The Suspension System cont.

Two months have passed now and things have been slow. In that time the first thing I did was get some wire. The 0.265mm wire was too thick resulting in ridiculously low impedance, not to mention it scraping the the top plate in the air gap. So I got some 0.150mm (38 SWG) and 0.125mm (40 SWG) wire. Most of the voice coils wound in this period have all been usiing the 0.150mm wire. Winding has become a big problem and machines to do the job are very expensive, so I'll be looking at making a jig to do the winding soon.

The most notable finding in this time has been that of the suspension. The major drawback in the majority of conventional dynamic driver designs is that of the spider, as it is in essence acting as a second diaphragm. The spider however is good as what it does - returns the diaphragm to zero position as quick as possible, stores as little energy as possible, and provides lateral positioning of the voice coil (centre the voice coil in the air gap). Its importance is now obvious. However, there are alternatives which some crazed DIY folk have found that may overcome the spider's drawbacks.

The one I'm most interested is electromagnetic suspension. The idea is conceptually simple, and potentially has some great advantages. Basically this involves winding two addional coils on the former. These are above and below the voice coil and supplied with opposing DC current. In doing so, a magnetic force now keeps the former in the centre of the air gap. By altering the DC voltage we can select a target compliance.

The extra coils mean that an underhung topology would be better to keep all the coils closer in the magnetic field. This could be seen as a limitation, but since an underhung topology is supposedly better within excursion limits, I'm not complaining.

This only solves one of the two major jobs taken on by the spider. The second job of laterally positioning the former is still something that requires a fix. A bit of research has led me to believe that ferrofluid might be up to the task.

Ferrofluid is an ineresting substance. It provides many welcomed properties in addition to what I want it for, such as transferring heat away from the voice coil and providing additional damping.

After a couple of messy attempts, I realised that the best way to apply it is before the voice coil is inserted into the air gap, and with a syringe of some sort. I got some dirty looks when asking for a syringe at the chemist. I don't think she believed me when I said it was for applying fluid to a speaker.

The Basket

I have recieved a sample pack of ferrofluid with different levels of viscosity. But now I've found that to test the effects properly, I can no longer live without a cone attached to a sturdy basket.

The simplest method I could think of was to use two rings in which I could sandwich a surround held up in place by some bolts. In this way I could adjust the height as I see fit to try different types of cones down the line.

It just so happens that the material easily available was aluminium - which doesn't have the strength I'd like for the rings, but at least it doesn't conduct flux. It turns out that the whole basket structure was impressivley sturdy.

The Cone

Now for the cone. The only method that came to mind was paper mache. This would enable me to make crude diaphragm of roughly the right shape providing I could find a suitable template. The only thing I actually found was a funnel which had a shape that was too deep, but I tried it anyway.

Now instead of using the usual mix of flour water and glue, I decided to replace the flour with some cement powder. Here's the first cone, it's a huge improvement over the paper/card diaphragm used up until now. I'd go so far as saying that the first signs of non-objectionable sound are appearing. Off course I need one that's the right shape now, to find a material to use as a surround, and put it all together to test the suspension properly.

The Motor

If electromagnetic suspension proves to be the way forward then I will need a better motor to provide the required energy. It's laughable now that I simply clamped some magnets inbetween some steel to produce a motor. Playing around with a Finite Element Analysis tool, FEMM, I managed to design a motor that efficiently concentrates magnetic flux into the air gap to achieve a density of ~2T.