SET and Fostex BLH Speakers

On the topic of the EC Studio, and again, I'm not one to mince words: but STOP FARTING AROUND WITH HEADPHONES. THE TRUE reason for the Studio for powering speakers like this.


A few people have asked me about high-efficiency speakers. You can shoehorn the Fostex 6" Sigma drivers (very good) into the Madisound kit. It's not ideal, but you can tweak around it. Yes, even though BLH have their problems, the sound of these is truly amazing. @drfindley needs to get two pairs of speakers for his setup. The Martens for his EAR, and something like this for the EC.

 

Depends on the speaker, but there may be no soldering required. Maybe just need to remove quick connects from driver terminals and re-attach. The FE168EZ has eight mounting screws instead of four, but you will find that four of the holes line up exactly with the fe166en holes.

I found the FE168EZ slightly too midbassy in respect to the Fe166en in the same cabinet. To fix this, it was a matter of stuffing fiber material into part the first leg of the horn behind the driver. Use speaker stuffing like Acousta-Stuff or wool, not pillow material or cotton. If it's still too midbassy, put the speaker on stands away from the floor or move away from the back wall. You may also find that a supertweeter is less necessary with the FE168EZ. On the other hand, since a supertweeter is less necessary, integration of a supertweeter is actually easier and will be more cohesive.

 

Since these are approximately 6" drivers in a BLH, they will suck for low and sub bass. A subwoofer is highly recommended. I would suggest a REL. I sold a REL over ten years ago and I totally regret it. They are maybe one of two or three subwoofer brands that will be fast enough to integrate with speakers. Ask @shaizada - he's way more up to date than me on the subs.

After you get a sub, you have two choices, run the BLH full range or implement a simple first order crossover (200uF should do it). The theoretical high pass point of the 200uF crossover may be rather high, but the effective roll-off really isn't that much more significant that what it is naturally (this is because of non-linear impedance characteristics of the bass for the speaker / horn design).

Running full range without a crossover will net you maximum resolution, but at the expense of muddier and less clean bass to the lower mids. Running with a crossover will clean up the bass and lower mids, but may impact resolution, add grain, or impart a small veil. In other words, do you want "Joni Mitchell mode" or "Daft Punk mode"? The crossover's effect will depend on the components; and I would advise don't go crazy on boutique caps, especially since the speaker isn't super expensive. On the other hand, you need to do what you need to do, especially if you like the sound. High efficiency wide-banders do some special things that other speaker types cannot do, namely experience SE DHT magic with the high-frequency filaments.

 

In this case probably no. The output transformers have a max output of 10W before they start distorting. It's a good idea to keep a little bit below this limit so everything remains clean. The Studio is 6W out into 8 ohms, so this is about right. I guess in theory we could get better performance if we used a small transformer with a max output of 1W and designed it specially for the HD800; but really, the difference between 1W and 6W isn't that much. I guess you can say the EC Studio was purposely built for HIGH-EFFICIENCY speakers AND headphones (which are actually less efficient than HE speakers). The Studio will not work with moderate or low efficiency speakers.

In any event, if you recall the original custom 2A3, the Levi prototype, that amp used huge huge output transformers that could take quite a bit more power. Even then, it was probably the most resolving amp I have heard back then at that time.

 

Probably fantastic. I have a pair of Altec multi-cel horns waiting to be put together in the garage. This is not my first foray into horns. I had a pair of JBL 2440 mated with Beyma woofers in a garage listening room many years back. The horns are probably 100+ db sensitive. The woofers are 94db. The only downside to front horns is that you need a big room, otherwise imaging will suffer. Also, in the context of sensitivity specs, JBL does not bullshit or fudge figures like how some audiophile speaker manufacturers will tend to do.

In EC labs, we have huge Moth Audio enclosures with coaxial 18" woofers / 1" compression driver / horn. Haven't that time to put that together. But that should tell you some of the ideas behind the Studio.

 

It's really hard to explain, but since everything is interrelated, there are different kinds of resolution. I've always had the Fostex 168EZ hooked up to lesser amps; so I was super surprised how resolving they were with the Studio. I haven't compared directly back and forth to resolving headphones, but I can say for sure that I've heard certain things on the 168EZ that I had never heard before on headphones. Even something as simple as the spoken voice on Daft Punk's Giorgio Moroder revealed a certain kind of not-so-subtle spatial / microdetail that I had never heard before ever on headphones. Likewise, I think it's probable that the HD800 will be able to reproduce the last few nanoseconds of the tail ends of decay or string transients that the 168EZ might be not able to. Yet again, much of the HD800's "detail" comes from it's uplifted treble region and top octave. As you said, a supertweeter may help.

