|
Post by MartinT on Oct 8, 2014 5:41:33 GMT
Thought I recognised that cleaning pad - I have a Clearaudio RCM Nice conversion of the pickup arm, there.
|
|
|
Post by MikeMusic on Oct 8, 2014 12:26:12 GMT
Reminds me of a Paul Merton quote from years back So after seeing the Keith Monks RC "Hey, I could l knock that up in my shed !"
|
|
|
Post by speedysteve on Oct 9, 2014 16:51:06 GMT
The next instalment. I was at this point happy with the upper channels for the time being. I was convinced that front loaded horn bass was the next thing to do. So, circ saw and birch ply at the ready I started on the mid bass horns. These would fill in below the JBL2482's below about 350 or 330Hz and play down to about 90Hz. Some simulating was done first in Hornresp. It's a great tool for bass horn simulation / modelling and can do all the major profiles and helps with backbox / driver chamber, throat and mouth sizes and flare rate. Here's how the conical mid bass modelled. Only a few of dB down at 330Hz, and I figured the 2482's would fill that in... Here are the parameters in Hornresp. I modelled using an Eminence Kappa 15A - widely available cheap robust 15" driver, that is used in real bass guitar amp / monitors and has found a place in the hearts of a few Hifi DIY hornists too for it's tone and speed. Note the throat size are 325sqcm (about 20cm diameter) and the mouth at 6400sqcm (about 90cm diameter)... These were not going to be small. Gluing and build Constructed the back box - all 21mm birch ply... Where the Kappa 15A's would live Set up as a 4 way - lacked deep bass but gave very tuneful mid bass. Cello / Viole De Gambe was great! Vertical is best for imaging so I tried it all in pyramid order first. The GRF's are used here as bass bins one - Crossing at 90Hz. This gave great mid bass and put the deep bass back to where I was used to - Actually slightly better than Tannoy bass, as I was using 15" Eminence Kappa Pro LFII's instead of Tannoy 15" at this point! Next I tried the horns in a more integration friendly configuration Now a word or two about crossovers. At this point all was passive. I needed 2 x 24mH Janzen inductors and 2 x 126uF Crosscaps for the deep bass and 2 X 24mH inductors for the cutoff at 90Hz of the mid bass horns. I could perhaps have used the mid bass horns natural rolloff but did not! This gave me a 2nd order X/O on both but cost a packet. The inductors alone cost over £70 each. I later moved on to 4th order which needed even bigger 350uF caps! But that was for the GRF's replacements - more on those later;) Listening and all was not perfect by any means. It was clear by measuring that the Eminence drivers has rising impedance with frequency issues - which meant the upper X/O was not bringing the frequency down anything like 2nd order! A Zobel was what was needed. Putting that in and things started to work nicely. I was still using a single amp with 5 X/O's, Zobels on the lower two channels and loads of L-pads to balance vols to each driver - that all added up to quite a bit of wiring and 'stuff' in the signal path! I learned that each 2nd order X/O inverts the phase, so swapped the polarity of the mid bass and upper mid drivers and things started to get really nice. So now I had a true 5 way system with each driver / horn working in it's frequency comfort zone. However the deep bass was not front loaded - that would have to be addressed in some style! More about that later.
|
|
|
Post by speedysteve on Oct 12, 2014 8:39:19 GMT
So for some real deep bass - I liked the idea of tapped horns. Folk talk of horn bass being it - I had to find out... Danley uses them extensively in pro sound reinforcement - Danley sound labs I chose a tall design that would go in the corners of my room. I wanted to use the existing 15" Eminence kappa pro LFII drivers that I had used in the Tannoy cabs if possible... A bit of Hornresp modelling later and I had a design scaled to fit my room height. As a straight horn it would look like this and be 4.3m long! How the SPL / frequency response modelled Pretty flat all the way to 20Hz and sensible enough up to 90hz - where I want it. Don't worry about the combing as that will all be killed off by a 4th order X/O. It can also show the diaphragm displacement. ... Here is the horn shape if you unfolded it. It would then be over 4m long. About 2.4mm max at 10Hz, remember this is 105dB! In normal music frequency range less than 1mm is the norm. The Kappa pro will easily do 6.5mm - so no risk there. In fact the DIYaudio boys who built tapped horns to test to destruction report that the piston of the driver would punch through the cone first! Each cubic meter of air weighs about a kilo. The are probably 3-4 kilos in the tapped horn. The cone is pushing against that 3-4KG! Tapped horns are known for very little excursion and we have a 15" cone to move lots of air. Fast short through rather than slow long throw is the order of the day here! So the build started
|
|
|
Post by speedysteve on Oct 12, 2014 11:44:56 GMT
