emails from Ukraine, was Radiotron SE2A3

Bach Chaconne
from
Partita No. 2 in D minor for Solo Violin

An important lesson I learned as a classically trained violinist was that music theory evolved from the analysis and study of great masterpieces from Bach to Bartok. Great composers broke theory to reach new heights in musical expression. Just like any artistic pursuit, music continues to evolve through individuals who are not afraid to go against convention for a greater good!

Tube load line simulator
from Trioda.com

On the other hand, audio engineering, for the most part, has always been theory driven. Audio design simulators are significant modern conveniences but don't guarantee musical-sounding circuits even if the metrics look great. Creativity in audio design is when human ears are still the final arbiter, at least in my experience and opinion, which is why my hifi system is a hodgepodge of mostly obsolete technology, some of which date back to almost a century ago.

Last year, a Juilliard-trained pianist turned lawyer visited my man cave and listened to various amps + a couple of speakers I built. On his second visit, he brought his amps for comparison. At the end of that session, he commissioned a pair of Altec 753C speakers, the second pair I built in collaboration with John Piro of Hommage Audio, who crafted the speaker cabinets.

The chosen amp was the Radiotron SE2A3 - a topology from the pre-WWII Radio Designer's Handbook, 3rd edition. This circuit was from an era in which calculators were but a dream and engineers computed complex equations aided by a slide rule. 

!!!WARNING!!!

The voltages found in the following circuit can be lethal! Build at your own risk!

Although this pentode-driven circuit sounded fine as is, to my ears, its fullest potential was only realized after I adjusted the plate and screen grid voltages to achieve a 2:1 ratio as suggested by my Ukrainian internet buddy, Sergey Didkovsky in the winter of 2007. 

"Try please, to increase the cathode resistor (1,2k-1,5k ohm) and reduce anode resistor (to75k-82k). It will allow to increase Ua up to 150V and to receive Ua:Ug2 = 2:1 ratio. The sound should become more 'interesting and old.' That is especially good for jazz. Still mine 5 cents...use good NOS oil capacitors in G2. I use TCC Visconol. I think oil caps in the grid - adds bass. Reduction of the anode resistor - improves the hi range a little."

Click here for the 2007 Radiotron SE2A3 article 

I'm pretty sure the 2:1 ratio wasn't a number crunched by a simulator, it was a figure painstakingly chosen by Sergey's ears because of his love for music!

 

Sergey, wherever you are, I pray that you and your family are safe. Freedom-loving people around the world are very grateful, yet very sorry that Ukraine is all alone in shielding the free world from the atrocities of a megalomaniacal autocrat!

Tube Rolling Fun (and Peril) + Adapters

Tube rolling is a fun aspect of the audio hobby. It only gets tedious when audiophiles start splitting hairs about a VT231, 5692, etc. I draw the line when an audiophool claims, "...my line stage preamp resolves the R train rumble under Carnegie Hall better when I use a 6AS7 in place of the original 6SN7..." 🙄  

Those two tubes may be pin compatible BUT a power transformer 🔥melt-down🔥is imminent since the 6AS7/6080 draws 2.5 amps @ 6.3V. That's over 4 times more filament current drawn by a 6SN7, which only draws 0.6 amps @ 6.3V.

 💣 💣 💣 Rolling tubes is dangerous unless you have a working knowledge of electronics to comprehend tube data/specs. 💣  💣  💣 

It's been a while since I listened through Telefunken 12AX7 smooth plates on an RCA phono. Since I used to graft this circuit on a Dyna PAS preamp, I ordered a pair of 12AU7/12AX7 > 6SN7/6SL7 adapters to relive the experience on my RCA octal preamp.  

Back in the 90s, Steve @ Angela Instruments had a boatload of Raytheon 5755s and shared a sleeve with me. Since I never found a project requiring these tubes, I got a pair of 5755/WE420 > 12AX7 adapters to listen to them in the Shure M65 clone. Piggy-backed to 12AX7>6SL7 adapters, they also worked nicely with the octal phono preamp. 

Despite slight differences in transconductance and plate impedance between the 6SL7/7F7 and their younger 9-pin siblings (ECC83, 12AD7, 5751 and 5755), there was no visible variation in square waves nor frequency response aberration when I bench tested those tubes in both circuits.

Admittedly, loctals don't look as glamorous as their octal cousins with bakelite bases. But I'm very fond of them to the extent that I built and use a full function loctal preamp. I've been adopting unloved loctals at radio shows and hamfests. Essentially the industrial looking 7F7 and 7N7 are rebased 6SL7 and 6SN7, respectively! My only issue with loctals is, the tube pins are prone to corrosion. This is easily remedied by a wire brush + DeoxIT before use.

I purchased these tube adapters from the seller xulingmrs. They are very well made and fairly priced. Shipping cost and transit time from China was also reasonable. Highly recommended!

Darling Amp, JE Labs Style

It's been 20 years since jc morrison started hosting the new york noise audio design exhibition at his Hoboken, NJ loft. One of the highlights of that year's exhibit was Darlingfest '99 as reported by Jeremy Epstein in the final issue of Sound Practices. The Darling amp was the brainchild of Bob Danielak and was published in Sound Practices Issue 15.

