1. Things you need to know about the new ‘Conversations’ PM system:

    a) DO NOT REPLY TO THE NOTIFICATION EMAIL! I get them, not the intended recipient. I get a lot of them and I do not want them! It is just a notification, log into the site and reply from there.

    b) To delete old conversations use the ‘Leave conversation’ option. This is just delete by another name.
    Dismiss Notice

What do you think of this tweeter response?

Discussion in 'd.i.y.' started by ToTo Man, Mar 6, 2019.

  1. ToTo Man

    ToTo Man the band not the dog

    I'm interested in what folks think is going on (or, more aptly going wrong) with the following tweeter responses.

    I'm not going to name the make or model of the speaker, all I'll say is that it's a two-way sealed bookshelf/standmount design from the late 70s / early 80s with a 6.5-inch mid/bass cone crossing to a 1-inch silk dome tweeter at 3kHz. Its anechoic frequency response was published with generous +/- 5dB margins.

    I've owned four pairs over the years (if I see them at a bargain price I can't resist buying them!). I thought they were pretty decent for their time with a lovely warm but openly detailed balance with impressive solidity and depth, and I had no complaints apart from the high treble frequencies. Not having had the ability to measure their frequency response in the past, I couldn't quite pin-point what it was I was finding odd about the treble, my best analogy at the time was likening it to a "citric zest".

    Now that I've measured them it is clear what the issue was, they have a +9dB spike at 13kHz! This spike is present (to varying degrees) on two of the three pairs I recently measured, and I'm pretty sure a similar spike would have been present on my very first pair that I owned about 15 years ago.

    [​IMG]
    [​IMG]
    [​IMG]

    Interestingly, the third pair I measured (Pair C above) don't have a sharp spike at 13kHz, but a more gentle +4dB plateau that's consistent with the quoted "+/- 5dB" anechoic response. This makes me suspect the +9dB spike wasn't part of the original tuning but has instead developed over time, either due to natural age-related deterioration of the tweeter and/or crossover components, or an abuse of signal input level. I'd be interested to hear your thoughts on what might cause such an odd looking HF response? I'd normally look at the distortion graph for clues, but this only goes up to 10kHz so isn't of much help.

    I should note that all measurements were taken on-axis with the tweeter and at a distance of 1 metre. Pair A and Pair B are no longer in my possession, I've only kept hold of Pair C as the frequency response was more agreeable to my ears.

    [​IMG]
    [​IMG]
    [​IMG]
    [​IMG]
    [​IMG]
    [​IMG]
     
  2. Tony L

    Tony L Administrator

    At that frequency it is likely to be either the driver itself or bad cabinet/grille-frame difraction or similar issues. It is far enough away from the crossover that I can’t really see how that would be to blame unless there is some seriously weird stuff going on. What is the impedance plot like? Are there any similarities with other speakers using the same tweeter? Was it a well regarded tweeter (1” seems very big for a two way crossing so high)? Did it once have ferrofluid that might have since left the building?

    PS Pair C don’t look too bad to me!
     
    Arkless Electronics likes this.
  3. ToTo Man

    ToTo Man the band not the dog

    All good questions, Tony. I'll oblige with a little more info.

    The spec sheet for the speakers says: "the diaphragm of this well-proven 25mm dome unit has been given a special treatment to improve its damping, giving an even smoother response when flush-mounted in the cabinet". What might the special treatment be, ferrofluid? The tweeter I believe is a Goodmans DT4, but I didn't think this used ferrofluid. I don't have an impedance plot for the speakers but the specs say: "8 ohms nominal with a minimum of 6 ohms at 2kHz".

    My measurements were made with the grilles off so no diffraction from those. The tweeter is however flush-mounted in the cabinet and has a metal plate surrounding it that has been glued onto the cabinet baffle (the plate is an aesthetic design feature). I guess the plate could be causing diffraction but the plates are the same on all three pairs of speakers.

    FWIW aside from the weird treble spike they really are fabulous sounding little speakers, very underrated IMO, and I'm more than happy for them to stay that way! ;)
     
  4. Yank

    Yank Bulbous Also Tapered

    Ferrofluid wouldn't be used to treat the diaphragm, that stuff goes in the magnetic gap.
     
  5. Tony L

    Tony L Administrator

    True, but it could still restrict an overall mechanical resonance by damping the voice-coil. I’ve never owned a speaker that has used it so it isn’t a subject I’ve researched much, though I am aware that it dries out and impacts treble response (e.g. Kef 105, certain ARs etc).
     
