Ok, let dive a little bit further into my examination of the subject of Lamm ML3
What is important is to undusted that my view less targets the Lamm ML3 but rather it targets the ML3 as a representative of entire class of amplifiers. When I told that Lamm ML3 is “fundamentally improper design” I did not mean to critique the ML3 specifically but rather the entire family of expensive Single-Ended amplifiers where the “cost became to be a major design impediment”. I will try to explain it further. People who have brain understand that everything in audio is learning experience and I would like to describe why I feel that “expensive SET” dose not comply with my vision of a “correct direction”. It is juts my opinion and it based on my own experience. Your mileage might wary but where did you see me to care about your mileage? Anyhow, I have my opineon and I agree with it, so it comes…
For beginning I would like to observe a SET amplifier as a progression of quality and price, in this case prices will be very much an applied tool. We all know that SET is very simple amplifier: usually 2-3 stages, few parts, short paths etc… Let pretend that price DOES have direct relationship to the quality of amplifier (or sound), in fact why shouldn’t it? In today world a simple OK-performing SET made around “simple” 2A3, 45, 6C33C or 300B cost under $1000. There a number of Chinese companies that do chips SETs (under $500), AN does their 300B set for under $600 I believe, there are a number of American companies who do very inexpensive sets (Bottlehead, AudioMirror and many others). The all sound better or worth but they in fact closer to “better sound” (with an adequate load) then many $20.000-$60.000 solid-state monsters build by Morons with diplomas of electricians (a long list of the companies goes here).
Do these $1000 SETs sound as good as SET could? Of course not – there is a quite a room for Sound from here. Now we begin to add price/expense and we will have a more or less proportional growth of quality (if the designer knows what he does). The specials case when people employ phony “expensive solution”, like use of 0.001 precision resistirs, gold transforms, or the chassis made form Agarwood with platinum bumper around the amplifier I will leave out of scope, as those actions have no relation to Audio. So, regardless of the topology if your SET has more properly used chokes, better transformers, better power supplies, better amps design, better assembling techniques, better drivers (very critical), better protection and control, better tubes and many other factors then the amps could add quality and price. Still, any SET is basically is 2-3 dozens of parts and soon or later, regardless what you do you reach a “price threshold” where better parts, better supplies, better assembling techniques, better tubes do not add anything anymore within any MEANINGFUL scale. My estimate, based upon my experiences, that a self-cost of a pair of SET monoblocks at hits the “price threshold” level would be around $4000-$5000 for the amps that use low voltage output tubes and $7000-$8000 for amps that use high voltage transmission tubes. From here there is only one way to add quality and consequentially price to your SET – to improve further the only thing that maters in SETs: the output transformer (OPT).
The OPT is the main bitch of any SET and in the same time it is a subject of glory of any SET. Something very positive happens to Sound when goes through transformer, the key is that the transformer should be good and it arguably should be the only transformer in a unit. So, the OPT: the frequency response, articulation, dynamic, tone, inner-tonal connectively, balance between “resolution” and “space” and whatever else you might value in Sound came in it’s majority from the quality of OPT of SET. A person whose amps is at the “thresholds prices level” from here might jump into all imaginable OPT exotica, chasing in Sound whatever he fells needs to be chaise. A good OPT might cost a few thousand dollars it is all depends of the budget and the intentions/expertise/experiences of the person who design and who makes and who build the transformer. However, regardless how fantastic SET transformer would be it still hit its own topological limitations. To get bass you need inductance. I’m not necessary taking about better number of bass but rather about bass as quality. The OPT do not just need inductance to do bass but the inductance as “something else”. You can increase the core size as much as you wish to pay, building up inductance but inductance kills HF. It would be simple if the inductance kills juts HF numbers (there are ways to fight it) but the inductance kills HF “quietly”. Higher frequencies are opposite: they heed fewer turns and lower inductance…. So, people got into many different more or less sophisticating techniques (complicated core materials, tricky sectioning, intricate winding techniques and many others) to get out of SET’s OPT proper reproduction of boundaries. When people go into the high voltage tubes the situation become even more complicated as high voltage requires more isolation between the turns, which increases space within winding that builds up parasitic capacitance. Those capacitances dehydrate HF’s transients. When people listen all of those high voltage amplifiers with the “big tubes” they report “Big Sound”. (I call it the “Elephant Sound” – search my site I have written about it before). However, a nature or this “Big Sound” is dehydration of HF. Take any speaker, increase its tweeter output for a 3db and then place in front of the tweeter a soft hairy fabrics that would eat this 3db up. Now you will have a mimic of that “Elephant Sound” – that in fact is not the “Big Sound” but rather a sound with unevenness of subjective transients across the range (MF are “faster” then HF). So, retuning back to the transformer – it is very complicated to do the OPT transformer for any more or less mindful full range, that would presumably also sound properly. Very few people out there know how it might be done. Very view can actually implement it if they even told how it might be done and very few, if any, go for a recursive subjective assessment using PROPER LISTENING TECHNIQUES of achieved results when they build transformers. Still, making even a theoretically perfect transformer for a given SET any person hits the dead wall of the fundamental bandwidth limitations for OPT and it is imposable to fight them while keeping the aim to the exoteric properties of sound in the same time. Russians have a good old fairytale about a village person who caught a fox and decided to make a soup with it. He put the fox onto a pot but the tail was sticking out. He pushed the tail into the pot but the fox’s head moved out of the pot. He pushed the head in the pot but the tail went out…. The very same with OPT and people cook their foxes disregarding the fact that heads and tails in OPT are improperly cooked….
