Fontek A-4 Minifon electrostatics with C-4 coupler

[Precaud]

I thought all my headphones had been stolen when my home was burglarized 3 years ago. But last weekend I found a bag with the dynamics (AKG K240DF, Senn HD560, Grado SR-60) buried on a closet shelf. And now the box with the Fontek electrostatics hidden on the floor. Needless to say, I am happy to see them again. These have been sitting in a box in the closet for more than 20 years.

There isn't much info out there on Fontek or the A-4's, a Goog search turns up very little. Apparently a Mr. Niwa who used to design for Stax left and started Fontek Research. I heard them at a Winter CES show in the mid-80's (where we also had an active display) and was impressed by the sound and build quality. A bit rich for my budget at the time, though. The next year I listened again and decided I had to have some. The Fontek sales rep liked our speakers, so we arranged a swap + some cash for the show demos.

A couple years later I bought a second pair of A-4 headsets so the then-wife could join in the fun.

I don't remember why I stopped using headphones and put them all away. But enough history. Yesterday I put pink noise from a mono amp through them. One of the headsets is bad, levels are low on both capsules, one more than the other. The other set still work fine. I won't have an extra stereo amp to drive them until this weekend when my son comes to visit. So for now I'll just post some pics.

Here's the whole setup. The handle from an old Tektronix portable scope is temporarily put to good use.

The manual is a small (6cm x 18cm) card, English on one side, Japanese on the other.

More to follow.

 

[Precaud]

Some more pics...

 

[Precaud]

Inside the C-4 Coupler. All passive - two transformers (in series) and two resistors per channel. Unlikely that anything will ever break in here.
What's not clear to me is how the capsules are biased/charged from this arrangement. The Stax "SB" models rectify the stepped -up voltage from the transformer, smooth it with capacitors, and apply it as DC to the center tap of the transformer. Clever arrangement. But something else is going on here. These transformers have no center tap.

 

[Precaud]

So I got the power amp from my son - it's my old PS Audio Model Two - only to discover that it does not have a volume control. Ugh. Took a couple days to find a dual 10k audio taper pot in the piles of stuff here, and install it. The amp probably needs recapping but still sounds remarkably good, especially for a 35-year-old.

BTW, the reason there is no bias arrangement in the coupler box is that these are electrets.

First listen was ok but something was not right with the right channel. Level down a tad overall, and sounded thin, and something wrong with the low end, as if everything below 100Hz was out of phase with the other ear. So I went hunting for issues.

First took apart the C-4 coupler and measured the 5 watt resistors. There's a 22 Ohm across the input and a 4.7 Ohm in series between the transformers (it uses two transformers in series per side). The 22 Ohms measure ok, but the 4.7 on the right channel measured 5.2 . Closer inspection showed it had a crinkly surface compared to the other one, a sure sign of it having been overheated/overpowered at some point (who, me?). I just happened to have some Koa 4.7 Ohm 5 watt metal films from the same series, only later. Installed it, and the level came up and thinness improved. But the bass phasiness was still there. Started poking around with the headphones on and noticed that I could press in at a certain spot and the phasiness went away. Examined each earpiece and saw that the baffle under the earpad was installed backwards on the right one! There's a raised section molded into part of it for a better seal, it was on the front of the ear and should be on the back side. Turned it around and voila, full, well-balanced bass! Now we could listen.

From the start it is obvious these are really nice, smooth-sounding headphones. They draw you in. Just a hair forward in the 400-600 Hz range but brief experimenting shows that can be fixed with felt. Like all 'statics from that era, the sub-bass is rolled off but the bass that is there is clean, tight, and punchy. They sound great with all kinds of music. Soundstage is excellent. As I mentioned earlier, I have some very good dynamic 'phones here, and none of them touch the Fonteks for engaging, smooth, coherent presentation.

Very nice to have them back in action after a 20-year break. Now to figure out where to put them... next to the computer, or with the stereo system...

 

[Precaud]

Ooops, duplicate post

 

[Precaud]

As I've been listening to these, two of the things I've wondered are: What is the stepup ratio of the transformers, and what kind of load do they present to the amplifier?

At 400Hz, with 1V RMS in, the output was 50V, so the C-4 coupler is a 50:1 stepup device.

It's DCR is 5.1 Ohms, and here is the input impedance and phase of each channel, from 20Hz - 100kHz:

(The forum software is making this GIF larger and less clear than it really is.)

As you can see, the channels are well-matched, and this is a very easy load for an amp to drive; 20 Ohms from 20Hz to 3kHz, dropping to 5.1 Ohms at 21kHz. At 10Hz it is still 12 Ohms. Bottom line: it does not take a monster amp to drive this.

And here's the schematic of the C-4 coupler with HP attached (sorry, I'm not good at drawing with a mouse...)

Two transformers in series connection. They appear to be identical; there are no markings to suggest otherwise.

From 20Hz - 5kHz, the 22 Ohm resistor is mostly what the amp is seeing. And with a DCR of 5.1 Ohms, the 4.7 Ohm resistor comprises most of the DCR. That means the primary winding resistance of each xfmr is about 0.2 Ohms. Getting this high and flat an input impedance with a 50:1 step ratio is impressive. Mr. Niwa definitely understood how to use xfmrs!

To be clear, the two resistors are there mostly for the power amp, but they do impact the 'phones. It is likely that they would sound better without the 4.7 in series, but no power amp likes seeing 0.4 Ohms at DC. Since the 22R is the dominant load for the amp, one wonders if lowering it would flatten the Z even more and extend the response a bit at either end. I may play around with that and hear what happens.