Some further pics of the tapped horn build 15" Kappa LFII baffle - some compression there - the model tells you the size needed to get the right size. One finished I tried it in place The way they work is that the front cone throw sound pressure wave goes up the inside of the horn, all the way around the top and down, where it meets with the sound pressure way of the same pulse coming off the back of the cone - this is trick of the tapped horn - it amplifies the SPL by this tuning. From the model you can see 1w should give about 105dB! Not bad for a sub;) I then made the 2nd one so proper testing could begin. The sound was pretty impressive. Doing test tone measuring sweeps had the room shaking at reasonable SPL's. My wife was in the garden a good few metres away and came and knocked on the window - "what are you doing?" - ans:"I'm doing test tone sweeps down to 20Hz", reply - "stop it, it's annoying, I can feel it through the ground" This was promising - but I'm not into boom bass - I wanted fidelity. Now for the tuning and refinement work. No sign of any rattles or resonance from the cabs - nice solid birch ply and central bracing sections:) I started of with 2nd order X/O's on the upper shelf of the tapped horn at 90Hz and the lower shelf of the mid bass horn. I then bought more caps and inductors and went 4th order - the result cleaner! Next step was to tame the room modes... Most rooms have these and with deep bass at the same SPL as the rest of the music - you can set them off and they give massive bass bloom or big suck outs. Most speaker manufacturers quote things like 30Hz - 20KHz. They may even show you very damped SPL frequency plot. What they will sometimes quote is a figure like 35Hz at -6dB. This means a whole 6dB down at 35Hz compared the the rest of the frequency covered. Even 25Hz at -6 or -3 is going to be well down at 20Hz - where you want to hear / feel that 64ft pipe Church organ. So the deep bass is well rolled off! This reduced the chances of room modes, but it won't give stonking Church organ or synth bass on modern stuff. I was aiming for a true flat down to 20Hz or as close as possible.
|
|
|
Post by MikeMusic on Oct 12, 2014 14:29:32 GMT
The set up itself is impressive Steve
To see the build from acorns up is even more so
|
|
|
Post by speedysteve on Oct 12, 2014 16:04:22 GMT
Some raw in room measurements of the bass and the other drivers / recap of X/O's - all measurements are at listening position. This is done with Holmimpulse measuring SW on laptop, a calibrated Behringer mic and Phantom pre. Keep in mind that we are aiming for +/-5dB in room. All X/O at this point were all 2nd order passives, no correction, no room treatment. All measurements done at the listening chair, ear position. Bass 20 - 90Hz - It can be seen that the 2nd order X/O struggles bit in bringing the SPL down evenly and there is a massive peak at 90Hz Also a pretty big one at 35Hz / suck out around these two peaks... The mid bass horn - 90Hz to 330Hz not much better - peaks at 110Hz and 350Hz! The mid horn JBL2482 2" large format driver (conical horn) Upper mid JBL2435Be 1.5" driver on Conical horn - 1200 - open (no upper X/O) Tweeter 11KHz - open Emience bullet tweeter The whole lot together - above 250Hz it integrated pretty well really. Next up - how to improve...
|
|
|
Post by MartinT on Oct 12, 2014 16:22:14 GMT
I've had plots like that myself, and worked hard with room treatment and speaker port stuffing to tame it.
|
|
|
Post by speedysteve on Oct 12, 2014 16:30:48 GMT
I've had plots like that myself, and worked hard with room treatment and speaker port stuffing to tame it. Yes, with a pair of conventional speakers or indeed conventional X/O's that is what you do and it can sort things, but it is room and sometimes WAF dependent of course:)
|
|
|
Post by speedysteve on Oct 12, 2014 16:33:24 GMT
The set up itself is impressive Steve To see the build from acorns up is even more so It's been an evolution to what you see now. The reasons for the evolution will be shown / demonstrated in the measurements (I hope). Of course measurements are one thing - raw sound quality and coherence etc are another - it's quite therapeutic doing this. Reinforces / confirms my decisions and what I like in a system as well as keeps my interest in it all fresh. Once I've caught up with current status, things will slow down to real/now-time:)
|
|
|
Post by Sovereign on Oct 12, 2014 16:50:00 GMT
Very impressive Steve, I also use the Eminence drivers in my OB's, at some point soon I'll be swiping out my saba 8inch mid driver for a 15 inch Altec Lancing 418B and the 100mm Saba tweeter will be replaced by a Altec 608 Horn,
|
|
|
Post by MikeMusic on Oct 12, 2014 17:02:03 GMT
If I retired I still don't think I would have the time, will power, patience and probably skill to do it all
|
|
|
Post by speedysteve on Oct 12, 2014 19:57:55 GMT
