Rerurn to Romy the Cat's Site

One more thing that I would like to point out in context of the recent Marc Mickelson’s review. Marc wrote:

“The Atma-Sphere amps are certainly more powerful than the ML3 Signatures, but not noticeably so unless you clip the Lamm amps. “

Here is where Mark, as most of his dim-witted and easy on mouth ruining colleges, just mentioned the clipping of ML3 as some kind of self-understandable fact. He, I am sure, had clipped ML3 with his Wilsons and knows what he is talking about. Sure, you dream!!! Those idiots-reviewers just toss the words without any meaning and most of the time without any understanding of what they are saying.  Anyhow, forget about the Mickelson – he was spreading his stupid weeping about ML3, trying make Lamm to give him the amp for so call “long term landing” – Lamm might just do so even despite to the fact the Mickelson sold his former “long term landing”  (Lamm L2 preamp). It is just a cost of doing busses in high end industry – somebody shit on your head and you kiss them in the ass juts because it is profitable…

So, as I said forget about the Mickelson the Reviewer - there is more interesting fish to fry. The subject of my attention is the Lamm ML3 clipping as I feel ML3 might be a very good tool to research and finally to answer the question of SET clipping. 

Let pretend that we have ML3 driving some kind of reasonable load and the signal voltage goes up. What would restring the power of the output and what would make this SET to clip? Living the various moments of bad and stupid design out we would have the following suspects:

1)  Input stage clips
2)  Driver stage clips
3)  Output  Transformer Clips

The stages might clips by current or by voltage but we presume that Lamm know what he does and ML3’s stages are properly powered, loaded and current or and voltage clip at the same time. So, the ML3’s input stage would be hardly a suspect of clipping as it sit on feedback and the cathode voltage will be lifted up with the output voltage goes up. The Output Transformer might be the one that clips first but we have no control or knowledge how it was designed and where it intended to begin to clipping. Then we have the Driver stage that might go into clipping – here is where the subject of ML3 become very interning.

Lamm made the ML3 driver with 4 paralleled 6H30 assuring that the driver stage can produce very high current. Why we need a lot of current in there?  Because the next stage has only limited amount of voltage on the GM70’s grid. As soon the positive voltage of the signal goes up and become equal to the negative voltage on the GM70’s grid then the tube enter to class A2 operation and develop currents. To navigate the tube further the source stage shell be able to battle the grid current with own current capacity. Having the 4x6H30 is a very powerful current source, so the ML3 has a good provision to have driver stage to be very clipping-resistant.

Now, let compare the clipping-resistance of ML3 and Melquiades. Melquiades driver stage is very current-weak-it has only 18mA. Melquiades was design to have absolutely different objectives then ML3. Melquiades use voltage driver stage – the tube with very high gain (35) that takes sub 3.5V of input stage (-4V on bios)  and with just one single amplification stags it might develops over 120V to drive the output stage. There are two reasons why Milq never care about currant.

First reason, the Milq meant to drive very high sensitivity loads. Take a look the Milq’s driver performance. It is the performance of driver of the bass channel, loaded to the bass section of Macondo and paling heavy bass music for 10dB-12dB louder than I ever play in ordinary.

 The 0V AC is at the bottom of the screen between the first and second square. The scope is set for 10V/div and 9 divisions up we have -92V of the bias voltage of my output tube. The positive AC voltage of signal runs up and doe but it never apaches the square 10, the rise of 90V what it would go into class A2. So, with the sensitivity of Macondo more current out of the Milq’s driver stage is just not necessary

The second reason why Milq never meant to have currant capable driver is because Melquiades using the 6C33C. Let pretend that that in my measurement above I loaded the amp to an equivalent resistor and drive it with even more voltage. Then the driver stage will send to output tube the voltage that would be equal to more than the 6C33C’s bias and the 6C33C go to A2 operation. In fact it was what I did and what happen?  6C33C responses very none linearly and clips quite fast. I was driving the driver tubes with 35mA and once even paralleled them trying to see how 6C33C would behave if the driver can push over the grid current. The 6C33C responded very stubborn and not very good. It looks to me that 6C33C hated grid current.

OK, how it all related to Lamm ML3. Well, Lamm is in a very different game as he has no two Melquiades’ reason I descried above.  Obviously Lamm would like his amp to be able to drive as wide range of speakers in tern sensitively as possible – in fact he knows that most of them will be in 90s dB not in 100s dB. And the second is that ML3 uses not A2-hating 6C33C but the A2-loving GM70. The GM70 is direct heated and it has no problems to work positive voltage on grid – many transmission tubes were designed to do it. So, the only this that Lamm case in his case was to build a good current supply for the GM70’s A2 operation – if it ever goes there… Here is where the 4 paralleled 6H30 come handy. BTW, many do not know it and fell that 6H30, or the BAT’ Super Tube is the wonderful discovery of Victor Khomenko from BAT. It is not the case. The 6H30 was suggested to Victor Khomenko by Vladimir Lamm before Victor knew the tune and the business -savvy Victor converted it into the BAT’s marketing success.

That all is relatively know and simple things but there is one big aspect behind all of it, the aspect that I do not know how to judge and it might be exactly where ML3 might be  wonderful tool to research. We have a tube that is capable and that “likes” the A2 operation. We have the driver stage that is perfectly cable to support the tube in the class A2. So, what would happen with Sound when the tube favorably inclined to the A2 conditions does enter the A2? This is VERY interning question and the Lamm ML3 might be a wonderful playground to research the subject. I do not really know if the A2m operation is bad by the definition or it is bad juts because in most of the cases it is just badly implemented. I still do not think that crossing between A1 and A2 would have no consequences to Sound but I never had the environmentally and methodologically clean condition to do this assessment, nor did I need it with 110dB sensitivity…  The ML3 would be a good opportunity to try answer this question but it would be too much for Marc Mickelson to ask. Not to mention that 4-other band bass woofer at the bottom of the Wilsons would certainly disqualify Mark from making any more or less lucid appraisal on the subject of class A2 entering….

Rerurn to Romy the Cat's Site