Something stirs in the Undergrowth

[simon]

Nothing's ever finished when you DIY, it's always evolving

[IslandPink]

I know it's a bit of a minority interest on the forum these days, but I do quite like plugging away at Audio.

[steve s]

Nothing's ever finished when you DIY, it's always evolving

Too right simon, i must confess that there are many threads i only occasionally read, but i always read marks exploits

Like andrew i used to think he was going about things rather slowly... but in reality I'm sure Marks just enjoying journey

[andrew Ivimey]

I'm sure he is...!

[IslandPink]

Just wanted to re-cap on the bass shoot-out.
The slot-loaded OB was definitely in a distant third place. Quite disappointing. The bass efficiency seemed lower than the other two , very odd , the mids if anything higher than the quasar, so quite uneven. there wasn't much impression it reached any lower in frequency even at reduced efficiency , despite the lowered Fs. I have now dismantled most of it, kept the slot assembly upstairs on the off chance it might be useful for some future idea.
It would be good to hear the multi-driver versions built by JC some time to understand how it's supposed to sound.

The Tapped horn seemed somewhat better than the 285GMF in Quasar-OB mode, but only over its limited two octave band. It was more efficient at low frequencies and a bit quicker and better separated on fast bass runs. As is known, it's really a sub, not much good if you want a slow cross to a midrange or even upper bass driver.

[chris661]

As a thought - did you test with the Quasar crossover in place?
A large inductor would certainly start taking the midrange down, and with that more out of the way, the bass might be better revealed. Just a thought.

Chris

[IslandPink]

No, but I used the left amp channel with its 700Hz low-pass. The differences would be easier to hear if I'd put something lower on there, but the tapped horn would have gone a bit 'wibble' with an 8.2mH choke. I think I got the gist of the bass differences with suitable material.

OK, should re-cap on plan at the moment :
Midrange experiments need to be concluded but small drivers are good.
2 x FX120 is very good indeed and works with series cross to 285GMF in a Quasar-style. Have got fairly far towards an optimised crossover solution here.
Getting Yuichi horn to work with this is still not solved and may always suffer from the further phase rotations taking out tone & snap.
Will be relatively easy to further engineer this OB combo with a 5kHz low-pass on the 'Quasar' bit and G3 ribbon tweeter coming in to clean up the upper 2 octaves.
Urgent need to get stereo working & listen to music , so I will work now on setting up ribbon and thinking about how to best do the extra 2 amp channels.

Horns only on back-burner - have new ideas forming but some visits to certain places/people required to confirm best way forward.

[IslandPink]

On with the back-burner now....
Still wondering how a driver like the FE208Ez with a low Fs of 42Hz can be made to work for a horn delivering 80-100Hz or so with an undersized mouth. My efforts with conventional horn expansion & design norms have run into problems but I've looked at the Back-Loaded Horn designs on & off and seen how different the philosophy is there. Having heard Keith's Fostex 208 horn used with 206 at Owston and found some good sound there, I looked first at that. The expansion is very slow , these horns have no 'back volume' to play with, as the 'back volume' is the open air in front of the driver ! But, the Fostex horn looks like a bit of mess on Hornresp.
Just in the last few days I came across Ron Clarke's design for 206/208, the Dallas II. I'm intrigued by the comment he designed the horn path with fluid-dynamics software, plus some other details on the balancing of the rear horn output to the front, including baffle-step slope etc. Seemed an interesting guy and design worth investigating.
The horn is here :
http://frugal-horn.com/downloads/Dallas ... 251215.pdf
It's quite a lot different to the Fostex horn and has a much better reputation.
I have had a quick & rough go at modelling the back-horn path with Hornresp. The thing starts with a back chamber of ~5 or 6L, then goes into a small throat about half the driver area ; then it goes on via a very slow ( ~6 Hz ) first section, then a 33Hz section, then a slower ( again ) bit of 85cm with 21.5Hz, then a fast final expansion of about 75Hz to a mouth nearly twice as big as the Fostex cabinet. it's about a 2.5m length.
While the amplitude plot is not that impressive ( and needs the driver front output added to look meaningful ) the phase response looks like this :

This is impressive in that the phase is within a 45 degree envelope between 60Hz and 400Hz. It's more or less impossible to get this with a bass or bass/mid combination of mass-loaded drivers.
What also was very interesting is that the phase does not rotate wildly as the horn 'unloads' and goes into TL mode below 150Hz. The very good length of this horn ( half-wave at 67Hz ) may explain why the phase holds up. There could be some very good coherence from bass to mids with this design. The lack of low bass could easily be rectified by adding in some tapped-horn output from 100Hz down.

