Feb 26, 2016

MC1/60, E60M, TM100 SE amplifier - Part 2

Back on the air with amplifier circuit



Voltage requirements.

We need 600V for the final stage, 300V for driver, negative bias voltage, heater voltage.

High voltage is not very high, I can take advantage of vacuum tube rectifier (helps for soft start) and polypropylene capacitor for main supply. I prefer the sound of these caps over the usual electrolytics.

For driver stage I choose choke input filter with bleeder resistor. Choke input is less disturbing for the power transformer and ripple is very smooth.

Main power transformer windings.

1- 500_0_500 V will give me the proper 600V high voltage with GZ34 rectifier.
2- 350_0_350 V will feed the SRPP with 30H choke input filter and ultra fast soft recovery diodes.
3- 50V for bias.
4- 7.5V for MC1/60 to provide 4V choke input filtered too.
5- 6.3V for 6SN7.
6- 5V for GZ34.



On the bench



Sub chassis ease construction a lot.........




Amp finished





Tamura TN351 interstage transformer with it's 1:2 voltage ratio helps a lot to drive the final tube easily with minimum distortion from the driver.
The little black switch next to the 6SN7 socket is for a local NFB loop. I can choose between 0 / 3.5dB feedback. Personnaly I prefer no feedback.
Screw speaker terminal, grey paint and AEI silver plated socket give a retro fashion look to the amplifier.



The second black switch next to AC connector give me the choice to earth or not the amplifier, helps to get the lowest noise floor.



Screwdriver potentiometer and large panel meter for easy current setting.





The mighty E60M tube @ 600V / 75mA




Listening test

Amplifier is dead silent on my Klipsch, I have to put my ear against tweeter to hear a slight hisss.

First impressions .... punchy, very clear and detailed without being tiring. Beautiful and delicate mids and highs (as fine as the 10/VT25 amp !! I can make comparison on mono recordings VT25 amp on one side, MC1/60 on the other), voices are natural, very articulated and living like (Madeleine Peyroux_Careless Love).
Bass are impressive, this is the first time I can physically feel bass with my Klipsch (Bach_ Cello suites, Pierre Fournier). Excellent soundstage with lot of depth, very airy and detailed, I ear the very, very slight shift of the bottle neck on the guitar, the slight fingers taps on strings ( Pedro Soler_Sombras or Selmer 607 Gypsy guitar CD) never heard before! It's good, very good in fact even on very complex music ( Verdi_Messa da Requiem_Giulini).
I also listened to E60M and I have great difficulty in distinguishing the tubes. More body may be? It depends a lot upon input tube. Best with MC1 / 60 are the VT231 Ken Rad (sooooo punchy in the low end), Hytron VT231 or Brimar 6SN7 GTY with E60M or TM100.

Amazing to see after so many years that the 4 V tubes have always been killers compared to the other ones (2.5, 6.3, 7.5V …..). I have been playing with plenty of these classics (2A3, 45, 50,811,211, 845 and so on) and found each time they were inferior to the 4 V ones *. Is it a matter of cathode coating, is 4 V the best voltage for regular emission ? I have no answer, just listening facts that aficionados of the PX4, PX25 and AD1 perfectly know.

* Just one exception, the Visseaux A710, a 10 with BIG plate (as big as a 50) using a 7.5 V oxide coated filament. It's to my knowledge the only 10 of that kind and the only one that sounds like a PX4.

Have to wait for some friends to come and make a very serious listening.
At that time this amp is the very best I have ever built.

Some CD's for this test









Feb 19, 2016

MC1/60, E60M, TM100 SE amplifier - Part 1

In the DIY world of triodes amplifier lovers, most people know the best of the large power tubes like the 211/VT4C or the 845. These tubes are easy to source, still manufactured by Chinese, Czech or German. It's even possible to find some NOS ones from GE, RCA…
There are plenty of circuits to try for the hobbyist and parts like output transformers are easy to find.
This topic will make you discover some of the best European triodes and especially french ones. For audio use everyone, almost everyone, knows the tubes from Mullard, Philips or Mazda, but very few have had the opportunity to get on Neotron, Visseaux or SFR.

These manufacturers made tubes mostly for administrations and are rarely present on the tubes vendors stalls still active today. In fact, most of these tubes were bought by the Japanese in the 80's and are now selling at indecent prices. Understandable when you see manufacturing quality that may turn you pale. Philips MC1/60 or SFR E60M where built to a very very high standard even for the tube industry of that time.

Let me introduce our competitors.

