jessie.dazzle wrote:

I should mention that with all the wonderful things it does, the L2 (preamp) definitely deserves a fresh pair of 12AX3s and a 6C19P. (I have not changed them since starting work on the horns... all tubes are exactly one year old, which with my listening habits is more like three years).

Jessie, the retubing of L2 is necessary but it will not affect the noise that preamp output. The MF nose that is coming from it is the noise of MOSFER output stage and it is intrinsic with the topology of the given preamp. Well, I do not know if is it really correct but it was what Lamm was truing to convince me when I complained about the L1/L2 noise.  Still with 109dB sensitively and ~25dB gains in ML2 you might mitigate the noise and make it practically auditable. You have too much gain in your playback for 109dB sensitively. So you burn ~8-12dB in the L2 attenuators, use it at the very first clicks. However, it is the L2’s out stage the one the produces the noise and the noise is constant with change of volume (I called it “bias noise” but it was juts my name). So, what you can do it to let the L2’s output stage to run full throttle (the L2 has unity gain when attenuator points at 1PM) but kill the gain at ML2, or AFTER the preamp. So, this way you will have your ~8-12dB voltage divider at ML2’s input that will kill voltage ALONG with the noise of output stage. It means your noise will be ~8-12dB less but you will add volume of music by more opening up the L2 attenuator (not to mention the L2 sound better when it more opened).  I used ML2 with 109dB sensitively and L2 with 12.5ddB voltage divider at ML2 and I had no annoying noise at nearfiled.

Also, a very few words slightly off the subject. If you did not replace the 12AX3s and particularly the 6C19P for a year then your preamp is absolutely dead sonically. A few years ago Lamm in response to my publics criticism of extremely low lifespan of L1/L2 power supply made a note on his site recommending to change the 12AX3 and 6C19P once a year (it is after 10 year of L1 production). I still disagree with what Lamm said. The 12AX3 in L2 must be changed after 6 months and 6C19P after a month or two maximum (I would say 200-400 Hours, no more). I have personally witnessed L2 preamp where the 6C19P was dieing after 20 hours, I am not kidding. Sure, it is still regulated voltage and tested any my tester with perfect transconductance and cathode emission but it begun to sound with compressed white-noise injected of any tone… truly disgusting... So, buy a box of the 6C19P (that’s God in Russia they cost pennies) and replaced them as I did it – each time when I paid my apartment rent….

jessie.dazzle wrote:

Romy, you may be right in suggesting that I am "going a little crazy in there" (I consider this a compliment btw)... so for now I decided to try some locally sourced precision metal film resistors (Vishays are not that easy to get over here in small quantities, even though I seem to remember the founder of Vishay being a Frenchman!).

Romy,

Thanks for the info on alternative manufacturers.

The resistors I found locally seem to be working super well (precision metal film... didn't ask who made them...they are all black and cost EUR 2.80 each). I may eventually replace these with RN60s just for form, and also go for a bit more attenuation, as I am discovering that the 1:00 position (unity gain) is really still a bit too loud, and that the L2/ML2s combination does sound better with the volume pots further open. I made the fixed attenuators so that changing resistor values is a very simple operation.

Also, I have been meaning to mention that normally I would not choose to discuss an issue which is really related to the L2, in this ML2 thread, but I originally thought the hissing was a product of the ML2s.

So to bring the discussion back on course, these little attenuators have done a lot to get me closer to the ML2s.

jd*

Paul S wrote:

Today I used the re-capped/improved K&K phonostage as a source and I drove the speakers first from 8 Ohm and then from 16 Ohm taps, and I got a very pleasant surprize...... .It certainly gave a clue as to why Romy would opt away from feedback in the Milq's.

is has absolutely nothing do with feedback. The ML2 uses tabs and moving from 4R to 8R, or from 8R to 16R you are juts more unloading the output tube. The more tube loaded the less power and fewer hammocks you get out of it. Do not feel that idling tube irrationally would necessary bring all kind of benefits as if you go too far for the capacity of the driver driver/crossover then the Sound become too much “extreme” and in many instances almost like digitalized.

