom and the vacuum obviously compromised (thus the white getter). The Reissue (which is what I’ll call the reissue “Mullard”) looks to be healthy, but didn’t test as so.
The first part of the tube encountered is naturally the glass thereof. This isn’t a really fair comparison since the Mullard had already broken itself, but it’s worth mentioning.
The Mullard was already broken, but still put up a decent fight when I clipped my way around the break in the bottom to free the structure from the tube. It didn’t shatter as violently, but made a lot more smaller chips and shards. It was indeed thinner, measuring 0.0275″ at the same point on the side of the tube.
By placing the edge of a triangular file on the tube just below the bottom mica and hitting it with a hammer three or four times, the reissue tube shattered. The getter took an unusually long time to turn white (I’ve opened up several old tubes before), and the glass seemed rather thick. The best measurement my calipers gave me was exactly 0.04″ measuring a piece from the side of the tube.
Overall, the glass doesn’t really say much about the tube. The heavier getter flash on the reissue did take longer to fade out, but by the point the whole thing is used, it doesn’t matter anyways as the tube is history. The thicker glass is probably needed to meet some safety standard, and again has nothing to do with the tone of the tube. The only benefit I can think of is a possible increased durability in terms of physical abuse.
The getter is the part that makes the silver getter flash on the tube. It’s really got nothing to do with tone, and merely serves to keep stray gas molecules out of the way of the operating vacuum tube. The comparison here is for flat-out accuracy of the reissue.
The Mullard has a typical round getter, I believe known as a halo getter (please correct me if I’m wrong). Attached to an upright copper post, which is mounted to the corner tab of the plate.
The reissue getter is nothing like the original, and looks more like a saucer than the ring of the real Mullard. It’s mounted on two thin wires to either tab of the plate.
Again, the getter doesn’t do anything to the tone of the tube, the comparison here is just how accurately they copied the real Mullards for the reissue. Clearly, no attention was paid to the getter style.
The Micas are the white-ish things at the top and bottom of the plate which hold the structure of the tube in place and against the glass. Tonally, all the mica will do is control how likely the tube is to go microphonic (which also is affected by plate size). This is mainly another comparison for reproduction accuracy.
The Mullard’s top mica, measuring 0.020″ thick, is round with 8 small ‘arms’ that hold it in place against the glass, offering 8 points of contact [editor's note: all Mullard plates are made like this as far as I know] to stabilize the plate structure of the tube. The bottom mica, again 0.020″ thick, also has 8 little contact points, 180 degrees out of rotation from the top mica, which provides support from all sides when in the glass. In all, there are 16 points of contact with the glass, all evenly spaced around the circumference of the structure.
The reissue top mica, measuring 0.017″ thick, is a rounded square, providing 4 points of contact with the glass (the corners) the stabilize the structure. The bottom mica, measuring 0.021″ thick, is the same shape, aligned with the top mica providing another 4 points of contact, however they are directly in line with the top mica, as opposed to ‘bridging the gap’ as in the Mullard.
The main function of the micas is the support the structure of the tube and keeping it from moving around inside the glass. Clearly the original Mullard did a much better job of this, offering twice as many points of contact, all of which were evenly spaced around the perimeter of the tube. Microphonics are much more likely with the reissue due to its poor mica design.
After removing the top mica (something easier said than done) it is possible to look directly down into the structure of each half of the tube. This makes it possible to see element spacing and other aspects of the design that can affect the performance of the tube, and thus its tone, significantly.
The first thing to make itself obvious in the Mullard is that the grid wire is supported by two copper posts, and is very precisely located around the cathode as to be extremely close, but not touching it. The cathode is perfectly round. The plates have a definite indentation, getting them closer to the grid and cathode but allowing them to be full-sized for dissipation purposes.
Compared to the Mullard, the reissue is poorly built, to say the least. The grid wire is supported by two metallic posts, and seemed less accurately wound (in-depth analysis of the grid comes later). The cathode is oblong, and crumpled on the ends from hasty crimping to the top mica during production. The inside of the plates are a perfect rectangle, with no indentations or protrusions, and in general seem farther away from the cathode and grid assemblies.
Overall, it is clear that internally, the components of the Mullard are of better quality and design than the reissue, which seems quick, cheap, and boxy. The differences here will be explored further later, with in-depth examination of the Plates, Grids, and Cathodes.
The plates are the main part visible from the outside of the tube, and are what receive the electrons emitted by the cathode. Their quality greatly affects the performance and longevity of the tube.
The Mullard plates are extremely consistent. The halves are crimped together, folded over, and have two tabs at the top and bottom which mount them to the micas. They have a sort of ‘ladder’ design in them [editor's note: all Mullard plates are made like this as far as I know], providing ridges for some purpose. The material measures 0.0042″ thick, and internally, the measurements are approximately 0.1335″ by 0.2550″. The halves are well connected, and have a small hole (0.064″ across) on the side for ventilation, presumably.
