Monday, July 8, 2024

AliEx amorphous, Hashimoto HL20K6 + Sony TamRadio line output transformers

L>R: 6AH4 + AliEx, 12B4A + Sony/Tamradio and 6AH4 + Hashimoto HL20K6

The last time I saw my buddy J-Rob was pre-COVID, but we keep in touch via email and text messages. During one of those hello/how you doin' moments, the chat progressed to line preamp transformers and how the $100/pr. amorphous jobs from AliExpress perform. 

I used to own a pair of Tango/Hirata NP216N nickel line OPTs, which were featured in my Sound Practices issue 17 homebrewer article. In hindsight, I don't think I gave them a fair chance so this was a good opportunity to revisit the topology.


The voltages in this project are potentially lethal! Proceed at your own risk!

Inspired by Thomas Mayer's blog, I installed the amorphous pair of line output transformers in the above circuit.


AliExpress Amorphous

6AH4 + AliEx amorphous
Top trace = Kenwood AG-203 signal generator
Bottom trace = preamp output
100 Hz, 1 kHz, and 10 kHz
square wave tests conducted with
the secondary terminated with 600 Ohms
Click here for AliEx Amorphous literature and specs

Given the impressive sales literature and specifications, I wasn't expecting to see ringing in the 10 kHz square wave. At least there was no overshoot like what I found on one of the four Hammond 125ESE transformers I tested in the early 2000s. Thus, I'm not sure if I can ascribe the unwanted sibilance and graininess I heard in the upper frequencies to that visible ringing. Considering that it's the most massive transformer in this survey, I was also surprised by the tilt at 100 Hz. It looks better than the Sony/TamRadio but quite inferior to the Hashimoto HL20K6. Subjectively, the bass response of this transformer is at par with the Sony/TamRadio but the Hashimoto is superior to both. 

Inasmuch as I wanted this transformer to succeed, it sounds mechanical, dry, flat and lacking in overtones. I'm also disappointed by the lack of microdynamics. In spite of these reservations, it didn't induce outright listener fatigue. It also resolved slightly more detail than the vintage Sony/TamRadio discussed below. 

In the video above, I rewired the circuit to use the more organic sounding triode connected 6F6G. However, the pervasive dryness and lack of harmonic richness I've come to associate with solid state doesn't excite me musically.

Due to my high expectations, I was quite disappointed by this amorphous line output transformer. It's merely a mediocre budget iron with premium transformer pretensions by using an exotic core material. Finemet products from General Transformer of Japan like the FM3WS-H and FM6WS output transformersFM-ATVR1 AVC and FM-MCT1 MC SUT are on a totally different level in terms of quality and performance.

The silver lining of this AliEx shopping experience is the above homage to the Patek Philippe monopusher chronograph by Cadisen, which I ordered together with the amorphous OPTs. I've been wearing this for over six months and it has been keeping excellent time.👍 As it turns out, J-Rob is also a watch enthusiast!🍻



I wasn't going to give up just yet because even after two and a half decades, three of my SETUP buddies in Manila are still listening via their Tango NP216Ns. I also wanted to try the 600 Ohm taps in my extra pair of Sony/TamRadio OPTs courtesy of my friend Hamingredient aka Redboy in the MC SUT world.🙏 They were installed in a triode'd 6V6 line preamp project during lockdown that had a nasty hum.

6AH4 + Sony/TamRadio OPTs
Top trace = Kenwood AG-203 signal generator
Bottom trace = preamp output
100 Hz, 1 kHz, and 10 kHz
square wave tests conducted with
the secondary terminated with 600 Ohms

I replaced the AliEx amorphous pair with the Sony/TamRadio OPTs. From the outset, it was dead quiet due to the outboard power supply. Since there are no manufacturer specs available, I did frequency response sweeps guided by the AC VTVM meter of my old school Heathkit IM-48 audio analyzer cross referenced with my Fluke 87. With the secondary terminated by 600 Ohms, it's flat from 1 kHz all the way down to 20 Hz, going up to 6 kHz, it's - 1 dB and by 10 kHz, it's - 3 dB, which explains the poor 10 kHz square wave rise time. Into a high Z load like the typical input Z of a tube amplifier = 100k Ohms, it's flat from 20 Hz until it starts rising by 1 dB around 15 kHz and continued to rise wildly past 20 kHz until it hit a brick wall at ~ 75 kHz. Measured performance and the square waves would perhaps improve with the use of feedback and/or tuning pF caps here and there. However, through the years, I've learned that more often than not, such band-aids are detrimental to good sound.