Philips MC1/60






This Dutch power triode was specifically developed for audio, rated at 75W plate dissipation,
very close in specification to the 211/VT4C, but easier to use due to its lower trans-conductance.
The MC1/60 have a second advantage with its 4 V oxide coated filament rated at 3,3A (to compare with the 10 V/ 3,25A of the 211). We will talk latter on an other advantage of this oxide coated cathode in term of music sweetness. Sweetness and naturalness shared with the very best of the 4 V triodes like the PX4 or the AD1.






SFR E60M

French military triode made by Société Française Radioélectrique in the 40's. Incredibly well made with cost no object materials, twice the weight of its Philips relative, these tubes can bear the 3X75B military code. Very close in specification to the MC1/60 which is very helpful due to the lack of information concerning the SFR company products.
Like the Philips it's a 4 V / 3,3 A oxide coated cathode, anode power rating 75W.






MAZDA TM100

French military triode made by Mazda in the 30'/40',
also can bear the 3X75B military code.
Very different construction with glass rods to support filament springs.
Despite the old fashion and less impressive construction it is probably the best sounding triode of this group, and probably one of the best sounding ever.
Oxide coating cathode, 4 V / 3,3 A, power dissipation 75W (better not to go beyond 60W).
All these triodes can be switched as they have the same filament ratings, use the same socket and are very close in specifications.


Top view shows the very different construction.




The 6 watt monster amp

My first intention when I decided to build an amp around these tubes was to get a reasonable output power with the best possible sound. You will say this is a very personal and very subjective goal but during the past 20 years I built a lot of amp, preamp but only kept 3 ones after intensive listening test with golden ears friends (10 / VT25 SE and VT25 / VT25A AB1 PP amps, 6J5 / NP206 Tango transformer coupled line preamplifier). These materials serve as references to my future constructions.

My requirements are small in term of power. I mainly listen to Klipsch La Scala speakers upgraded with JBL 2470 drivers and Alk filters and at 105 dB I can shake hell with just 1 watt.
So I decided not to push my tubes too much and make them sing at 5/6 W output.
Looking at the curves I found a satisfying operating point at 600V / 80 mA with a 3.5K load,it may be weird at first sight but at that point internal resistance is mainly 1.6K (S=8mA/V and μ=12.5). This gives the maximum output with a 3.5K load (and I had the transformers on the shelf for a long time, it helped a bit to make a decision).



The driver was a more difficult choice. I needed very low distortion, large RMS output voltage, low output impedance, large bandwidth. Not easy to combine. After several tests with the most commonly used drivers and tubes I stuck to a 6SN7 SRPP feeding a transformer. In that way I had a well defined load for the SRPP, a very good bandwidth despite a coupling cap (no current through the transformer) and the possibility of less final distortion using a 10K/40K step-up transformer that double the driver voltage. Moreover I could use a negative voltage bias and the possibility to set the current depending upon tubes to be used.

The 6SN7 choice was not a choice. This tube is well known for its linearity, its sound qualities and I have plenty on hand.
Long time ago I have modelized the SRPP based on a very interesting thread by Merlin Blencowe from AudioXpress ( The optimized SRPP ) and I just have to fill some fields in a Excel file to have my operating points, resistor values, gain and output impedance.

Now some math considerations to see the different AC voltages involved in this project.

Power output should be 6W/8 ohm or 6.9 Vrms.
Output transformer have an impedance ratio of 2.28 10-3, a voltage ratio 4.78 10-2 (-26,4dB).
It means for the power tube a plate voltage of 144 Vrms or 408 V peak to peak to get my 6.9Vrms.
Looking at the MC1/60 curves this can be done with 48 V peak to peak on the grid (17Vrms).
Gain of power triode is 8.5 (+18.6dB), confirmed by calculation where

                                                  G = µ x Rload/Rload + ρ.



Now to get 17 Vrms for MC1/60 complete anode swing I just need 8.5 Vrms (24 V peak to peak) out of my 6SN7 SRPP due to a 10K/40K interstage transformer used in that circuit. Easily achieved, under 300V anode supply the SRPP circuit give a gain of 13.4 (+22.5dB),means 0,63Vrms at input (1.80 V peak to peak) and very low distortion.





The amplifier overall gain is :
+22.5dB (6SN7) + 6dB (Transformer) + 18.6dB (MC1/60) – 26.4dB (Transformer) = +20.7dB


To be continued......
full amplifier circuit, construction details, listening test, tubes to be used, mods and tweaks.