The reissue’s plate structure was much harder to remove from the bottom mica than the Mullard, but the improvements end there. As stated before, the inside of the plate structure is completely rectangular, with none of the intricate indentations of the Mullard. The material is 0.006″ thick, but shows signs of heat damage on the inside surface. The coating of the plate is also much less robust than the Mullard. The plate halves are crimped onto two upright posts, which then secure the plates to the micas. Inside dimensions are roughly 0.1379″ by 0.2641″. The plates have three marks of slight ribs, but no holes or gaps.
Overall, the plates of the reissue are boxy and simple compared to the Mullard. How the precise shaping of the plate affects tone is something I don’t know, but being that the signal comes out of the plates, there will be some slight effect on performance of the tube.
The grid is what the input signal to the tube goes on to, and what regulates the flow of electrons from the cathode to the plates. The quality of the grid is very important to the quality of the functioning of the tube.
The grid in the Mullard tube is supported by two copper posts, and is made of an extremely fine wire wrapped around them (Calipers say 0.0015″) in a tight manner, with no more than the thickness of the wire between any two rows. The wire is soldered to the length of both posts in an even manner. Despite being made of such tiny wire, the structure is fairly rigid.
The reissue grid is supported by metallic posts, presumably made of the same metal as all the other metallic components in the tube. The wire isn’t wound as neatly onto the posts, and measures 0.0015″ thick. There is more space between windings than in the Mullard, and only appears to be attached to the length of one of the two posts. It seems much more flimsy than the Mullard as well.
Overall the grid is a very important part of the tube, functionally and tonally. While it makes a good effort, the reissue grid doesn’t live up to the real Mullard grid in terms of rigidity and accuracy of the winding. The slightly thicker wire of the reissue’s grid also will add some capacitance to the grid, affecting the performance of the tube.
The cathode emits the electrons that later are picked up by the plate (after regulation by the grid) to amplify the signal. In a cathode follower, it sends the signal out into the rest of the circuit.
The Mullard cathode is perfectly round, with a thick, even coating (except where it got scratched removing the grid). It measures 0.0319″ across.The material itself is only 0.0032″ thick. For a small tube of thin metal, it is again very strong.
The reissue cathode is oblong, with a relatively thin coating. It is 0.035″ wide on the smaller axis, and 0.048″ wide on the larger. The top of the cathode is crimped shut, presumably trapping heat. The material itself is 0.011″ thick.
Overall the cathode of the reissue came nowhere close to the quality standards of the Mullard. Maybe the figured nobody would ever see it to know?
The heater resides inside the cathode, and heats it, causing the emission of electrons by the coating. The heater doesn’t have much effect on tonality of a tube, but it does control the overall does-it-work aspect. This is the last component to be examined in detail (pin connections are kind of a yes-or-no sort of thing, and clearly the reissue ones work).
The classic flash-on-startup of Mullard tubes is due to the heater’s lower resistance when cold, or something like that. The heater pretty much is a piece of wire with an electrically insulating coating on it (so it doesn’t short to the cathode). One triode’s worth is about an inch and a half long. The wire itself has a diameter of 0.0059″ and, with the insulation, 0.0105″
In the Reissue, each triode gets about 1 3/8″ of heater. The wire itself has a diameter of 0.002″ and, with the insulation, 0.0110″
It seems they got it right at least here, the heater is the simplest part of the tube, but they managed to get something really close to the Mullard in the reissue. A heater’s pretty hard to screw up, though…
I’ll let you draw your own final conclusion, but I think the end result of the comparison is pretty clear as to the accuracy of the so-called “Mullard Reissue” by New Sensor. I’m sticking with my NOS. Thanks for taking the time to read all of this, I’m going to go wash my hands (who knows what kind of chemicals are in these things) and have some ice cream… I’ve been standing at my bench for two hours -Matt
(The answer to the little trivia at the end, in case somehow you didn’t get it: The real Mullard is on the left, the reissue on the right)
Editor’s Final Note
I can’t help but wonder if New Sensor purchased the trademark of Mullard merely for brand marketing purposes with no real intention of creating an authentic reproduction or reissue of the original ECC83. It saddens me that they have missed a fantastic opportunity to honor Mullard’s glorious name and legacy – all the people, the engineers, scientists, chemists, draftsmen, technicans and staff who worked to make fine-quality vacuum tubes over the decades. Mullard was a great British Company, a manufacturer and also a research institution that was constantly endeavouring to improve and develop vacuum tube technology. I feel this is what is lacking today in modern tube manufacture – the will to push forward, to improve and achieve the absolute best in terms of materials and design ethos. I’m sure I’m not the only guitar player that’s disappointed and disenchanted. I would love to see genuine good quality reproductions of Mullard tubes with some heritage being made again. Until a tube manufacturer realises this there will be no substitute for 1963 technology.