So, I would rather like to hear this. Not too many ( but a couple, I think ) in the UK.

[IslandPink]

Bit more work Friday and today.
Now have a three-way ie. a full-range speaker ( still mono ! ) .
Aurum Cantus G3 ribbon mounted up above the FX120's.
I re-worked the low-pass in the amp left channel to give about -3dB at ~7.5kHz . This consists of a 2.7k carbon comp resistor after the 4P1L and an 1800pF cap to ground, before the coupling cap. The Zout of the pentode is about 7250R , which adds to the 2.7k.
On the separate channel to the ribbon there is a 3.9uF cap in series. Have done a bit of level tweaking by ear and it sounds pretty nice - though may try a bit of up/down adjustment on the 2.7k resistor. Also have one or two adjustments to try on the mid/bass crossover.
Very listenable though ...

[Cressy Snr]

Cool

[IslandPink]

..ish ! - still a bit rough & ready, but 'less lashed up than last time ' !
Last night adjusted cap in bass/mid down to 100uF again, and from the usual position at the computer near the doorway, not concentrating ( standard test method ) I was surprised by a bit more upper bass tone and timing more convincing - 'Mingus' ( Joni ) sounded really nice. I think the mid/tweeter crossover needs a bit more tuning though - should be more tone in lower treble, slightly 'crude' sound compared to what's possible. Think the amp low-pass needed a bit of double-checking when I set it. See if I can install a preset pot in there for easier tuning.

On the 'back burner' stuff, I managed to find the fullrangedriver.com forum on Wayback machine. What a mine of info ! For instance, there was a bloke called Steve Cresswell crowing about some metronome speakers or something.
This thread ( if link works ) is one of those that needs saving for eternity and re-reading whenever possible :
https://web.archive.org/web/20060712074 ... .php?id=91
He was suffering from cancer around that time, and I wonder if he decided to try & write down as many details of things he'd learned as possible, so they weren't lost ?

[chris661]

Just a thought - ribbon tweeters tend to suffer if driven low, resulting in a bit of "hash" in the kHz range from harmonic distortion rising.
A steeper crossover might clear things up for you.

Chris

[IslandPink]

No, it doesn't sound like that . ps. please search for 'Pinkzilla' for our last discussion on this subject !

[IslandPink]

Andrew said :
"The owner of these speakers - simply the best speakers I have ever heard - understands almost nothing about music though he has interesting and diverse tastes. They all tend to be rather dynamic though! Phhhhrrrrroooooarrrrrrr and then a Bach harpsichord work or medieval choral piece followed by Amy Winehouse so..... wow! These Tannoys are in home made (though a Tannoy design) monstrous boxes - magnificent 15inch Gold drivers. The crossovers have been completely re-built and re-wired over years of careful work. I would say very much improved having heard and owned the odd 15inch set up from London SE27. "

So would these cabs be horn loaded- something like the Autograph for instance ?
It funny that I found myself suddenly back thinking about Tannoys yesterday while looking at Ron Clarke's stuff , it all linked back in. The 'Austin' A166 -
http://p10hifi.net/FH/download/Austin-A166Mk2-plans.pdf
..design has the usual BLH attributes of a horn lower-mid and upper bass which blends at the bottom frequencies into Transmission Line behaviour, and at 300Hz or so upwards blends into the front direct output of the driver. This one is a little odd in that the bass exhausts backwards but has a 'roof' reflector... anyway ....
Ok but then there's this 'supra-baffle' around the driver. This is designed to alter where the driver front output 'comes up' because there's the baffle-step slope also in the mix , which has to be included in the overall simulation , at least in Ron's work. So on this one he chose to extend a larger baffle around the driver to lower the frequency at which the baffle starts to help the front output.
On the later A126
http://p10hifi.net/FH/download/RonHorn- ... 260207.pdf
which was the last one he did, the extra baffle is even a little bit concave and acts as a shallow horn or waveguide for the same effect.
At this point I realised we're looking at something very much like the Tannoy Autograph & subsequent, where the back horn action blends into the front output which is bolstered by a shallow horn operating from 300Hz upwards. So it's got all the same elements, in a rather smaller & less extravagant package.

I do remember really liking the Westminsters the only chance I had to hear them. Oddly, James D doesn't like them, says he can hear the offset around 300Hz affecting male vocals, in fact he was quite rude about Tannoys in various ways. I concluded I must have a better listen some time.

[andrew Ivimey]

Tannoy GRF - He has a lovely big room dedicated to lissnin'.