SpeakerTalk

clubsport911 · Posted: Mon Jan 09, 2023 1:51 pm Post subject: Reference Three's

Good afternoon folks.

As a satisfied owner of 105/3's, and listened to Reference 4's & Blade Meta's at my mates house, I have decided to move into Reference Three's - I was tempted by Fours (4.2's) but they are a touch too large.

Three's seem to offer me a little "more" of what I want - a few more Hz in the lower registers etc. I'm aware that Andrew (on the KEF design team ) used to post here and I think he moved to Elac ? If he is still on-line and active, I'd love to deep dive a little into the ethos of these designs as they have some slightly difference design concepts compared to the 105/3's.

Needless to say, my 105/3's will be up for sale soon as Director of Domestic operations has a limit to the number of boxes allowed in the AV room.

Any other Reference Three owners here ?

speakerguru · Posted: Fri Jan 13, 2023 11:29 am Post subject:

Andrew Jones left KEF in the 90s when he moved to the USA to join Pioneer, if I remember correctly.

That was around the time before we'd even started work on the Ref Series 1, 2, 3 and 4. I don't think he had much input on that series before he left. AFAIC recall. most of the work was done by myself and Richard Hunter.

At LF, I think they all used the double coupled cavity configuration, where a reflex box feeds a second ported cavity and it's this second cavity which ports to the outside.
Advantages; more extension and excursion reduction
Disadvantages: higher order cut off.

Earlier coupled cavity designs had a closed box feedning into the second cavity.

clubsport911 · Posted: Sat Jan 14, 2023 12:46 am Post subject:

Thanks. Yes, I have (over the years) become very familiar with the CC bass system on 104/2, 105/3 etc.

The Newer Reference One, Two, Three and Four had the so-called interport as you mentioned. On the KEF enthusiasts FaceBook page, I have documented the dissassembly of the Three's comparing them to the 105/3's I have. I was surprised (but not really in fact) to see just how much cheaper the modern speakers were compared to the earlier designs and it shows up in so many ways. I also noticed the Three crossover is much simpler which surprised me as from a post by Andrew, he mentioned just how hard it was to align the crossover on the interported design.

More thoughts ?

Welcome !

speakerguru · Posted: Tue Jan 17, 2023 12:27 pm Post subject:

It's too long ago for me to remember much detail. I don't recall the crossovers being simple.

WHat I can say is that compared to a reflex's two impedance peaks, a raw interport (or double coupled cavity) design has three peaks in the impedance curve as well as the usual hf rise due to the voice coil semi-inductance. This is a difficult load for any passive filter to "see" when designing a xo.
In order to get the LF section of the crossover to converge to a target response, with this load, the upper peaks would have to be, at least, partially reduced with conjugate networks (like the 104|2 which ended up as a substantially resistive load).

Can you provide a schematic of the Three's XO?

clubsport911 · Posted: Tue Jan 17, 2023 11:52 pm Post subject:

Yes, I do have a cct diagram for my Reference Three's. Interestingly, the speakers are labelled as "Three's" but have the boundary layer compensation device and the "5th wire" to a low value resistor in the crossover. As far as I can ascertain, if you short out the resistor, bass is flat (but speakers need to be away from walls / corner). If you leave it open circuit, bass is reduced for boundary placement.

How can i attach a picture on this forum ?

audiolabtower · VIP Contributor 500+ Joined: 06 Jan 2009 Posts: 686

Upload the pic to a site like Flickr which will give a url to the picture direct. That url will work in a post using the IMG links in the posting box. Google photos tend to be a pain to get to actually display.