GAMMA 30 experiement

[Adda]

Bad news, it seems the suspension in one of the BK3013A's is sagging, and the other one doesn't have a centered magnet, or some other voicecoil/magnet related problem.

 

[Adda]

This is how the GAMMA NLP-60A will look with a 4 ohm driver except this is a sealed box (still 8 ohm VLD-12).
I may be able to get a almost completely flat impedance response in the bass region on a aperiodic box and the 4 ohm LA1231, maybe.

I really like how this looks, very flat impedance:

 

[Adda]

Another advantage to using a 4 ohm woofer is significantly improved phase integration (was decent before, not sure about accuracy of this graph though):

Edit: I'll order 4 ohm LA1231's.

 

[Adda]

I have linked to the thread over at penkiller.com, where Toshko shared these measurements for Vissokogovoriteli VKN12211/VKN112311 drivers, but I'll post them here as well:

 

[Adda]

The "new" VLD-12's I bought are in mint condition and have the VLD-13/40 style pole plates without holes, the ribbons had some residue on them though and will be replaced (red powder from the tape that held the ribbon protector).
I have ordered new ribbons form Howard Dawson Audio, and will mount them using Arctic Silver 5 paste.
I replaced the binding posts on them with better ones and plan on doing the same with the GAMMA woofers, no loose ends this time.
The only thing I'll largely ignore at this point is wiring, I'll be decent speaker wire though.

 

[Adda]

I have ordered two 4 ohm GAMMA LA1231's, now all I have to do is wait for the parts to arrive so I can get started.

 

[dumptruck]

Another advantage to using a 4 ohm woofer is significantly improved phase integration (was decent before, not sure about accuracy of this graph though)

The accuracy of that depends on a few things. First, that you are using the same type of phase data for both drivers (and that it's useful data), and second that you have an accurate setting for the offset (green circle) for both drivers.

For the first thing, most DIY people are using some type of minimum phase data, derived from the response either in their measurement program or later on in some other program. If you want Boxsim to use the phase data that was imported, you need to check the red-circled box. If you want Boxsim to derive it, uncheck the box. There are lots of workable combinations of ways to do it, and lots of ways for it to go wrong.

For the second thing, the most foolproof way a DIY person has to find the offset between drivers is to take three measurements (for a 2-way) from a single mic position: LF driver, HF driver, both drivers, or four where you add a second both-driver measurement with polarity swapped on HF. Then you simulate with single driver measurements, and adjust the offset (green circle) until it matches the both-driver measurements. For boxsim this needs to be measured on the axis midway between driver centers, which could be a problem for you because of the large size and off-axis ribbon rolloff.

In any case, the easiest way to check if the sim is at all accurate is to make a big null in the sim, such as by inverting one driver, and then try to measure it. If that isn't working you can be sure something is pretty wrong.

This is all normally totally critical stuff for a crossover design. I'm not sure how much difference getting it right will make here, though, because with such a high xover point and that spacing, getting it right might just be the difference between a vertical axis response at xover like shown below (except the off-axis lobes would be decreasing in magnitude quickly), and a more chaotic version of that.

You've probably got more of a challenge in finding a balance in sound power, since horizontally off-axis you're losing the middle of the spectrum but not much off the top, while vertically you lose the top and middle at least somewhat equally, or at least that's how I imagine this pans out, and Boxsim is not able to show you that at all (except when you have full off-axis information in the driver files like when you load the Visaton driver files).

 

[Adda]

Earlier tests where I could make sharp dips in the frequency response, caused by inverted phase, showed that the dip was near where boxsim predicted (slightly higher freq).

I typed the data in to boxsim manually, and used old graphs as my source of info, so there are things I simply cannot simulate accurately.
Boxsim can make some guesses though.

I haven't done any measurements yet, all I have done is listening tests and tonesweeps heard by ear.
Everything is good with the crossover, it sounds wonderful, all I need to do is finish the box and fill it with perfectly functional drivers.

 

[Adda]

Just a few pics for your enjoyment, great two ways with horn loaded ribbons:

 

[Adda]

The 4 ohm LA1231's have arrived and the people at hifikit.se where kind enough to remove the duscaps!

I'm still thinking about switching the new ferrite magnets for the older AlNiCo ones, but I'll have to listen to them before making that decision.

Edit: on close inspection, it seems the voice coil hasn't been properly mounted in one of them, it sits too high so it's not properly centered in the magnetic field, this is getting tedious...

Edit2: hifikit tested the drivers before removal of the dustcaps, and thinks the voice coil problem is insignificant considering the price of the drivers... sigh...

 

[Adda]

BK3013A magnets are not compatible with LA1231 drivers, the mounting holes don't line up well.

 

[Adda]

The Howard Dawson ribbons have arrived and have been mounted in the VLD's, the pase plugs have been mounted on the LA1231's, so only the box remains.

 

[Adda]

New MDF baffles have been made, and frames for the VLD's, made of plywood
Now all that's left to do is to paint the new parts, assemble everything and stuff the boxes with rockwool.

So I need to buy rockwool and screws.

 

[Adda]

Just a little something for you DECCA and Western Electric aficionados and admirers.

http://www.youtube.com/watch?v=nv0OVMHauMI

 

[Adda]

Screws and rockwool has been bought, holes drilled and paintjob begun, the speakers should be in one piece by the end of tomorrow.

 

[Adda]

More later...

 

[Adda]

That's better:

 

[Adda]

A picture of the LA1231's with phase plugs and ferrite magnets.

 

[Adda]

One of the ribbons has a problem, the woofers seem fine, I may have to experiment with stuffing though, but burn.-in first.

It seems I may have to make a new crossover, or maybe the ribbons are different from my old ones, donno.

Edit: no wait, the crossover is fine, it's just the faulty tweeter (the ribbon is rubbing the backing material).

Edit2: they need burn-in, they sound too tense and bass extension is rather poor, no different from last time I got unused woofers (Skytronic 902.222).

 

[Adda]

One tweeter down but you get the idea.