Jeremy's octal driven Darling amp made a lasting impression on me but I never got a chance to build my own version due to a busy performing schedule.

While I was organizing my workshop in the basement this past winter, I found power and output transformers from a Sony TC500A tape recorder + other parts to build a Darling amp.  

The chassis cleaned up nicely with a wire brush. Then I drilled, cut and punched the necessary holes before spraying a fresh coat of silver Hammerite.

I chose the 6F5 hi-mu triode with a cool grid cap as the input/driver tube.

To qualify as a Darling amp, the indirectly heated 1626 transmitting triode  output tube is de rigueur.

The TC500A donated a healthy 6CA4 rectifier.

!!!WARNING!!! 

The voltages found in this circuit can be lethal, build at your own risk!!!

My interpretation of a Darling amp is ready for testing and listening!

Bench Testing

Both channels driven

P = V²/L

2.415 x 2.415 = 5.832/8 = .729 or 729mW per channel

Hum and Noise = 1.8mV

100Hz

1kHz

10kHz

I did not expect this level of performance from the cheap looking Sony OPTs, which I also verified while listening to the amp!

Bass extension is surprising from this 3/4 watter! It is deeper and tighter compared to my Tango'ed SE171A amp. However, in the midrange, the true Directly Heated Triode shines. The 1626 was just a touch veiled in comparison. Although it sounds more powerful than the similarly rated SE171A, it also doesn't overload as gracefully. I remember hearing a similar phenomenon when Steve @ Angela Instruments asked me to build the Simple 5691 > EL34 SE amplifier. The Simple 45/2A3 clips with more finesse. Is the cathode causing the relative lack of transparency and harder clipping?

The days of $5 Type 45s and 2A3s are long gone. But NOS 1626 tubes can still be acquired for under $10 a piece. Since it has a cathode, with careful wiring less than 2mV noise is achievable on AC heating which is quiet enough for me even on headphones. There's no need to hunt for hard to find 25-75 ohm, 2W hum nulling pots. That's just a couple of reasons as to why a Darling amp derivative is one of the cheapest and easiest routes to a DIY SET amp nowadays.

Did I mention that it also does a great job driving headphones?

headphone adapter/attenuator schematic for 
efficient headphones ex: Grado SR80, Koss KPH30i/Porta Pro
Lower efficiency/high impedance classic headphones like the 600 ohm AKG K240 Monitor, 
Sennheiser HD420, HD6xx and Orthodynamic/Planar magnetic: Echo TDS-16, 
Fostex T10 and Yamaha YH-1, can be driven directly from the speaker terminals

Happy listening!

Morikawa SE2A3 in Jose's Listening Room

Back in the 90s I used to hang out with Ding and Jose at Hamfests in the suburbs of Philly. Jose isn't your typical audio nerd who subscribes to the latest is best mentality. He goes for value for money and although not a DIYer, has an eye and ear for quality design! Even if he has a Technics SP10 + Sumiko MMT turntable combo, most of his listening nowadays is through a high quality pre-bitstream a Denon DCD 1500 and a stable platter Pioneer PD65.

Having heard that a well designed SE2A3 amp is a synergistic match for his Altec 604-8Gs, he spotted and won this Chuyu Morikawa designed Audio Professor SE2A3 amp on eBay. However, the amp had two issues. The input level controls were no longer functioning, and worse, the amp was unstable manifested by extra-musical popping and whistling sounds. Since we're practically neighbors now, I took it home for a check-up.

Under the hood 

Beautiful workmanship, typical of a handcrafted amplifier from Japan

After installing a dual ganged 100k Alps audio taper input pot, I realized that the cause of instability was the step down transformer the eBay seller sold with the amp. It was hardly stepping AC down to 100V. During the amp's residency here at JE Labs it was plugged to a General Radio Variac set to 100VAC. It was stable and quiet. Fortunately Jose also has a Variac on stand by, which he acquired during our Hamfest hunting days.

6Z-DH3A

The amp uses 4 x (2 for each channel) 6Z-DH3A, UX-6 based, high-mu triodes as input/driver tubes. This tube is electrically equivalent to an octal based 6SQ7. AFAIK, there are no plug-in compatible US or European equivalents. 

The input/driver circuitry is configured in SRPP, very similar to what I did in the JE Labs SRPP 45/2A3, albeit with textbook operating points.

5 Volts rms to an 8 ohm load at onset of clipping = 3.125W per channel

Luxman SS5B-2.5k OPT

100Hz

1kHz

10kHz

These square waves show that the Luxman SS5B 2.5K primary SE OPTs are of very high quality. Not much information is available on the internet regarding this transformer aside from completed auctions, but I've encountered this model in schematics and pictures in Asano's two volume tome on tube amplifiers.

Since Luxman SE OPTs were no longer available during the Sound Practices era in the 90s, I would date this amplifier to the 80s, at the latest. That's before Cary introduced the 6SL7 SRPP driven 300B or 2A3 AES SE-1 kit amp to the mainstream market.

I built SRPP circuits back in the day, but was never a fan of this topology. However, this amp made me reconsider my stance. Now I'm asking myself - could it be the textbook operating points and/or the 6Z-DH3A input/driver tubes which made this experience memorable? 

I truly enjoyed listening to this amp, thanks Jose!