I have heard many SETs and all of them had issues with wide bandwidth, it is not that they did not have bass or did not have HF but when they try to get it something else turned to go wrong. The only one wide-bandwidth-interesting amp that I’m familiar with was Lamm ML2 (old production not the ML2.1). It is not perfect SET from the perspective of “as good as it could be” but it is “better then anything else that I heard”. The ML2 was the only amp that was trying to do HF and LF (still with it limitations as I learned eventually) and do not go apart in anything else at the same time. So, a cogent reader would ask: “Romy, if Lamm was capable to design the ML2’s OPT and get the result the he got with that amps then why you feel that he might not do the same good job with ML3?” It is an excellent questing and I have two answers: first is the Lamm’s attitude and second is the “fundamental design flaw of ML3”.
FIRST: THE LAMM’S ATTITUDE.
Lamm might do the OPT as good as the ML2 was but also there is nothing that prevents Vladimir to design the ML3 transformer in the way how he designed the ML2.1 transformer. Considering that GM70 transformer is many times more complicated and more expensive then 6C33C the question would be: where Vladimir would say “it is enough for them”. I have no doubts about Vladimir’s capacity to do interesting things but I have quite uninspiring knowledge about his decision making pattern. I would like do not go personally into the subject but stay “outside” and use only the publicly available facts. So, was the ML2.1 transformer “good enough for them”? Defiantly it was: the amplifier had glorious reviews, being sold nicely and generally considered a success model. It is ironic, that the very same person who brought to us that miserable “ML2.1 sound” (reportedly because of the “better” transformer) tries to impress to us the “ML3 sound”. You see where I am coming from? It is too simplistic to see that I appointed Lamm as my person bitch on my site but the realty is way different if to look at the fact. The facts are facts and the LP2, ML2.1 and L2 that came from Vladimir AFTER Vladimir made M1.1, L1 and ML2 are the facts and their performances are self-evident. If a person is an idiot and do not know better (many audio manufacturers) then they juts idiots. The problem is that Lamm is not an idiot and he knows better. Therefore the question with Lamm actions would be: will Lamm, as a commercial products maker, willing to do better? I do not know the answer and what I am observing later regarding Lamm makes me suspicious. I personally very much understand the Lamm’s attitude: “It is it good enough for them”. Sure it is. If I were a manufacturer then (considering my respect to the audio Morons out there) I would take it even further them Lamm does and I would make my amplifiers to explode like a bomb and to kill those “listeners”. Still, I am not a manufacture but a person who is critical about the fact of audio performance –all the rest is secondary. In the end, the ML3 will be as good as Lamm’s attitude would alow it to be, however the Sound of ML3 will be still absolutely and unarguably limited by the “fundamental design flaw of ML3”
SECOND: “FUNDAMENTALLY IMPROPER DESIGN OF ML3”
Here we go, here is the very major and the most important part in the entire article and I would return you attention back to the subject of price as an applied tool of SET design.
Pretend you have a perfect SET. Let presume that word “perfect” implies the unlimited amount of money, unlimited amount of design skills and unlimited amount of knowledge about the nature of sound and audio (very critical) that were used for implementation of that perfect SET. Since a complexity of SET is limited by its default simplistic topology you puff up that perfect SET with I would say $8000 worth of implementations. (we are still under presumption that prices = quality). Then you do crazy and designed an amazing OPT. Let presume that it is high voltage tube and you went for as high as $5000 for a transformer. Right now you have a stunning, full range SET. Let presume that this SET would be your Lamm ML3 for $125K, or AN for $165K or Wavac for $350K. So, you have a very good amp that cost you a LOT of money, a LOT of affords, years and years of learning about Sound and that your Perfect SET is able to address the most demanding requirements existing in audio reproduction and the most demanding your demands. Now, who would use that Perfect SET?
