Rerurn to Romy the Cat's Site

Yesterday I got two emails asking me about Lamm ML3. I replied as I usually do, saying that I that I do not practice in satisfaction in audio-consumer curiosity.  Still, it made me to think again about a hypothetic amp around GM-70.

I am not a fan of GM-70 as in my view high volume should derive from a sensitivity of acoustic system not from power of amplification. Still, GM-70 is a very linear triode as some people might consider it. So, if to have an abstract task to get “better” GM-70-based amp then would I make/buy something similar to Lamm ML3?

Nope I would not.  My rational would be not just me suspicions that after the Lamm ML2.1 the ML3 will be loaded within a bunch of the “Easter Eggs”. My rational would be purely architectural as I personally would like to see in an “ultimate” GM-70 implementation (if I personally have inters in that tube) something different. So, what my version of a “stepped-up” ML3 would be?

1) Two stages only. It looks like ML3 has one superfluous capacitor-coupled stage – unnecessary.  For better architecture search better tubes.

2) Grounded cathodes. There is nothing further should be said.

3) The Input chokes filtration with SS rectification. The contemporary Silicon Carbide Schottky diodes can run up to 1200V, have no reverse or forward recovery and they can switching incredibly fast. Behind choke filtration is way more preferable powers source for any constant current drowning amplifier.

4) Class A1 operation only. What Lamm was doing in his ML3 making GM70 to run with grid currents? Why is it necessary? Because we want to drive GM70 deep into A2 and get more power from the amp on order to punch the dull speakers? It would not be my objective and I would like my GM70 to stay only in class A1 operation.

5) Voltage. It is obvious why ML3 has that powerful current buffer built up in second stage – because Vladimir wanted to make it to stay at low voltage with his ML3. I would estimate ML3’s GM70 has 850V on plate and here is where Lamm’s needs to dive into A2 in order to get 32W. At 1000V GM70 does 40W in A2 and at 1300V it can do 32W in A1. Many, very very many people who used GM70 report that at 1200-1400V the GM70 begin to sound much better. I do not know why they report it – it is possible that they have low-sensitively acoustics and they need more power in pure A1. I would not look for a lot of power but if to go for a powerful triode and demand a lot of power then I would prefer A1 power instead of A2 power.

6) No feedback. I really do not find it is necessary to use DHT and run feedback, particularly in the way how ML3 implemented (circuit board assembly). It looks like ML3 runs 2 feedbacks and none of them are truly necessary. The global adaptable feedback with which Lamm let people to play with damping to make the amp to work with various LF sections of loudspeakers. I do not feel that it should be there. It would be OK solution for $3K amps what a person would like to put one ass on all stools but it is not a solution not for an “objectionable” amp. If you want to exact-load your bass section with your SET then go DSET. There is no needs to change the operation of the enter amp, applying a global feedback, juts because the port of your loudspeaker does something funny. The second feedback ML3 runs most likely from second stage to input. Why was that necessary? Because Vladimir was too lazy to find an input tube with higher bias voltage and decided to jack up the 12AX7 cathode with rise of signal I see no needs for it and call it patching. Search better input tubes an do not cure them with feedback.

7) No DC filaments. There are so many solutions nowadays how do not drive DHT with DC on heaters that using DC in an “objectionable” amp became kind of cheap.

Rerurn to Romy the Cat's Site