Building some speakers

[Fran]

So as I mentioned before I got the panels cut and I gotta say, that makes a hell of a difference.The accuracy that CNC gives means that when you line up panels for glue-up they just all fit pretty much perfectly together. All those lining up perfectly means that you have less work to do later.. which means less chance of mistakes and a better end result. Good foundation = good result.

So before assembly, I had to cut the driver holes. The mid/woofer sits out front "on" the baffle, but the tweeter is recessed so that the plate is flush with the front of the box. This actually makes a difference sonically. So I used a router with a home made circle cutting jig to get the fit right, and then relieved the back of the hole. On the woofer panel I also needed to make the slot for the port. Again, I made a jig for the router to cut the slot neatly, and maybe more importantly identically on each speaker. I mounted the front panel of the tweeter box on the lathe to machine the rebate for the tweeter which worked out OK. The last step there as to drill out the holes for the drivers and mount the T nuts on the rear of the panel. Using bolts and T nuts means that neither the boxes or the drivers get damaged when you are dropping in and out drivers during testing.

IMG_20200523_154634-01-01.jpeg (67.14 KiB) Viewed 1059 times

IMG_20200523_160124-01-01.jpeg (54.15 KiB) Viewed 1059 times

IMG_20200523_172453-01-01.jpeg (63.37 KiB) Viewed 1059 times

IMG_20200523_174852-01-01.jpeg (75.85 KiB) Viewed 1059 times

IMG_20200523_183224-01-01.jpeg (69.4 KiB) Viewed 1059 times

IMG_20200523_184658-01-01.jpeg (53.69 KiB) Viewed 1059 times

Glue comes next....

[ConorW]

Fran - very nice work getting that tweeter in snug and flush. Looks great.
Having done my Lenco plinth in several layers of MDF I know how difficult it can be to get the cutting done accurately. Nice that the supplier did the first cuts well.
Good luck with it.
Conor

[Adrian]

Great stuff Fran! Thanks for sharing!

[Fran]

Thanks for the comments lads!

I usually use the glue and screw method of assembly but in this case I did the nail and glue approach as it means there is much less cleanup to later. I decided I would need to trick about with the box stuffing a fair bit, so I wanted to be able to remove the back if I needed to. So I thought I'd screw on the back to leave it accessible. So it was really just a case now of straightforward glue up and let set.

I'm a big fan of a stiff front panel as I think it makes quite the difference based on past experience. Although I'm using 25mm here, I thought it could be improved on, so I glued in a piece of (thin) plywood to the back of the front panel using carpet glue. This approximates towards a constrained layer, and that glue doesn't set glass hard, but rather retains a tiny bit of give. So I think that should add to the front panel. I also cut some bracing for the box, although I've often seen more than whats used here.

[Claus]

Great project! Always wanted to something like this in my dad’s workshop...
Can’t wait to hear the finished product as soon as we are allowed to all come and impose on you!

[Diapason]

Can’t wait to hear the finished product as soon as we are allowed to all come and impose on you!

I assumed there was one for everybody in the audience, no?

[Fran]

You guys are funny :)


Next up is applying some panel damping and stuffing to the box interiors.....

The idea here is to line the box to reduce reflections, and to add some "limiting" which is invisible to the low frequencies but damps the high frequencies. This means that the port only produces the low frequencies which helps clear up the sound.

Too much damping and the sound can be affected, not enough and you get other problems, So this is a case where you need to be able to add or remove stuffing as you listen. Essentially you are doing some tuning.

[Fran]

After this stage, I moved on to lots and lots of listening and tweaking. This was very much an iterative process, listen, tweak, listen, then put on an unfamiliar track and try to hold a neutral stance while analysing. So there's a test CD (now a test download) out there called "binks" that is really useful. It gives you 30s tracks at a host of frequencies, sweeps etc etc which are really useful for testing. It told me for example that I had an issue with the seal around the (temproarily held together) back, bur also that the port was working correctly as I could hear only low frequencies from it as I ran through the sweeps. I also got to find out a little more about their characteristic features. As an example, I realised that sitting back a little further gives me a better soundstage, and toeing them out very slightly helps too. They are quite directional, the sweet spot really is the sweet spot. I also figured out that angling the tweeter to ear height worked out best. Another thing is that often the tweeter crossover filter can affect phase, so in this case, reversing the phase for the tweeter gave the best most coherent soundstage.

Once this tweaking became more "mature" and the values of the crossover parts varied less and less, I felt I could move forward with buying better parts and start to think about the finish etc.

[Fran]

As mentioned above, a lot of listening and tweaking meant I arrived at the crossover values I posted earlier. This might still be subject to some change but as it stood it was good enough to go ahead and start finishing them up.

I ordered 2 suprisingly expensive options next. These were the better crossover parts - jantzen copper foil inductors and cross caps and mundorf resistors and then some veneer to finish the cabinet. The inductors are quite expensive and not that easy to find, but a supplier in Belgium sorted me out (Benny Glass - diyparadiso.com) at a decent price. Would recommend him and I know he also does tube amp repairs.

After much looking around at all sorts of options for making them look presentable, I decided to veneer the cabinets. I wanted something to make them pop a little, so I went for 2 contrasting woods. Paduak is an African wood, rich red-brown in colour with a fine grain. Most of us would call it rosewood or honduras mahogany like in appearance and that's what I ended up picking for the back and sides. So then I decided on some birds eye maple for the fronts. The birds eye figuring is subtle but yet jumps out when you look a little closer.

Both packages arrived not too long after, and in a nerdy way, it was pretty exciting to get a look at both. The inductors look serious quality, and if you subscribe to the rule of hifi that heavier is better, well, these will please you. The veneers surpassed my expectations substantially, bot were really beautiful. While waiting for these, there was much filling and sanding going on in the background so I would be ready to get working when the supplies arrived. I added magnets to the cabinets in case I wanted to add grilles later and also put in threaded inserts for feet etc.

An interesting thing happened. When I took the half finished boxes out of the system and slotted the horns back in, I was curious to see how I reacted to the horns again (I have been using them for a long time). I found that I was getting all the detail that the prototypes here gave me, but that the separation and depth were not as good. I suspected that bass ability is probably more nimble with the new speakers too.

[DaveF]

Lovely work Fran. I'd love to get a listen to them sometime. Some of the characteristics you mentioned above sound very similar to the Audioplans - a narrowish sweet spot and a very nimble bass. Out of interest do you test/tune them with valve or solid state based components ?