The Driver/Enclosure relationship

[Lindsayt]

How do the material costs compare for the MDF vs the double plywood for your project, Docfoster?

[Docfoster]

...Sound of hurried scribbling on back of envelope...

I think that in B&Q ply is about 50% more expensive than the same depth of MDF.

I've used some 25mm MDF (around £35 per sheet). And I'll use 9mm ply (£25). But I want to double-up the ply. So make that £50. Looks like the PU18 glue that will be the filling for the ply sandwich comes in at £5 per tube. I don't know how much I'll need. 4 tubes...? So £70 altogether. A mathematically convenient doubling of the 25mm MDF cost.

Tho in terms of skill, time and effort investments the plywood will be significantly more demanding. I'm not convinced it's within me. But I want to give it a go!

[Docfoster]

Joinery went OK. Not perfect. But fine enough for me.
I'm awaiting some gasket tape and some proper bolts and inserts to properly secure the drivers so it's not quite done yet.
But initial listening tells me everything works and , with some digital attenuation of the compression driver, that it sounds promising.
Once those last bits arrive, I will do some proper comparisons with the original ported cabinet and Tannoy crossover of the Eatons (bottom right of photo).

[Docfoster]

Jolly good. I love watching a diy project take shape. No bean counters to spoil the pudding. Hope you are having fun (of course you are).

[Docfoster]

Thanks Stu.
Yes indeed. It is enormously satisfying when things don't blow up, then work, then sound good.

I am going to have to put an L-Pad on the HF compression driver. Listening now, I've attenuated it (digitally) by ear by 16.7dB.
But I'll get the gasket in-place to seal the driver and let everything bed-in for a while before making a final determination and sourcing the appropriate high-wattage resistors.

[Docfoster]

Gasket tape in place. So the units should now be all sealed.

Digital attenuation of the HF unit has settled (subjectively) on -16.6dB.
To achieve this electronically I need an L-pad with a 6.8ohms in series and 1.4ohms in parallel.
I've ordered the necessary for these values.
I tend to go for high powered (at least 50 watt) wire-wound resistors. Like these.



I'm very impressed with the HF compression drivers in the 1977 Tannoy 295A units. Every other 1970s mid / HF drivers I've heard I have disliked.
Also a good amount of bass from the 10 inch parts of the drivers. Which seem to be working well with their new DIY sealed cabs.

EDIT: I use L-pads as opposed to single resistors as an L-pad keeps the resistance seen by the crossover at the same value (in this case 8 ohms). A single resistor will inevitably increase the resistance in front of the crossover, which has the effect of decreasing the crossover frequency to that driver. It is possible to correct for this by re-designing the crossover for the new resistance value, but at that point, one starts chasing one's tail.

[Docfoster]

L-Pads all fitted.
Listening is telling me the attenuation is actually too high (about 2dB too much).
Maybe too little care paid to measurement in the planning! (Possibly the resistors are at the limits of their 10% tolerances (unlikely), maybe the resistance of the compression driver isn't exactly 8ohms as my calculations assumed (highly likely).

I have a 2 ohm resistor here already. Am going to replace the current parallel resistor (1.33) in the L-pad with that. That will increase the relative flow to the driver. But that will also decrease the overall flow a bit, and reduce the crossover frequency a bit. So who knows what the overall subjective effect will be. Only one way to fine out :-D

[Docfoster]

...And that's better.

For anyone interested in a bit of DIY-speaker-porn here is a graphic inside shot. (Please ignore my inability to keep colour-coding of wires consistent.)



You can see the crossover I put together is very simple.
It's a 2nd order Linkwitz-Riley. 10uF capacitors and 2.5mH inductors.
Crossover frequency is 1000Hz, which I chose as it is the frequency of the original crossover for this driver.

The L-Pad top right is probably reducing the signal to the HF unit by about 14.3dB.
The crossover on the original Tannoy Eatons ported cabinet has a 5-step adjustable level control for the HF unit. I don't know what the steps on that are. 3dB maybe...? I liked it best when on its lowest setting (so possibly -12dB) which I though still sounded a bit bright, so that would tie-in with why I've settled on around -14.5dB for my DIY jobbies.

Oh, and to ensure everyone's got some good news, here's a shot of the reverse surface of the Tannoy 295A speaker cone...

[Docfoster]

But not good enough.

A little too much HF now...
So switched from that parallel resistance in the L-pad from 2 ohm to a 1.5 ohm.

Seems the best so far.
Will leave it there. (For now).

Just the edging around the cabs to do .
Will get that done tomorrow.

[Docfoster]

All done.



Prefer the sound of the 295A drivers in the new cabs to how they sound in the original Eatons (which are absolutely fine.)

Can't say what makes what the difference here as load of variables as well as ported v sealed. (I've used thicker MDF, and built brand new crossovers from brand new components, and have attenuated the HF unit to exactly where my ears want it.)

The bass sounds typically sealed though. Nice and tight and tuneful.

Also, the cream coloured cabinets are less obtrusive and more acceptable to others in the house.
They probably look better standing vertically. But I'll have them lying down just to keep the speaker stack as low as possible!