Surely in a compromised context of a full range SET (not DSET) it is seldomly possible to find a proper loading for each driver…. Anyhow, when I used my ML2s I always used 16R tabs with whatever speakers I had at that time. In fact I remember I disconnected the 4R tabs and had 2 pairs or 16R binding posts…

I was beginning to wonder if I'd ever get the new listening room checked out electrically and the new system ground installed, etc., but I finished this work yesterday, including a soldered-rather-than-clamped ground wire-to-rod connection.  While I was at it, I also changed my tonearm cable ends from female to male jacks so I could plug the arm wire right into my tranny instead of using jumpers; but that's another story.

The new room is smaller than the old one; in fact it's the smallest room I've used with the DEBZs, and I wondered if this might help the ML2s, if not overall sound quality.  I measure the new room at 20' 6" X 17' 2" X 8' = 2,829 cu ft.

The system was down for about 7 weeks for/from the move.  I had finally figured out just before I shut it down for the move (after futzing with my tonearm, looking for bass,  AGAIN!) that the 6N6Ps were going sour (again), after only 3 1/2 months, so I put in some new ones (again).

It took ~3 1/2 hours for the system to charge up and start working as a system again, and then I tried Romy's suggestion to try dropping current (and raising vlotage) on the ML2s.  I experimented again and stuck with the 16 ohm taps, and as I listened I moved the V/A points around, eventually deciding on 54 Watts as a pivot, and I settled on 29.5 mA and 183V.  I love pretty much everything with respect to the music at these settings.  Bass is not only harmonically richer and more articulate, it is deeper and amazingly strong with my non-greedy 15" BRs.  Oddly, overall timing improved quite noticibly, too; go figure.  Basically, there is no down side for me with this change, except you might say that the "image density" is somewhat lessened by it; but this is more than made up for -  in my system/IMO  -  by the much greater tonal variation at the "new" settings.  YMMV, of course, but this does mean the the V/A parameter can be used as a sort of tuning device.  I have to say, it changes things more profoundly than just tube rolling, although it is not like two different amplifiers.

I was afraid the DEBZs might overload the new room, but this was not the case  -  at all.  By the time I concluded today's session I was wearing the DEBZs like headphones, sitting darn near up between the speakers, which are presently about 6' apart and about 7' off the back wall, in this room.  I have not done much with speaker/room loading yet, but I quickly found the best bass, which spots will serve as my reference points.  If it goes like it did at the last place I will leave the speakers where the sound is most "interesting", then re-tune for bass and/or take a hit on that score, if necessary for the greatest tonal/dynamic range.  Last time the two spots -  "interesting" and best bass  -  were not far apart.  I am getting a nice sense of "immediacy" right now, which I like, and yet I can listen to crappy records again, too.  Perhaps the system has sounded more "clear" before, but it is as musically "articulate" as it has ever been, despite the apparent paradox.

The other day I saw an online photo of Wilson Wamm speakers sitting next to what may have been a pair of ML2s.  I sure hope they were 1.1s, or something.  Sure, I am enjoying myself mightily with the DEBZs; but it is already obvious that although the ML2s do better with tougher loads than doubters might suppose, they also continue to up the ante if you go to them, electrically.  I'm not looking for horn parts yet, but I am thinking about it more and more...

BTW, about 90 degrees out today, and all is well, or at least OK.

Best regards,
Paul S

Paul S wrote:

It took ~3 1/2 hours for the system to charge up and start working as a system again, and then I tried Romy's suggestion to try dropping current (and raising vlotage) on the ML2s.  I experimented again and stuck with the 16 ohm taps, and as I listened I moved the V/A points around, eventually deciding on 54 Watts as a pivot, and I settled on 29.5 mA and 183V.

Paul I do not think I suggested dropping current and raising voltage as I have no idea what loudspeakers you are using. I might suggested to play with taps and play with different current and voltage settings  but I would defiantly would not propose to go toward explicit dropping current and raising voltage. BTW, the 295mA (I ma sure it was not 29.5mA) and 183V is very close to what Vladimir suggests to use (175V and 300-310mA).  BTW, a few years about I was asking Lamm what was max current his ML2 transformer is gaped (to know what would be my current boundaries when I was dropping voltage) and he replied that it was his proprietary information that he is not wiling to divulge. What a hoot!

