6AS7 on the left, and 6336B on the right.
Notice the size of the graphite plates in the 6336B
Semi trailer in background for scale!
I discovered that Mark's 6336A mod proved to be even more extensive than described but worth it. My timing was perfect because Mark posted the Ultimate M-60 Tweak just before I started working on my M-60 kit. I don't think I would have had the ambition to rework 16 sockets to ceramic. I was really excited about the mod because after living with a borrowed pair of M-60s from Bill Layer for 6 months, I was hungry for a little more drive strength (lower output impedance) for my 4.7 ohm impedance speakers. So I set off to build my kit according to Mark's recommendations preparing for the 6336A tubes.
Finding 6336A tubes actually became a bit of an obsession for me. I think I lost 4 weeks of kit building time by spending my evenings on the computer treasure hunting. I ended up gathering mostly 6336Bs and some 6336A and a few metal plate 6336s. My favorite pull came from a marine surplus business on the East coast. A guy named Tony had a bunch of 6336A & B tubes that were installed into military power supplies for aircraft carriers that were put through their test procedures, plastic wrap sealed and set on a shelf for the next scheduled power supply maintenance rotation..........which never came. Tony cut the 30 year old plastic open, pulled the tubes and shipped what I could afford to me. All 16 that I am running now came from there.
I found out early on that the 6336A/B tube is very prone to shorting. I went through a lot of fuses the first two evenings of getting some of my first pulls up and going. I was only getting 1 out of 3 tubes to come up clean. I later discovered that these tubes like to wake up first by cycling the heater on and off every half an hour throughout the evening and then left on over night before applying plate voltage for the first time (in 30 years). Using this technique I got 90% of them going.
I didn't want a fuse in series with the music signal going to my speakers, so instead I'm running two fuses, one each between the 1 ohm plate resistors and the positive of the power supply cap for each half of the circlotron. It ends up that this offers more protection as well. A fuse in series with the load protects the load, but then nothing protects excessive current in the outside circle of the circlotron. When a tube shorts grid to cathode and turns on hard, all the tubes on the other half of the circlotron now get almost all the voltage of both output power supplies. And the shock of this may cause one of those tubes to short grid to cathode turning it on hard. About half the time when a fuse blew, the fuse in the other phase went also. I'm glad that I am past the fuse blowing phase of this project.
The largest caveat I ran into had me stumped for a long time. I did everything Mark described and I was getting a nice low output impedance that was taking good care of my speakers. What I wasn't getting was output power, it only jumped from 60 to 70 watts into 8 ohms. So what is different between my amps and Mark's? Mark said that he has all his plate currents matched within 10%, mine weren't. So I called up Tony and bought everything he had. But even with a matched set I was still getting only 70 watts???
According to Bill Layer, early M-60s did use the same output supply transformer as the MA-1s as Mark describes, but current M-60s now have their own smaller transformers. And the 6336A/B tubes simply suck the B+ voltage down. The stock transformer can only supply enough power to develop 70 watts.
I found a great replacement transformer from Toroid Corp. of Maryland at:
It is rated for 470VA and its part # 749.1202 for only $84.60 each. They threw in two mounting washers and two rubber pads for each transformer which was necessary to fit them into the M-60 chassis. It is a dual primary / dual secondary isolation transformer that holds a strong 150VDC when rectified and loaded. The interesting thing I discovered is that the transformer is not a 1:1 winding ratio. It is 1:1.05, a 5% step up. I suppose this is to compensate for losses. So if 150VDC is to high, one could run the transformer backwards to slightly step down the voltage. The stock M-60 output supply transformer weighs 5.1 lbs and got to hot to touch. The new transformer is 8.1 lbs, only gets a little warm, and just fits into the chassis with only 1/16" to spare. The 6336A/B tubes seem to really like the extra voltage. The amps sound became bigger, stronger, and more exciting/dynamic, and I was now getting 145 watts @ 8 ohms.
I am a little concerned about the output filament transformer because it is now also smaller than what is used in the MA-1's. Although I believe it has the same rated output current, just at a higher temp rise. They run HOT yet the voltage does not droop, it holds 6.4 VAC even with 8 6336B tubes loaded. I plan on leaving them.
I have one finding that is different from Mark's. Using a 0.01 ohm resistor in series with the filament of one of the sockets, I found that both 6336As and 6336Bs pulled 5 amps, and only the metal plate 6336s used 3.75 amps.
I now consider this tweak finished and successful with even better results than Mark described. I don't consider a 6336A/B being equal to two 6AS7s, it is equal to 2.4 6AS7s. The output impedance dropped buy 2.4 as if I was running 19 6AS7s per chassis. I actually think that when fully biased up at 1.2 amps the amplifiers over dampen my speakers. Backing off the bias to 800-900 mA raises the output impedance a little and seems to critically dampen my speakers. I suppose the bias could actually be used for a tone control, less bias for thicker bass, more bias current if the bass gets a little to sloppy. Mark led me to this excellent article on critical damping: Visit:
I would love to hear from anybody else that has tried the "Ultimate M-60 Tweak". In the next "Speltz Discovery" I will detail out my bias/balance switch arrangement that enables adjustment in 10 seconds on the fly, and detail out my output B+ power supply.
Enjoy the music.