  6. h.g.

    h.g. pfm Member

    I would suggest perhaps looking at the cone, suspension and surround for some form of mechanical failure.
     
  7. ToTo Man

    ToTo Man the band not the dog

    What would mechanical failure look like on a 25mm dome tweeter?
     
  8. h.g.

    h.g. pfm Member

    Split, perished or perhaps soft and soggy.
     
  9. RustyB

    RustyB Registered Ginga

    Why the secret squirrel, for such an old speaker?
     
  10. davidsrsb

    davidsrsb pfm Member

    Impedance measurements of the tweeter will show if there is any motor resonance
     
  11. fatmarley

    fatmarley Registered User

    Bit of a long shot but the wavelength at 13000hz is 26mm. If there's a small gap where the faceplate meets the cabinet or the faceplate isn't perfectly flush, you could try some tape over it and try another measurement.
     
  12. cpg

    cpg Active Member

    Damar varnish ( google under artists supplies) is sometimes used to change tweeter characteristics. You paint the tweeter cone (not the surround!) with the varnish and this is supposed to damp the cone resonances.
    If you assume that something similar was used on the above tweeters then perhaps the following has occured:
    1) Tweeter without "special treatment". The 13kHz peak is at a higher frequency but very dominant. (check on Goodmans spec sheet)
    2) The tweeters are treated with the varnish, this increases the mass of the cone slightly, resonance reduced to 13kHz but now damped, see pair "C"
    3) Pairs A and B - After 30 years the varnish has now hardened and does not damp the peak anymore. The resonance is now very apparent at 13kHz (mass still higher than (1) )

    If you go onto the Hiquphon website (tweeters used by Linn for the Isobarik) the owner of the company says the most important thing about tweeter build is coating the dome correctly. If would be interesting to see the frequency response on an uncoated dome v. a coated dome.

    I don't think ferrofluid has anything to do with the issue, this mainly affects the frequency response at about 1- 2 kHz and the impedance curve at low frequencies. (compare tweeters with and without ferrofluid on the Seas website)
     
    Last edited: Mar 7, 2019
  13. demotivated

    demotivated pfm Member

    Tweeters often have absorbent wadding in the cavity behind the dome. Any made on a Friday or Monday in the 70s might have the cotton wool missing.
     
  14. ToTo Man

    ToTo Man the band not the dog

    The tweeter in question (DT4 I believe) has what I assume to be a wave-guide around it. On other speakers that use this tweeter, it is usually front-mounted so that the entire unit stands proud of the cabinet baffle. On this particular speaker however the tweeter is rear-mounted, so that the wave-guide sits flush with the cabinet baffle.

    [​IMG]

    I cannot locate any published frequency or impedance plots either for the DT4 on its own or for said loudspeaker. I no longer own the two pairs that exhibit the 13kHz peak, which I realise is an effective dead-end as far as being able to troubleshoot the issue, but I was really just curious about what folks think is the most likely explanation for such a response anomaly.

    PS - I will reveal the name of the speaker in due course, I chose to withhold it partly to create a sense of intrigue but also to prevent potential bias entering the initial discussions.
     
  15. Tony L

    Tony L Administrator

    Interesting, it has quite a deep waveguide/horn type thing going on, which is rather unusual for the period. I don’t recognise it/can’t recall ever having heard a speaker using it.
     
  16. ToTo Man

    ToTo Man the band not the dog

    The DT2 and DT3 were its predecessors and these tweeters were used in several loudspeakers in the 70s, including most Goodmans designs, and I think they were also used in Radfords.
     
  17. ToTo Man

    ToTo Man the band not the dog

    Time to let the cat out of the bag.... the speaker is the Goodmans Achromat Beta. I'm curious to hear @Beobloke 's thoughts on the graphs above. :)
     
  18. Beobloke

    Beobloke pfm Member

    My first guess would be dried-out ferrofluid.
     
  19. ToTo Man

    ToTo Man the band not the dog

    So the DT4 contained ferrofluid?
     
  20. Beobloke

    Beobloke pfm Member

    I would have thought so, yes. - pop it apart and have a look!
     

Share This Page

  1. This site uses cookies to help personalise content, tailor your experience and to keep you logged in if you register.
    By continuing to use this site, you are consenting to our use of cookies.
    Dismiss Notice