Quasar Crossovers

Christmas rather got in the way but the inductors and capacitors for the Quasar crossovers finally arrived. A rather reassuringly heavy package turned up from Hi Fi Collective with two Mundorf 6.8mH air cored inductors and two Mundorf 150uF Evo capacitors.

The series crossover is an integral part of the Quasar design, crossing over the drivers around 150Hz. This removes the bass frequencies from the wide band driver (which can only be a good thing) and limits the bass helper to the bass notes. The series crossover does some nice things compared to a parallel crossover too.

So this is James' suggested design for my Fostex 208E sigmas

And as I intend to use the 208s eventually, this is the design I used to buy the bits. Fortunately the values are near enough for the Alpair 12Ps I'm currently using too. So without too much thought I soldered the inductor and capacitor together, along with a resistor to drop the efficiency of the Supravox bass helper a touch and clipleaded to the drivers.

I wasn't expecting much of an improvement, but was rather surprised by how much things cleaned up. Before, I was running the Supravox full frequency, so rolling off about 5kHz, and using a 6uF PIO capacitor to roll the 12P off at about 4kHz. And it sounded great. There was a bit of a muddle which I put down to the band between 4kHz and 5kHz where both drivers were operating, but the clarity of the 12Ps shone through, and the underpinning of the 285s was firm and taught.

With the proper crossover in place the muddle disappeared which helped to clean the sound up quite significantly. Not only did the muddle go, but the overall detail improved as the 12P was running lower and seems a touch more articulate than the 285 above the bass helper frequencies.

But then James reminded me he'd designed a parallel crossover for the 12P as his experience with the original Alpair 12 wasn't good with a series crossover. Mark Fenlon does suggest that the 12s need as clean a supply as possible. So I changed the components around for the parallel crossover below and reconnected the clipleads.

I wasn't sure what to expect, and didn't think it would make any difference in all honesty. Well, it's early days but I think there is a change. At first I thought there was a little less punch and dynamics, and possibly a little more sibilance. The more I listen though the more detail I can hear coming from the 12Ps. I need to listen over a longer period but I think I can hear greater subtlety and tone with the parallel crossover. Just shows how small differences can have an effect.

Quasars - Open Baffles Par Excellence

Okay, okay, it's time I wrote about some proper DIY audio, enough shirking and flattering to deceive. And if you're thinking "he's a bit full of himself - first blog about DIY and he's telling us how wonderful his stuff is" then you'll be relieved/disappointed to know that I make no claims to progeny of these speakers at all. No, speakers are not my "thing". These open baffles are the design of James Doddington and we are fortunate as a community that he has been happy to share his design. Mine are fairly rough prototypes, but these are James' own speakers

The drivers are fullrange AER Mk 1s which cross over around 150Hz to Supravox 285GMF bass helpers. Both are very nice drivers and not cheap either, though the AERs are no longer available.

James has spent a long time fine tuning the shape of the baffles. By their nature baffles need a large surface area for low bass notes, even allowing for the pi/2 effect of using the floor. The shape of James' baffles use the cardiod effect to improve bass response for the given size, down to 40Hz. Nevertheless these are still large speakers at 1350mm high and 420mm wide. The size is the downside of baffles, but the upside is the lack of boxy honky colourations that cabinets usually have, and well built baffles shouldn't.

Whereas I used yucky 18mm thick MDF for my prototype build James' finished articles use 20mm thick perspex which being see through minimises the visual intrusion compared to wood. The front baffle is shaped, the sides of varying depth and the base profiled as they all enhance the sonics. There's some discussion too as to whether the perspex enhances the sound too. Here's the dimensions of James' Quasars.

In comparison my prototypes are an easy build. The front baffle is flat and 450mm wide x 1350mm tall sloped back at 5 degrees, the base is 500mm deep, and the sides are a simple triangle. It's important that the joins are all air tight. Sadly two baffles won't come out of a single sheet 2400mm x 1200mm, but spare MDF is always handy. The bass helper is mounted approximately 100mm above the base, and the full range driver at 750mm.

The crossover is usually a series one as this offers benefits over parallel. It's fairly simple, or rather there are few components: an inductor, a large polypropylene capacitor, a much smaller paper in oil capacitor to bypass the PP and bring back a little tone, and likely a resistor to balance the relative efficiencies of the two drivers. The inductor wire benefits from a cross sectional area of 2.5mm^2 as this minimises the DCR of the coil, though smaller diameters will work.

Currently I have a pair of Alpair 12Ps mounted as they're new and I thought I might as well run them in in the new baffles. My intention is to make some MLTLs for the 12Ps and use my Fostex FE208E sigmas and T90A supertweeters in the baffles, but one job at a time. I'm waiting for the crossover components to arrive, it'll probably be in the new year now unfortunately.

According to the Mark Audio datasheet for the 12P the drivers should be treated gently for the first 100 hours, then slowly increase volume and bass content and the drivers will improve for up to 800 hours! So at the moment I'm trying to resist the temptation and turn the wick up and listen gently. But the speakers already show tremendous potential.

I wasn't expecting much soundwise as the 12Ps are new and I don't have any crossovers yet but even now they're really very nice, albeit quiet. The pedigree of the 285GMF is clearly evident; it's a wonderful driver - clean, fast and tuneful. A number of James' Quasar designs have used 15" Eminence Beta drivers to excellent effect and are substantially cheaper, but the Supravox are really nice. It's early days with the Alpairs, but initial signs are good; they sound a sweet driver and very capable.

So only another 95 hours to go and I can turn the volume up a bit :-). It will be interesting to compare the 12Ps with the 208s in the fullness of time.