My old school test bench

Scope jockeys dismiss this OPT for more than just the above reasons. Let's not forget though that audio measurements use static test tones from a signal generator, which couldn't be more distantly related to the far more complicated nature of recorded music: complex harmonies intertwined with melodies played by various instruments and/or voices weaving with each other, varying in rhythm and texture, and performed in accordance to the dynamic nuances indicated in the composer's score. That said, if you're a music lover who's literally or instinctively aware that 90% of the fundamental notes used in music making barely range from 40 Hz to 4 kHz, you'll appreciate the lush 3D harmonic richness of this transformer and won't be deterred by the aberrations past 6 kHz. 

Analogous to this phenomenon is the reaction of a few fortunate people who have experienced listening to 1920s designed Western Electric horns with field coil drivers and are held in awe because despite their limited bandwidth and phase issues, the realism achieved during that era isn't easily replicated by modern transducers designed using the latest computer simulation tools available.

Although I use old school audio test equipment to facilitate sleuthing problems, I still believe that the most sophisticated set of audio test equipment is inferior to the most unsophisticated human ear. That's why I'm skeptical of audio hardware solely designed by computer sims or for that matter, even a slide rule.

Here's the schematic of my transformer coupled line preamp using the 12B4A triode + Sony/TamRadio iron featured in the video below. On the test bench, this measured and looked identical to the 6AH4 version. This preamp actually sounds tonally closer to the 6AH4/6F6G(T) + Hashimoto below than to the 6AH4/6F6G(T) + amorphous above.

I used to pick up Sony TC-500A RTR decks which contain these TamRadio OPTs at radio shows/hamfests and flea markets for under $50. Not anymore. In fact I haven't seen any in the past couple of years. I've used these OPTs successfully in single ended 71A, 45, 1626 and triode connected 6F6/6V6 amp projects which didn't require over 35 mA of unbalanced DC current. To be realistic though, I wouldn't recommend paying too much $$s for a pair even if they contain nickel laminations. Perhaps $150/pair tops, ideally less.

This TamRadio iron reminded me of another OPT I'm fond of, which also measures abysmally but I'll leave that for a future post.😉

Hashimoto HL20K6

Motivated by the Sony/TamRadio and the strength of the US dollar over the Japanese Yen notwithstanding, I decided to take matters a step further. I sent an inquiry to Mr. Isao Asakura of Sound Tradition in Michigan for a pair of Hashimoto HL20K6 line output transformers. His quote inclusive of shipping was 10% less than the listed price in his website and was also cheaper than importing them from Japan.

6AH4 + Hashimoto HL20K6
Top trace = Kenwood AG-203 signal generator
Bottom trace = preamp output
100 Hz, 1 kHz, and 10 kHz
square wave tests conducted with
the secondary terminated with 600 Ohms
Click here for HL20K6 specs

As soon as the Hashimoto HL20K6s arrived, I wired the primaries in parallel for a 5K primary Z. The square waves show a very high quality output transformer. Only a slight tilt is visible at 100 Hz, 1 kHz almost mirrors the input signal and the rise time at 10 kHz is almost as vertical as the input with virtually no ringing.

My sonic evaluation will be brief and direct to the point - extend the frequency extremes of the Sony/TamRadio, enrich the harmonic overtones, improve the resolution of detail, micro/macro dynamics and add more air to depict the atmosphere of being in a concert hall or jazz club = Hashimoto HL20K6!

Keen-eyed readers will notice that I omitted the input level control on the chassis fitted with the HL20K6 transformers. After extensive listening evaluations, this transformer coupled line stage sounds best with the Intact Audio 28-step nickel AVCs located at the output.

My nickel plate choke loaded 76 line preamp has been resting since I installed the Hashimoto HL20K6s driven by triode connected 6F6Gs in conjunction with Intact Audio AVCs in my main hifi rig. Is this combination better? Hmmm...🤔 maybe I'll tell you later. Right now, I'm enjoying going through my LP and CD collection!😊


  1. Hah, I've had a pair of HL20K6 on the shelf for 5+yrs... Now I know what to do with them! :-D

    1. You'll be listening soon...enjoy the build!