BL distortion from machining gap

[DevilDriver]

Just a question here that applies to XBL^2 as much as it does to anything.

I was talking to a member from another forum who said that TC Sounds told him that machining metal out of the gap (in an effort to flatten out BL) will cause BL based distortion. To me, this doesn't make any sense because I don't see how a driver would "see" (for lack of a better word) anything other than the BL field; it has no way of knowing that there could be more BL at rest.

Wouldn't you want to flatten BL by changing B rather than fooling with L and running into inductance, efficiency, and flux modulation issues?

Neil

[DanWiggins]

Sorry for the delay, I'm over here in China (in Shanghai, today) and haven't been watching to forums too closely...:)

You are completely correct; the VC only ever sees the B field of the gap. That interacts with the L of the voice coil to create the BL curve. And messing with the L part is tricky and of course much more open to errors than machining!

If you do it wrong, sure you can mess things up. Likewise with ANY topology. However, do it right and it works better!

The key is to understand where distortion from the BL curve comes from; it comes from nonlinearities in the BL curve. If the BL curve isn't symmetric, you get distortion. So you want it to be symmetric on BOTH sides of the resting position. A flat BL curve is kind of inherently symmetric out to the point where one side or the other starts to fall; if both fall off at the same, then it's symmetric there, too...

But there's also dynamics distortion - where the driver doesn't track the input signal. This isn't power compression (which basically scales the entire efficiency curve of the driver down, just like if you turn the volume down on the system), but BL-based compression. Essentially the driver gets weaker and weaker the further out it moves.

The ONLY way to combat this is with a flat BL curve; any other BL curve will not work! Not having a flat BL curve will result in BL compression, which is a frequency-and-SPL dependent effect, much like IMD. BIG problem.

So that leaves the only other issue - inductance. Steel around the voice coil increases inductance. Machining some out of the gap will lower the inductance, since there's less steel there. And then you also have a lossless place to put a shorting ring (if you do it correctly). So this only helps the inductance, not hurts it.

So I think the guy who posted that is wrong; was it TC that was wrong? Probably not, I would expect a speaker engineer to understand these basics. But I think the person relaying the information was confused. Removing some steel from the gap does not "throw away flux", or cause "BL distortion". Done properly, it steers the flux to where it's best used, and linearizes the motor.

Dan Wiggins
Adire Audio®