Now we hit a very interning point. Are we taking about juts an expensive piece of electronics that the short-penis boys buy in order to feel better or we are taking about the “most demanding requirements existing in audio reproduction”? If we are talking about using that “Perfect SET” to address some really demanding requirements then I am sure it won’t be a person who will be plugging that “Perfect SET” into a 30” tall single-driver mini monitor. It is most likely (not to say mandatory), that the person who can reach in his level of audio development the state of the “Perfect SET” would use as evolved rest of the system: the persons’ acoustic system will be as evolved as the “Perfect SET” is. It is my strong and firm conviction that any sound that worth more then a thousand dollars should come from a multi-channel installation where 4-5-6 channels produce sound within partial frequency range. It is complex, it is expensive, it is large, it requires a lot of efforts to make it sound PROPERLY (now a lot of people ever witnesses properly made and instated large multi-channel) but when it done as it should be then it capable to operate at the very same level as the “Perfect SET” - addressing the “most demanding requirements existing in audio reproduction”. Here is where the fundamental design flows of any “Perfect SET” (including the Lamm ML3) demonstrate its limitation in full glory: A SINGLE MORE OR LESS FULL RANGE SET AS NOT A DESIRABLE AMPLIFICATION TO DRIVE ANY SERIOUS MULTI-CHANNEL INSTALLATION.
Any single person who ever went for a PROPERLY IMPLEMENTED bi-amping and line-level crossovering never looked back, ever. An ability to writhe a perfect crossover curve with a perfectly predictable phase characteristic at the line-level is not imaginable at speakers level where impedance is “free running” (any attempts to lock impedance fuck-up sound, it is well know to any people who have ears). Furthermore, the amassing ability to manage damping for each channels, mange harmonics for each channel matching the EXECT need of the driver (by loading each output tubes differently) and at many-many-many other factors set any properly done multi-amping of multi-channel acoustic system very far apart form anything else. Now, tell me: what the hell a person with a capable multi-channel acoustic system would do with a single “Perfect SET”? The person system but its nature of being multi-channeled implies that it wants the best amplification within each channel. However, we have juts one SET, the one that was built like the cooking that fox on the pot with OTP the no mater what you do is juts compromise. I any “Perfect SET” 70% of its cost and efforts dedicated to be wide bandwidth… despite that any more or less serious multi-channel speaker requires…. narrow bandwidth for each channel. So, here is a clear waste of 70% cost and efforts in your use your “Perfect SET”.
A responsible question would be: why do not use 2 or 3 of that Lamm ML3 to driver a “better” acoustic system? Let again disregard the price, but even then, we might find that an amplifier like Lamm ML3 suddenly become as not attractive candidate for the task of milti-amping. What we need in order to peruse the “most demanding requirements existing in audio reproduction” would be different. We need a LF SET (it might be the very same Vladimir’s ML3) only with huge output transformer’s core mass and having very-very high inductance (many times more then Lamm has in his ML3), with huge among of capacitance in PS, with very precise ability to dial-in the tube’s loading, with as low output impedance as possible and with as much power as possible to get out the given tube (even switching in the end into A2). However, for HF we need absolutely opposite: low inductance and low capacitance in OPT, minimum amount of turns, sensitive “able to react” power supply with near minimal capacitance but large amount of inductance, capable for fast acceleration transformer core, heavier tube loading, limited power, no switching outside of the pure A1… etc, etc, etc…. What is the most remarkable that very same Lamm ML3 for instance still might be used but it would just need deferent output transforms. The Lamm ML3 for LF and Lamm ML3 for HF would have transformers for $400-$500 each but the level of performance of the theoretical Lamm ML3 DSET (Dedicated SET) would be so high that the full-range Lamm ML3 could not even imagine it in its dreams.
So, in the example above, looking at the way in which any expensive SET might be used in the real world of serious demands I hardly see any needs worth $100.000 SET. Quite oppositely, I feel that if the SET is more expensive then I would say $7.000-$8.000 then this SET should be divided on two frequencies-depended SETs (DSET), with price of $4.500 each. The Lamm ML3 with it’s (looks like) composite driver stage might most likly to pump a lot of current and it would be so rational to put in there a second GM70 on the same chassis with own transformer, driving it from the same driver stage. The PS and the rest “support” is easy manageable (ML3 has the filaments filters right inside of the amp instead of inside the PS). If ML3 were DSET then it would be phenomenally interesting to see what a direct-heated GM70 might do, that I would like to see the driver stage direct-heated as well – whoa the hell heed a half-ass solutions!? Also it context of DSET it would be much smaller space for the Lamm’s attitude as it would be a relatively not demanding OPT where Vladimir would have mush less space for his “it is enough for them”….
