WWDT, World's Worst Dome Tweeter
Copyright 2009 © Troels Gravesen

EXPERT loudspeaker of the Seventies. Chip-board cabs, a pillow of rock-wool for damping and two coils and a cap.
This speaker was (is) actually excellent on vocals. Bass was something to only dream about although the 100-200 Hz boom might suggest there was something "down there".

My friend Michael and I were discussing the pros and cons of dome tweeters and I don't recall who asked the rhetorical question of which dome tweeter was the worst ever experienced, but with one voice we said:

Philips AD 0160/T8

This dome was one of the first affordable dome tweeters of the Seventies and found in numerous constructions. Making crossovers in the Seventies was done by calculation, a lot of trial and error - and maybe we never really did get it right.... How bad was this mylar* dome really? Time to find out, and Michael went to the local second-hand shops and came back with a pair of EXPERT loudspeakers, featuring the infamous tweeter and a pair of SEAS 21 TV-G extended range midbass drivers. 30 DKK = 6 USD. Not too bad, and they were both in good working condition. Actually Michael used to work at the factory making these speakers for the Danish EXPERT chain stores. Low-cost speakers for mass markets and made under serious economical constraints. They would simply get an order of so-and-so many speakers, this-and-this size, 2-way or 3-way and fixed cost. Crossover components had to be cut to an absolute minimum and thin-wire coils and cheap electrolytics were the order of the day. Economic constraints can be a strong driver for innovation and often they were forced to be really creative to meet target performances.

*: Mylar and similar types of polyester film are composed of polyethylene terephthalate, a specific type of polyester. Mylar is still used for small speakers, e.g. computes, ear phones, microphones, etc.
I have found conflicting information on the nature of the dome material. The PD below states polycarbonate, which was in accordance with my own memory, but some google search stated mylar. Whatever, I guess Philips must know, thus polycarbonate it is.

Thanks to Nils-Åke from Sweden, here's the original Philips AD0160 Product Description file. Download here.


An unusual design. A one-piece mold with corrugated half-roll surround suppressing resonances. Paper-wound voice coil former glued to the dome and this all glued to a metal frame fitting into a wide magnet gab. The frame holding the dome is made from iron and it's major trouble removing the dome and even worse getting it into place again without damaging the voice coil. From one of the domes here, Philips didn't get it right either and the voice coil is scraping against the magnet gab. View impedance profiles below.
The pole piece has a hole stuffed with a foam plug and on the rear top of the dome a small piece of foam - to further suppress resonances, I guess. 13 kHz peak? Anyway, quite a lot on engineering went into this dome and despite not having a copper plated pole piece (symmetric drive) it displays a flat impedance profile.

Further info from Nils-Åke:
Thanks for the nice vintage things you still keep doing ones and a while! Slightly vintage nerdy as I am (unfortunately... or maybe not...) I though I´d share somethings about the Philips domes. The one you remembered using many years ago with the flat surface is (from all of them Ive seen) numbered differently, 0140 T4 or 8, not 0160. The 0160 exist in various types, some with a lager black reflector spot in front of the dome (I think 14 or 15mm if I remember correctly) and some with smaller (around 10..?) and the polepiece in some are slanted inwards, some are flat and the amount of stuffing varies... stuffing them harder usually lowers the impendance peak a lot and clears up some of the rubish some have around the peak also.. it also changes the FR slightly and Fs goes down. Room for experimentation there. Some have a small amount of foam (that usually disintegrates over the years) on the top of the dome, right under the black "spot" on the grille, some have not. Olle Mirsch also had some specially made for his speakers with different mods, some of those with diffusor fronts. Those I think have the modified polepieces and I think I tried putting a 0140 front on one of those with pretty nice result many years ago... dont remember exactly what I did with it though... the years go by... ;-/


How should this tweeter be mounted? Probably not like this. A major bump at 7.5 kHz due to the deep moat around the face plate.

This looks better (red curve) and is probably how it was intended. The version I had in one of my Seventies' construction had a flat faceplate for proper flush mounting.

Here the two domes flush mounted and displaying a nice response profile, steadily rising towards 17 kHz from where we see a rapid decline.
Some disturbance at 13 kHz, most likely related to improper damping of top and cavity of center polepiece.

Left: Impedance of the two domes, one compromised by skewed centering. Right: Step response from best dome.

Distortion test done at 0.25 m distance, normalised for 2.8 volts/1 meter.
Left: 2nd (blue) and 3rd (green) harmonics shown. Right: 20 dB scaling and 4th (yellow) and 5th (purple) harm. included.
These tweeters are doing remarkably well compared to any modern dome. Read distortion test

Left: CSD, 20 dB scaling mounted according to example A.
Right: CSD, 20 dB scaling mounted according to example C. Minor trouble at 13 kHz, otherwise fairly clean.

CSD, same as above with 40 dB scaling.

The AD 0160/T8 is a fairly sensitive tweeter, easily making up to a 90-92 dB midrange and as can be seen here, used with a SEAS midbass tuned to 95 dB.
The /T4 version would make 93-95 dB, I guess, having the same magnet.


The AD 0160/T8 is not an easy driver to model. Rendering a reasonably flat response takes a minor equalising step. Above a 3rd order filter @ 3 kHz.

Just for fun: The SEAS CA18RLY set up with the Philips AD 0160/T8 in a 2-way system.
I'd be surprised if this confirms our prejudice of "the world's worst dome". It most likely doesn't...