I tried listening for a while this morning, figured the electricity was not good enough.  When I went to switch off the amps, I immediately noticed that BOTH 5651s had failed spectacularly, with cracked bases and "white-out" on the tops of both tubes, IOW, no vacuum.  Weird, and weirder that it did not make it sound any worse than bad electricity.

I'd first suspect modern "value engineering" at its finest, but these are not modern tubes.

I would also suppose that other tubes got fried, too, for want of voltage regulation; but wouldn't that tell more noticably in the sound (which was not good, but it has been worse with good tubes and worse electricity)?

???

Paul S

Be careful with those 5751. ML2 became unstable with 6db more or less (I do not remember already) global feedback. Lamm told me a long and big story how critical for ML2 to have the global feedback of exact amount and how “properly” the feedback is implemented in this amp. At that time I was ignorant and was listening what Vladimir told me gullibly? Stressing the exactness amount of global feedback in his design Lamm however endorsed to use the 5751 in input stage.  Later on what I get some understanding of the process I was laughing on Lamm’s self-pomposity.

The ML2 has global feedback that covers all 3 states and 12AX7 in the first stage. If you put the 5751 in the first stage then the gain of the first stage gets reduced by ~25% that would make the gain of open loop reduced. The closed loop is still that same that would make the amps in a whole to change the amount of the global feedback that is being used. I did not detect any problem to use the 5751 but I found that it was very funny that Vladimir insisted that the feedback resistor mast not be changed because it will be “bad for feedback sound” but the change of the gain in the open loop is permitted - like it would not affect the close loop.

I am sure that for 3 easy payments of $699 Lamm would promise to reset the laws of physics in his amp….

Paul S wrote:

I don't know what you mean by "the demagnetizing coils in its output transformer".  What does this mean?

Reading at my site Paul’s narrative about his fight with ML2’s tube and the ML2’s tube sockets I need to confess that I always was suspecting that he is some kind of odd wacko with bizarre and sick approach to the subject of amplifier tubing. Too much of what he said just did not make sense to me. Well, until the last night…

Disclaimer: (wherever I say ML2 I imply the Lamm ML2.0 version not the ML2.1, about which I do not case)

I own two pairs of ML2 that I bought new and used them dally since then. I know dozens of ML2 heavy users who never had any problem with those amps. In fact ML2 were very stable and very secure in operation amps. The Lamm’s ML1, the preamps were with operational “issues” but M1.1 and ML2 were very stable.

Anyhow, I have this week a local friend of my brought to me a pair of ML2 and he need a help with inner-amp filter. The amps he brought were not used for a while and were sitting in storage for 5 years. I did the necessary things with filtration and desired to listen them. I have no way to drive Macondo full-range but I have my MiniMe prototype…

The plate currant on both of amps was too high. The further inspection showed that ML2 regulators stopped working. The B+ was not eaten by regulator tube and ML2 had 230V at .6A on plate – it is over 130W anode dissipation – bye-bye the 6C33C… Actually, I was rung the 6C33C like this for a whale, when I was searching the problem, the tube did not failed. The ML2 regulator is too simple and cannot work without a load. 

Anyhow, chasing the problem in the ML2 belly I was wondering why the regulator was down… on BOTH AMPS. The das tune was on, the error pentode was on the regulator tube was on. The design was very simple and very straight forward… I measured and confirmed that the right voltage was given to all tubes but it looks like the error amps did not corrected the voltage. I test the error amp’s pentode it was perfect. I went to my storage and brought my own WE 6AK5 – the same result. What the hell I was wondering…. I pulled my dummy loads to substitute the output tubes as it was glowing with 130 on plate like crazy and then the regulator … kicked in!!!

OK, guess what was the problem? The problem was the fucking frugality of Mr. Vladimir Lamm. This guy makes 30.000 worth amps with none-treated steel tube sockets. It was juts ridicules – the tube sockets on 6AK5 were rusted to the point of no contact with cathodes. The sockets were cleaned and the amps were working flawlessly… I wonder if Lamm would use not $.15 tube sockets but $1.15 then would it lead to extra $10.000 of the retail price… Anyhow, the idiotic economy on such a mall things in a flagman amp that cost $30.000 (the amp was made in end of the 90s with serial number in 40s) appears to me… to idiotic. It is not something that saves money to Vladimir and it is not something that Vladimir do because of his ignorance (his is a brilliant technologist). Those things that he does are because his attitude and he learned that his customers and his revisers are deaf idiots and he can go away with it. if not then how can in such case one explain that in early prod action of ML2 Lamm used Cube 950 caps for inner-stage coupling – the phenomenal-sounding Polypropylene foil caps. In the amp that I was fighting with regulator all coupling was made by bypassed Cube 935 that are Metallized Polypropylene and with very different, less sophisticated sound. It was $7.35 for Cube 950 vs. $5.11 for Cube 935 – I am sure it was well worth it… I was wondering how much Lamm “saved” in more important thing – on output transformer for instance. I wonder how far Lamm electronics sound off from the level as it MIGHT sound if the electronics was made with respect. I remember a few years ago Lars Fredell let me to listen his M1.1 that Lamm custom-made (just with better parts) for him. It was WAY more interning then the production M1.1. Well, I presume that if Lamm wish to kiss somebody else ass with sound then his is perfectly able to get better sound of his design. I wonder why it shell be a condition? Does Lamm’s series production is just the Lamm’s attempt to eagerly spit on somebody ass? Well, if you understand what I mean then you understand what motivated me to design Melquiades and do not be subordinate of that “subordination”.

Anyhow, the Paul’s suffering with tube sockets was understood now, at least…

Paul S wrote:

After a few minutes, I checked the amps with a (Fluke 189) DMM, just for drill. Voltage was set down to 175, but current was about 450 mA.  Dialing down the pots, I could not get  the current below 390 mA, even though the pots still turn freely.

Actually when you hit the end it will not turn freely. It will still tune but you will feel that it is wand and that something is flipping inside. With a new 6C33C it is normal to have excessive current what the tube is too fresh. Burn is a bit longer a d the initial rise of current will be stop. I would lower the plate voltage to 130-150V and let the tube to burn at lover currant, let say 300ma that would make 45W plate dissipation. After 10-12 hours your tube in this mode will not push more then let say 230mA. Then set the cruse plate voltage and play only with current. Again, it is normal behavior for this tube. One of the reasons why Lamm I presume charges $1.000.000 per tube that he sells is because he pre-burns his 6C33C – a slightly aged 6C33C has no effect of current rise during start up. BTW, I have written about it before.

What is interesting to note that observed this behavior only with 6C33C in ML2.  Milq with it’s fixed bias has no such an effect. I did not look I not ML2 and I do not know what kind out stage biasing Lamm uses. The plate is regulated for sure but with biasing it is not clear. Since that 6C33C behaves in my amp and in ML2 differently I might presume that ML2 does not use a straight fix biasing.