New Tweeters for my Mets

Dedicated to those large boxes at one end of the room
Cressy Snr
Amstrad Tower of Power
Posts: 10552
Joined: Wed May 30, 2007 12:25 am
Location: South Yorks.

#16

Post by Cressy Snr »

Andrew wrote:Nice job Steve, must have been a bit nerve wracking..
Bloody terrifying more like :lol:
Sgt. Baker started talkin’ with a Bullhorn in his hand.
Cressy Snr
Amstrad Tower of Power
Posts: 10552
Joined: Wed May 30, 2007 12:25 am
Location: South Yorks.

#17

Post by Cressy Snr »

OK ... more impressions.
These planar tweeters are really nice. Fine details, air and space are all present and correct, where a lot of it was missing with the horn tweets.
Of course I didn't miss that fine harmonic detail with the FT17Hs because I didn't know it was missing, if you see what I mean. Treble tone and texture are sweet and clean.
Imaging is first rate; easily as good as my original FE108ES mets from all those years ago. Also, the soundstage seems to have gained height information; maybe that is because of the line-source nature of the tweeters.

This exercise has revealed something quite interesting in that the requirement for a 4dB notch filter at 3KHz remains, meaning that the FF225WK driver is a major contributor to the mid-forwardness that has, off and on, been a characteristic of these speakers.
The published response curves show this 3KHz peak quite clearly on-axis. Thirty degrees off-axis flattens the peak according to said published curves. In practice however, this flattening fails to materialise; at least not in my small listening environment, hence the requirement to notch the damn thing out.
With the peak notched out, the sound is reach out and touch real; absolutely marvellous. With the peak uncorrected, it's far too in your face.

At the moment, I am using DSP to apply the filter, but clearly this needs doing via a passive method, otherwise connecting any source other than JMRC simply brings it back. I suspect it's that aluminium dust cap that is largely responsible for the peak.

That's the next job then; design and test a 3KHz notch filter, that'll not kill the sound. Meanwhile the planars or isodynamics whatever they are called, can be regarded as a success.
I love 'em
Sgt. Baker started talkin’ with a Bullhorn in his hand.
User avatar
IslandPink
Amstrad Tower of Power
Posts: 10041
Joined: Tue May 29, 2007 7:01 pm
Location: Denbigh, N.Wales

#18

Post by IslandPink »

Top-notch Steve. Great when speakers work out :roll:
If it's anything like the driver I'm working with, that'll require an air-cored inductor around 0.25 to 0.3mH .
"Once you find out ... the Circumstances ; then you can go out"
Cressy Snr
Amstrad Tower of Power
Posts: 10552
Joined: Wed May 30, 2007 12:25 am
Location: South Yorks.

#19

Post by Cressy Snr »

IslandPink wrote:Top-notch Steve. Great when speakers work out :roll:
If it's anything like the driver I'm working with, that'll require an air-cored inductor around 0.25 to 0.3mH .
Yep, that's what I have worked out. I get around 0.23mH - 0.28mH for the inductor, so near as dammit the same as yours.
Sgt. Baker started talkin’ with a Bullhorn in his hand.
Cressy Snr
Amstrad Tower of Power
Posts: 10552
Joined: Wed May 30, 2007 12:25 am
Location: South Yorks.

#20

Post by Cressy Snr »

Here's a wider view of the 5ft Metronome with its new tweeter.

All those converging verticals are a pig to photograph but it's the best I can do with the iPad camera.
The bottom front arch looks a bit odd, but never mind.
As we can see, I managed to get the tweeter front plate opening straight :roll:

Image
Sgt. Baker started talkin’ with a Bullhorn in his hand.
User avatar
Dave the bass
Amstrad Tower of Power
Posts: 12273
Joined: Tue May 22, 2007 4:36 pm
Location: NW Kent, Darn Sarf innit.

#21

Post by Dave the bass »

SSSSsssssssshody!

Nice work Steve. If you're selling the other Tweet's I 'could' be interested.

DTB
"The fat bourgeois and his doppelganger"
Cressy Snr
Amstrad Tower of Power
Posts: 10552
Joined: Wed May 30, 2007 12:25 am
Location: South Yorks.

#22

Post by Cressy Snr »

Dave the bass wrote:SSSSsssssssshody!

Nice work Steve. If you're selling the other Tweet's I 'could' be interested.

DTB
I'm looking at another smaller Metronome speaker design for a friend at the 'mo Dave so will probably use the spare tweets for prototyping. However by the next Owston, in June, I'll probably have finished with them. I can bring them along there for a tenner the pair if you wish.
Sgt. Baker started talkin’ with a Bullhorn in his hand.
User avatar
Dave the bass
Amstrad Tower of Power
Posts: 12273
Joined: Tue May 22, 2007 4:36 pm
Location: NW Kent, Darn Sarf innit.

#23

Post by Dave the bass »

<cyberhandshake> Its a deal!

Ta Steve,

DTB
"The fat bourgeois and his doppelganger"
User avatar
IslandPink
Amstrad Tower of Power
Posts: 10041
Joined: Tue May 29, 2007 7:01 pm
Location: Denbigh, N.Wales

#24

Post by IslandPink »

SteveTheShadow wrote:Here's a wider view of the 5ft Metronome with its new tweeter.
Phwooaar :!:
"Once you find out ... the Circumstances ; then you can go out"
Cressy Snr
Amstrad Tower of Power
Posts: 10552
Joined: Wed May 30, 2007 12:25 am
Location: South Yorks.