It would be interesting to compare this to the Stax passive couplers for their electret models.

 

[Precaud]

First it would be good to determine how much benefit is to be had from loading the inputs down more. I've had plenty of experience optimizing mic transformers in the past, and the same rules apply. Lowering the input loading broadens the flat impedance plateau in the middle of the curve, where the phase crosses zero, and flattens the phase as well. To optimize, you start from the standard loading and step it down until you start to see diminishing returns in flatness, then back off a step. That gives you the optimum-flatness input impedance at the highest impedance possible.

And that's what this plot shows. Stepping the load resistor down in 2 Ohm steps, starting with the standard 22 Ohm.

You can see the impedance and phase curves flattening with each lowering, until the 14 Ohm curve which is showing no phase difference from the 16 Ohm curve until the highest frequencies. That's the stop signal. So the 16 Ohm load is the optimum (the red curve). This is still a very easy load for an amplifier. It should open up the sound of these cans quite a bit.

I'll add the resistors and give it a listen.

 

[Precaud]

And the verdict is - FANTASTIC! The simplest way to put it is - more extended on both ends, and a cleanup of the sound across the entire spectrum. There is less of the 400-600Hz forwardness I remarked on earlier, but if I had to pick nits, I'd say some of it is still there. Some of it is probably coming from the amp, which needs recapping badly.

BTW, it is no accident that the "forwardness" at 400-600Hz coincides with the center of the plateau in the impedance curve, and with the zero crossing in the impedance phase. That's where the transfer through the transformers is most efficient. It is quite typical of transformers to sound strong at their impedance peak. One could come up with an elaborate passive network to dampen the impedance at that point further; not sure I want to go there, we'll see. These things need better earcups too. There is tons of leakage around these skimpy little things.

More to come in the days ahead. BTW, Stax electret owners should be looking at doing this for their steup transformers...

 

[B7RS4]

Sorry to necro this thread, the info about FONTEK RESEARCH is so scarce even in Japanese. Apparently their higher end model has some sort of built in crosstalk to achieve a forward imaging, have you got the chance to hear any of those?
It really is an interesting read!

 

[Precaud]

@B7RS4, no I havn't heard any other Fonteks. I am not a fan of adding crosstalk to simulate "forward imaging". If I need that I can listen to speakers. (BTW, the C-4 has signicant [and detrimental] magnetic crosstalk between the trannsformers...)

Since my last post, I have made some significant changes to the headphones and coupler, and like them even more as a result. I'll post details if there's interest, which I have not seen any of... I'm guessing most Fonteks have either died or been lost by now. I also built an excellent little 10W amplifier based on LM3886's to drive the C-4 and other transformer adapters i.e. Stax.

I have been tempted to make similar mods to the similar-size Stax SR-X Mk III's I have, but they are more difficult to work on. In theory, they should sound better than the A-4's but in fact they do not. The perforated plates Stax used for acoustic loading (equalize the pressure) on either side of the capsule are the major cause of the bright sound signature of these cans. Having all holes equally sized and uniformly spaced across the diaphragm is not optimum.

 

[B7RS4]

Nice! The price of these certainly make them very scarce (higher than contemporary stax, audio technica and aurex's offerings) and almost being DIY products don't help either. I also found most STAX way too bright without some extensive EQ work.

 

[Precaud]

Yes, there was an almost DIY aspect to the Fonteks. I recall them being about the same price as "equivalent" Stax back then. To be honest, had I better understood the potential lifespan implications of pre-biased electrets vs. external-bias electrostats, I might not have bought them...

And QC was an issue as well. I ordered a set of replacement diaphragms in the early-90's, and they sounded totally different than the originals, much brighter and more forward. This sound signature was very typical of so-called "high-end" audio coming out of Japan at that time. The hyped highs was the result of shoddy engineering but marketed as "detail".

 

[B7RS4]

Yes, there was an almost DIY aspect to the Fonteks. I recall them being about the same price as "equivalent" Stax back then. To be honest, had I better understood the potential lifespan implications of pre-biased electrets vs. external-bias electrostats, I might not have bought them...

And QC was an issue as well. I ordered a set of replacement diaphragms in the early-90's, and they sounded totally different than the originals, much brighter and more forward. This sound signature was very typical of so-called "high-end" audio coming out of Japan at that time. The hyped highs was the result of shoddy engineering but marketed as "detail".

To be fair the lifespan of electrets usually isn't as big as a problem as some myth made it seem to be, I have pioneer se100(two way back electrets) and sony ecr-500(uni-electrets) which both work just fine albeit made in the 70s. Also estats may suffer from coating degradation as some Beyer ET displayed.

 

[Precaud]

To be fair the lifespan of electrets usually isn't as big as a problem as some myth made it seem to be, I have pioneer se100(two way back electrets) and sony ecr-500(uni-electrets) which both work just fine albeit made in the 70s. Also estats may suffer from coating degradation as some Beyer ET displayed.

That's interesting. It could be as you say. I can't point with certainty to anything I've done that caused the two other pair of capsules to grow weak and then near-dead. My only other experience with electrets is a couple of Sony ECM-33P mics from the 80's. They're still doing fine, though one has lost a dB or two of sensitivity.

 

[B7RS4]

That's interesting. It could be as you say. I can't point with certainty to anything I've done that caused the two other pair of capsules to grow weak and then near-dead. My only other experience with electrets is a couple of Sony ECM-33P mics from the 80's. They're still doing fine, though one has lost a dB or two of sensitivity.

Seems like long term storage without usage is the main reason for discharge based on my observation. A lot of NIB electrets are very dead or have significant channel imbalance on the second hand market.