Wow oh wow, what a combo when it’s all done and all for less than $350!? I started modding the he-300/EF3 combo a few weeks ago and have just finished all the tweaking I think it could stand.
HifiMan has priced this combo at $250, which pretty sweet when you consider they sold the headphones alone for $299 not so long ago. I thought for this price I had a low risk $ risk system to play with. Here’s what happened:
I allowed the system at least 100 hours of burn in before I played with anything.
The mods on the headphones was pretty common to begin with. I removed the felt as suggested here
. However, the treble bounced all over the place without any dampening. So I created a thinner absorption layer with two types of wool fabric: at thick textured wool around the driver and a thin wool to cover the driver openings. That brightened things up for my liking without all the distracting reflected sounds.
Next I attacked the amp. First I replaced the original Chinese 6J1 tubes with Russian 6ZH1P-EV tubes from Yen Audio. That was an easy improvement. The change wasn’t as dramatic as I hoped, since the tubes were $40 on their own. Still there is no way I could create the final result with the 6J1’s.
The third step was to start swapping out the electrolytic capacitors. The original ones were a few cheap Elnas and a bunch of Rubycons. The challenge is that the original caps are tiny compared to Elna Silmics and higher quality Nichicons. I had a choice. I could leave them alone and have a pretty view from the glass top of the amp or I could do some funky placements with inconsistent lead lengths to get them all to fit. I chose the latter.
As you can see from the pics the size difference is huge, but you knew that. It took a lot of planning to find a way to place them. I wanted to have Silmic II everywhere, but it wasn’t possible. So I used a few pairs of Elna Cerafine or Nichicon FGs. I replaced a total of 14 capacitors, but not all at once.
First swap: 2 50v 100uf. The change was an immediate clarity in the bass range. The base drum and electric bass are now two distinct sounds even when they play on top of each other. I’m not an amp engineer, so I have no idea why that frequency range changed, but it was as dramatic as the tube swap.
Second swap: 4 50v 22uf. The size difference is monstrous. At first I thought, “No way.” However with a little redirection and looking at the board from various angles, I got them all in there. There wasn’t a big change here. In general, I thought it was nicer, but that could just be pride of making a mod.
Third swap: 4 - 25v 220uf Silmic, 2 - 25v 220uf Careen (Silmics wouldn’t fit in one stage), 1 - 25v 1000uf, 1 - 35v 2200uf Nichicon FG. I ran out of patience and replaced everything else that I could fit in.
I’m not sure the purpose of the single 1000uf cap, but there was only one and I swapped it. Due to size I could not replace both power supply caps. That made me uncomfortable, but it was for the power supply so maybe an unmatched pair wouldn’t matter.
The Final Ugly Result:
The result blew my socks off. Getting rid of those tiny Rubycons make all the difference. Everything has come to life. This is a completely different amp. I’m not adept at describing sound differences, but here’s my attempt.
The best objective result I could muster: I’m going thru the Philip’s golden ear program. I was stuck in a few lessons at the gold level. However, after this mod I had no trouble making distinctions for the quizzes and passed two more exercises immediately. The difference was that clear.
The easiest subjective way to describe the final result is clarity. Clarity in instruments, clarity in sound stage, clarity in separation, clarity in bass, clarity in the moisture content in the singer’s mouth.
Maybe the best was to describe it is with Diana Krall’s Live in Paris album.
This was a live recording and you expect some reverb. However, with the original setup I couldn’t tell if it was added studio-style reverb or natural reverb from the venue. After the mod there is no question. If fact I can even hear a stage left echo, some distance away.
During applause you can make out the different hand clapping styles, rather than sounding like cloud of white noise.
The brushes on the cymbals in ‘Deed I Do' sound like brushes and not static noise in the recording (which is what I originally thought it was).
Anthony Wilson plays a Les Paul, not just any guitar. You can hear the distinctive warm resonance in the guitar body. When John Pisano is playing a classical guitar it is obvious we are hearing an acoustic. You can even pick out the occasional string buzz. In other words, there is enough detail now to enjoy even the subtle differences between the two.
Vocals everywhere are improved. Nuances in voice is clear. Also, vocals are not blended in but stand out as they should.
In “A Case Of You” the empty frequencies are empty. In quiet passages, you can hear a hacking cough in the 20th row just left of center. I don’t think he’s a smoker, just a dry throat (hahaha).
I hope somebody can explain why tiny cheap caps restrict the reproduction so much. Are they creating noise or adding harmonics - or just generally muddy. I have no idea.