David L. Ralph

20 Feb 2002

I was doing some grocery shopping, walking in the area with the shoe inserts when I spotted the name of a product, Dr. Scholl's Lamb's Wool. I immediately grabbed a pack and bought it. This turns out to have been a very serendipitous event. I have been trying to tweak tweeters because there are often differences between two in a pair, and lowering the Fs and/or reducing other deviations is sometimes possible. The MDT-30 is a good example. On my page on Morels, I show how the foam in the pole piece vent was apparently the reason for the difference in that pair. I had also found differences in several MDT-20s. The first thing I did was to measure the impedance of one of them to have a good baseline, then cut some of the felt and put it into the pole piece vent. I re-measured the impedance and frequency response. The result was excellent. The midrange was smoothed a bit, but below that it was much improved. There was a bit more extension, with a far less pronounced dip in the 1800hz range, which is typical of the five units I have. I experimented with the thickness and length of wool until I found what seems optimum. For the sample of lamb's wool I purchased, and I assume that it is likely to be fairly uniform, a piece about 1-1/4" long and about half the sample thickness seems about right, though it may be driver specific, so experimentation is certainly not discouraged.

This prompted me to try every driver I have for which it's possible to have access to the vent. Vifa XT25 and Scan-Speak need not apply. They don't need it, though.

The results of all of this testing are shown in the other pages. I've tried this on the Dayton units (old and new version), the MDT-20 and the MDT-30, plus my own "Hybrid". I had also recently tried adding a chamber to an old version of the Dayton, with some improvement. Having Lamb's wool in the vent, but without a chamber seems to be the best modification, a surprise. The Morels, however, all showed improvement with the felt, even though they already have chambers. But the most interesting one is the Dayton motor with a Morel MDT-32 dome replacing the original Dayton dome, a.k.a. "The Hybrid". It shows improvement as well. It appears (to me at least) that there must be a set of T/S parameters for the tweeters (some manufacturers even provide them) which make some more amenable to having a chamber than others. There is almost always a low frequency improvement, but the change is dependent upon the amount, hense density, of the wool. Higher density stuffing will at some point reduce extension, but this same amount sometimes slightly improves the linearity elsewhere, so it may be a tradeoff. The only way to know with assurance is to measure, but a reasonable amount should improve almost any driver, IMO.

Check the other pages for the results of each driver tested, all taken on what I call my quasi-IEC baffle. It is 1m x 1.2m with driver centered. Measurements are taken at 0.5m to extend the low frequency measurements. The method and pictures of how to make the modification are described and shown on the "Technique" page. I found this to be easy and safe, as long as care is taken with the voice coil.