BTW, if you still looking for bass then you should dump you ML2 and but Lamm ML3 as according to some Morons ™ the ML3 has “better bass”….

Romy, at some point you did mention lowering current and shortly thereafter, if not in the same post, you added something to the effect of raising voltage.  You even mentioned values you have tried, and I printed it out.  But since I am usually pretty cautious, I took care to hold VA at or below stock, as you observed.  I did not have any preconceptions but went with what worked best for tone, and the bass was a bonus.

As for the BR bass, I think of basketball there, too, as though you slam the ball down at the floor, but the floor moves away slightly as the ball starts to touch it.  No, my bass is not the best, but the ML2 does a remarkable job with it, considering, and I have been futzing with foam inserts to +/- plug the vents, for reasons other than extension.

I would be interested to know if you can get good results with SET on your sealed units?  I have thought for a long time that either a "mostly sealed" unit or a non-greedy sealed would be best, practically speaking, but I never played much with the practical limits of sealed, at keeping the drivers "off the air cushion".  At this stage of my journey I am just glad that the ML2 can drive the BR about as well as can be done, given the built-in limits of both designs.

Let the suckers jump on the ML3.  I'm holding out for the ML3.1!

BTW, I have a "funny story" about losing track of a decimal on my Fluke DMM when testing (and then installing) some plate resistors...

Best regards,
Paul S

Paul S wrote:

Some time ago Romy threw me a fish in the form of two numbers, 280 and 185, meaning mV and V, respectivly, and I finally got around to chowing down.  I was getting annoyed with the sound this morning to the extent that I busted out the DMM, and I found the bias had drifted up considerably.  As I dialed it back I decided to just keep going, and I am glad I did. In my case, lower plate current and/or less power worked a treat.  It cleaned things up considerably and harmonics became better as well, including bass, with no loss in "articulation", placed in quotes because IMO bass "articulation" is not really present without proper harmonics.  In this case, bass was no more harmonically enriched than the rest of the spectrum, with the effect that sections were audible and interesting contributors in complex music.  Another beneift of the change was less HF hash, and in no way did the DEBZs lose power or "drive".  In fact, the net effect of less hash was more music and "drive" at any given volume setting. As I understand it, lowering plate current raises output impedance, so this is another move that "makes no sense", especially since the DEBZs are nominally 8R.  Still, the Lowthers certainly benefited from the change; but so did the big Audax.  The ribbons were the same, except they were dealing with less noise, so in that sense they benefited, too. Now (as usual...) I wonder how far I could go with this, at what point(s) the benefits would reverse.

First of all what do you do not controlling the plate current? Lamm drivers his out stage quite hard and it does require monitoring, or you easily jump to the too high currents with all know negative consequences.

Secondary, I do not believe that dropping the plate current to 250-280mA at the same (or higher) voltage might “cleaned things up considerably”. With lover current you should have in some instances deeper bass (because less power). All other things should not be changed too much. You might be very careful driving the ML2 at sub 35W or with one plate as the amp has global feedback and I detect that with dropping more then 3dB in feedback the amps become unstable, in fact you have no sensitivity in your speakers to go for lover plate power.

You might run across the ML2 taps, finding better harmonics but it is pretty much all the you can do with that amp. If you are not happy with harmonics of your channels with ML2.0 then it is absolutely undeniable evidence to me that you need to replace your speakers and do not try to twist ML2. The ML2 has stunning harmonic integrity, like no other amps I even seen, and they maintain the harmonics across VERY wide range of load. If you report that harmonic balance changes with dropping plate current from 300mA to 280mA then something is VERY wrong ether with your playback or with your perception. It is theoretically possible if you use ML2 for LF section and your bass damping is in “critical mode” but I very doubt that it is what it going on in your case.

Anyhow, you might review your findings… perhaps with other electricity. Do not mystify the output tube operation point. It is VERY simple to find the right operation parameters for your given load. You need to set your output 6C33C in the regime what at full power to your load the current and the voltage are clipping at the same time. Everything else is fantasies. I made countless experiments with that tube and find that with perfect semantic clipping the tube sound absolutely the best, not only this one but any other tube the I have tried.