Sumizing my feeling about Lamm ML3 I feel that I would be an OK amplifier but at which level? Certainly it will be expensive amplifiers with Vladimir attempt to place with this SET one ass on many chairs. I was explaining all those sentiments to a friend of mine but he told me that I’m delusional. He said: “Come on Romy, people out there do not think this way. They pay a lot of money for a brown track to deliver to them and bid expensive loudspeaker of a ‘good’ brand name and they anticipate that by virtue of large investment in a “marketingly developed brand” they would automatically get an interesting sound in their listening rooms.” Sure, those people would perfectly accept ML3 as a solution. He was defiantly correct but I do not see those people because I am delusional but because I refuse to acknowledge them. I do it because the actual result that that those people get in your listening rooms are not worth attention. Well, perhaps I am delusional but as I said in the beginning of this post – I have my opinion and I agree with it.
I do see good prospective for Lamm ML3 to become a suitable solution for many listeners that have dull speakers and some money to spend. Strategically I feel it is wrong direction to use dead speakers and powerful amplification, - I more inclined to “live” speakers and low power amplification, but it juts me…I have also concern that 32W of ML3 is NOT enough for the most of 90dB sensitive speakers it is not about juts volume but “quality” – most of the 90dB sensitive speakers need 100W -150W of class A to demonstrate something worth attention… If you need evidenced lock yourself in a small closet with 85dB sensitive speaker and a good SET (Vintage Lamm ML2 for instance) and play music. You will have more then enough volume but the sound will be still crappy. There is more to sound then just power-sensitively-room size ratio but it is a different subject. In fact I have some feeling that building that slightly ridicules driver Vladimir in his 3 stage design Vladimir did some interspersing bluff. The 32W is not enough to drive most of the speakers out there. For instance the 95dB sanative Wilson (the Wilson customers are the people to whom Lamm would like to pith the ML3) are not as will with his ML2, ML2 might do much better (but there are not good commercial high sensitivity speaks out there). However, the 32W of ML3 are not there also. So, I think Lamm in ML3 might made higher gain (look at that high gain inpit stage on 12AX7) but with lower power. Let pretend that ML3 had as much as 30dB gain not those damn Wilsons will sound “loud”. You will ask but how about the power – they will collapse into clipping in the peaks? Well, it would be in fact very … good for Wilsons. The Wilsons have 4th Order Bandpas enclosure at the bottom and if you drive it all the way up then it sounds like…4th Order Bandpas. So, it might be the Lamm thought to “compensate” of to mask the problem of the “contemporary woofers” with the “deficiency of his new amplifier? The people out there are deaf and stupid anyhow, and they will listen anything… they bought their Wilsons at the first place….
Anyhow, I feel the ML2 will be warmly accepted by audio public, the same public that very warm accepts the Ml2.1, Kelll and Dynaudio. It will be another stupid review from another Moron with Wilson MAXX and we all know that will be in the review and what will be next. The audio public is mostly ignorant and has reference points of pterodactyls and demands of vultures - why they should not love ML3? The “big reviewer” Lars Feudal was listening his ML2 with 88.5dB sensitive Veruty Parsifal and was pitching that ML2 was the best amplifier in the word. How he came to this conclusion I had no idea – I head his room and never understood it. Mike from “Audio Federation” (courtesy to whom the images above) has written in his blog:
“Lamm did play their one working ML3 on one of the Wilsons and Neli got to hear it for quite some time…. summarize, she liked it - it appears to be SIGNIFICANTLY better than the ML2.1 in ALL audiophile attributes, especially control of the speaker, dynamics, and transparency.”
I have no doubts that it was exactly how he described, however the following facts might be taken under consideration:
1) ML2.1 is VERY unfortunate amplifiers
2) ML2.1 with Wilsons Watt/Puppy (and partulasy not the 6.3 but 8.0 is not juts unfortunate but nightmarish.
3) What dos it means “better”: more like ML2.1 or more like ML2?
In the end – if the Lamm ML3 will help to some people to discover how SET might sound then it will be fine – the ML3 extends this SET application for a 3dB of sensitively form where ML2 was. Neli, Mike, Vladimir and whoever else who accustomed to the funny Kharmas, Wilsons, JM-Labs and the rest of the mid 90dB sensitive load might eventually with the help of Lamm ML3 learn how a first watt of the vintage Lamm ML2 sounded with a proper over 100dB sensitive load. I am glad that Lamm electronics continue to be educational. I learned a lot from ML2 and a lot among the expressed in this post came form ML2 made me to think. Now it is up to the others to care the torch of the education and since ML2 is no longer available nowadays the Lamm ML3 might be for somebody a useful tool in his/her educational journey about audio. Still, the election of W Bush in White House Lamm celebrated with release of his L2 preamp – the first “idiosyncratic” product from Vladimir. In 2001 I called L2 as the “Electronics of the fist year of nothingness in America”. Will the ML3 become the “Electronics of the seventh year of nothingness in America”? The answer will be pending…..