#25

Post by Cressy Snr »

OK...
After almost a week of living with the new Metronome configuration, I can safely say, the replacement of the horn supertweeter with the planar driver, is a success.

Today, I decided to try one thing before going for the 12dB/octave crossover.
I replaced the 2.2uF first order crossover with a 1uF cap, thereby moving the crossover point waaaay up. Removing the fixed L-Pad, meant that we were now relying on the efficiency of the tweeter to counteract the power drop-off at the lower end of the planar driver's operating range and so smooth the transition to the FF225WK's upper range.

With the horn tweeter, this had not worked as it screwed up the phase between the two drivers, no matter which way round the horn HF unit was connected, exposing the tweeter and emphasising the ragged top end of the wideband band unit.

With the planar tweeter, this phase response screw-up is absent. I am highly sensitive to phase errors and can spot them a mile off, whilst others hear nothing amiss.

Moving my head up past the driver array whilst music was playing, produced absolutely no phasey effects between the two units, just a smooth increase in HF content as my ear left the vicinity of the widebander and drew level with the tweeter, then a smooth drop off in HF as my ear continued on past it.

In contrast, doing this with the horn tweeter unit in place had caused all sorts of HF swooshing effects as the ear was moved past the array.

I still have the 3-4KHz passive notch filter to do to alleviate the need for DSP to be applied to the digital signal from the Mac Mini. The tone is now getting closer to where I want it.

The top end is beautiful, the tweeter change has worked, phase response is good, resulting in superb imaging, bass is fast, clean, deep and tuneful
Sgt. Baker started talkin’ with a Bullhorn in his hand.
Cressy Snr
Amstrad Tower of Power
Posts: 10552
Joined: Wed May 30, 2007 12:25 am
Location: South Yorks.

#26

Post by Cressy Snr »

Well, well, well....

I've just learned how to work the JRiver parametric equaliser and have put a -4dB cut in at 3.6KHz, with a Q of 3.4. The results have been startling. This is a fabulous tool. The notch can be moved along the horizontal by direct keyboard entry of the desired frequency and the width of the notch can be conrolled by the Q adjustment.

I had been using the JRiver graphic equaliser to try to tame the Fostex peak in the presence region but couldn't get the tone quite right, couldn't home in on the precise frequency that was causing the trouble and tune it out. Can now :)

Hours of fun.
Last edited by Cressy Snr on Sat Jan 25, 2014 10:00 am, edited 2 times in total.
Sgt. Baker started talkin’ with a Bullhorn in his hand.
User avatar
ed
retired
Posts: 5384
Joined: Thu Jun 21, 2007 4:01 pm
Location: yorkshire
Contact:

#27

Post by ed »

handle with care......they're almost like a box with the devil trapped inside.
There's nowhere you can be that isn't where you're meant to be
Cressy Snr
Amstrad Tower of Power
Posts: 10552
Joined: Wed May 30, 2007 12:25 am
Location: South Yorks.

#28

Post by Cressy Snr »

ed wrote:handle with care......they're almost like a box with the devil trapped inside.
Aye... I do feel a bit guilty for using digital eq but for now, it does the job admirably.
The fun will really start when I try to replicate that parametric eq defined digital filter, in the analogue domain, with passive components; something I am not exactly looking forward to TBH. OTOH, analogue passive filters and their phase issues are something I need to get a grip on, if I am going to get these speakers singing without electronic band-aids.
Sgt. Baker started talkin’ with a Bullhorn in his hand.
Cressy Snr
Amstrad Tower of Power
Posts: 10552
Joined: Wed May 30, 2007 12:25 am
Location: South Yorks.

#29

Post by Cressy Snr »

I have now widened the 3.6KHz notch so that it spans an octave either side of the tuning frequency. This sounds better than the narrow setting I had before.

I'm finding this a useful learning exercise as the effects of varying the filter Q factors are easy to hear. The parametric eq is highly useful as a tool to get a good notch config for the FF225WK.

I have entered the filter config created by the parametric eq into the useful online calculator linked to by Mark, earlier in this tread, which has given at least a decent starting point for a passive equivalent to feed the widebander.

Once I have allowed for the the driver's own self-inductance in sizing the inductor for the notch filter, I should hopefully make some progress this weekend towards a passive solution.
Sgt. Baker started talkin’ with a Bullhorn in his hand.
User avatar
ed
retired
Posts: 5384
Joined: Thu Jun 21, 2007 4:01 pm
Location: yorkshire
Contact:

#30

Post by ed »

does sound good doesn't it!

what I was hinting at was how do you measure these improvements you've mentioned.....

I was discussing this with Nick a while ago and I realised how dangerous this is in the hifi domain.......I use a bunch of parametrics to change the voice of instruments or disguise inconsistent playing(uneven attack etc)....but I've found if you use it in a hifi situation for instance:

an mls plot will show you dips and peaks at certain positions in the spectrum and you can target them with a parametric, varying the Q to match the offending blip......you listen to a piece of music to see what it's done and it either sounds brill or lacklustre...........it won't do the same for any other peice of music........you've possibly jeopardised the integrity of every piece of music you play in the future.....

I said with caution because it takes a long time to actually realise what you've done......a lot of listening to a lot of different things......

if you don't approach it from the mls point of view, i.e you are identifying peaks and troughs by ear then the situation is even more risky....

like I said, handle with care......you might like what it's done today, but you will likely have doubts in a months time.
There's nowhere you can be that isn't where you're meant to be
Post Reply