They will do "OK", but really asking too much for a 6" driver to reproduce orchestral music. The 168EZ has actually a little bit less surface area than the lesser 166en. I'd go to a 8" driver, the 208EZ, or over to horns to do massive orchestral or choral works justice. If you are stuck on the 6" for a while, go "DAFT PUNK mode", and hook up 200uF caps in series to let those 6" drivers off the hook for the bass and let a sub take over around 50Hz or so.

 

The datasheets are for reference and don't reflect performance once in a ported box, BLH, or otherwise. The 8" high-efficiency drivers tend to have a rising response. I've noted this with the Lowther, Fostex, Moth Cicadas, and others. The FE208EZ 8" driver has a free air resonant frequency at 42Hz (by memory, I could be wrong) which is 8Hz below the 6" driver. When finally in an enclosure, the superior bass extension differential will be maintained. Bass extension another 8Hz down is significant.

Finally, it's not just a matter of bass extension. Bass requires movement of air. Without getting into the details, the more a driver has to move, the more distortion and compression there will be. Small drivers really have no business reproducing bass, or at least clean or low bass. The FE168EZ has a cone surface area of 0.0113 m^2. The 8" driver, although just a nominal 2" more than 6" has a cone surface area of 0.0201 m^2, almost double. In other words, the 8" driver will have much cleaner (and deeper) bass than the 6".

The downside to the FE208EZ is treble extension. A supertweeter will definitely be needed. Even the Lowther 8" drivers need a supertweeter IMO; although the typical audience age for the Lowthers is probably 72 or something that like - so a supertweeter probably won't matter. I guess getting old can be a good thing.

 

Three driver options (there are a lot more, but these immediately come up to mind):

1. Peerless XLS 10". Linkwitz uses these types of drivers (or were they the XXLS) for woofers on his Orion. Always wanted to build a sub based on one of these. Maybe soon. The problem is expensive drivers, and after an enclosure and plate amp with the adjust-ability that I want, I am close to the price a used 10" REL sub.
2. JBL 2235 15". Will need a HUGE box. Super sensitive essentially getting "free" power and "fast" sound.
3. Seas L26. I haven't heard this, but I don't think I've ever heard a driver from Seas that I didn't like. Short of the Accuton ceramic drivers, Seas makes the best drivers in the world. I like the low mass and low inductance of the VC. If I don't buy a REL, I will probably build off of this. Knowing me, I will probably build off this.

There used to be more on my list, but I haven't kept up, and many solid subwoofer driver vendors I once liked are out of business. TBH, I am starting to think it's less the driver and more the tuning, implementation, and setup. (REL doesn't really use any special exotic drivers, in fact some of the drivers they've used in the past were dirt cheap paper).

If I were to use the L26, I'd probably do an EBS alignment. This will require a bigger box than usual (probably 2.0 or 3.0 cuft) and a lower port frequency. The EBS alignment will have better group delay too from 100Hz down to about 28Hz (where below that point starts to get much worse). It's tradeoff in my estimation - I have very little music with a lot content below 35Hz. The EBS alignments tend to work in real rooms better where there is room gain and room modes. The downside is less power handling down low, but this will be for music, not movies, and even then for movies, we are in a home environment, not a theater.

I'd suggest a plate amp with a wide range of crossover adjustability with minimum of 33Hz at its lowest point. This pretty much rules out anything PartsExpress sells. You will have to go to Madisound. The reason for this is that I have needed to cross over that low before, even with speakers that reached 45-55Hz per spec. As for type of amp, I would personally opt for class AB instead of class D. I've had class D pro amps before and hated them - granted that I used them only for woofers 500Hz and below and not in a pure subwoofer implementation; but that experiment left me allergic to anything class D. I've used (and still own) other class D type amps - they are sitting in my garage or my closet. Of course if you have a super high excursion woofer and a HT type application, class D might make more sense.

Finally, make sure the plate amp has speaker level inputs. Feeding the plate amp from the speaker outs of your mains amp is crucial to integration. I've tried line level from a preamp, it's just never quite as seamless compared to speaker level inputs. The argument is against is that the plate amp will also amplify any distortion from your mains amp, but this is exactly the point for seamless integration.