I just love making things and fiddling. So first step to improve was to put some 4th order crossovers on the bass channels and I bought 2 Antimode DSP active room correction units. I also refined the passive X/O's by adjusting the nominal calc'd values to what measured better. This got me some way to a better measurement Here the DSP Antimode room correction can be seen in action on / off plots The biggish bloom at 140Hz is still there of course as Antimode only worked on the subs. Still not so bad and not far off +/-5dB! Here is one of the Antimodes on top of the amps At about this time I started trying more amps - still passive X/O's but each amp I added meant I could remove the L-pads to those driver pairs and use the line level volumes to adjust the SPL's between the drivers. I used T-amps (as they were cheap) as well as the valve amp I had. Then I got my first EL84 SET amp - Icon Audio Lashed up crossover box - 5 in 5 out, weighed half a ton! Testing foam and felt damping Also I needed a more robust and flexible (for all the future driver and horn upgrade possibilities) rack system for the horns. I settled on 1" tubular steel that I would MIG weld. Pictured here is the test frame that would then have bracing pieces A borrowed pair of JBL2405 Alinco tweeters (there were smoother than the Eminence PA tweeters). Pictured here the stands are sort of complete and I'm using ratchet straps as a test rig. I also cut the upper mid horns down as definitely sticking about 1000Hz for these the extra size was not needed. This meant I could move the tweeter into a more rightful position to aid imaging and coherence. From the front, it's beginning to look like something:) To replace the rather unstable ratchet straps some threaded rod hangers - stable and very adjustable I beefed up the amps on the bass and acquired further SET amps as they sounded so sweet on the upper frequencies. Here pictured with some SPU's at a SPU evening:)
|
|
|
Post by Paul Barker on Oct 12, 2014 20:05:00 GMT
Reminds me of a Paul Merton quote from years back So after seeing the Keith Monks RC "Hey, I could l knock that up in my shed !" I have one of those
|
|
|
Post by speedysteve on Oct 13, 2014 11:34:19 GMT
Reminds me of a Paul Merton quote from years back So after seeing the Keith Monks RC "Hey, I could l knock that up in my shed !" I have one of those That's cool!
|
|
|
Post by speedysteve on Oct 13, 2014 12:14:44 GMT
The next thing I focused on was room integration. I read up about horn mouth throat termination. Rounded seemed better than square. Simulations of the wave forms exiting from the horns of various profiles are available in Hornresp too. Here's a lash up trial so you can see what I mean. This is foam pipe insulation cut and taped on. It proved a step in the right direction... As you can see horns can be used for many things - Swedish Christmas displays for example! A simple addition and some work with the rotary drum sander would fix what I had in mind and look pretty good too! Then there were two I then carried on and did the Mid horns - more hard work:) The bass horns do not need this mouth treatment for bass frequencies are not so picky (perhaps questionable if the mids did too, but hey ho!) Then Raal Lazy ribbons (the ultimate tweeter perhaps, came to be too). I had trialled a pair form fellow horn enthusiast and they integrated so well and sounded so sweet (compared the JBL2405's that were made to sound rough and ready). Here they are in top spec Amorphous magnet form. At 111dB / w they keep up with the compression drivers easily! Reminded me of The whole lot at that time Seen here with the tapped horns firing down the room - I learned that they worked / measured and sounded better firing across. Measurements were now improved... Left hand verse right hand Definitely getting to +/-5dB across the entire range. I modded the stands to clear the mid bass speakers (and repaired the cutouts I made in the latter) I also learned how to measure time alignment using Holmimpulse Here the first peak positives are physically aligned to ensure for example a snare drum hit creates a sound pressure wave from say all diaphragms and these arrive at your ear at the same time... Here's how it looks not aligned You just move the drivers / horns back and forth until alignment is achieved - Bass waves are long so you have to move a lot. Tweeter frequencies it is millimeters or less! This is how it looks - physically time aligned Notice the tweeter and upper mid horn are farther back - This is because you sit nearer them - also you cannot really align by eye or measurement to the diaphragms as it's impossible to tell where they actually are. I got 1 more SET amp, so had 3 5w SET amps and 2 60W T-amps to power it all. It was pretty much in this spec that I took the system to Scalford 2012. This idea came up as a few fellow enthusiasts who can and visited had commented quite early on in the game that "This lot should come to Scalford). As the system was now sounding pretty darn good - I swallowed hard and committed to Scalford! We hired a Long wheel base Transit packed up the 450KG of speakers the amps and front end and off. There is definitely a pecking order when it comes to room allocation, those who have been before get the better rooms, perhaps cos they know what to ask for... - I got a fairly crappy room called Quorn. 75SQM probably the biggest room there BUT with a low ceiling... Despite my best efforts to ascertain if the midbass hons would fit through the doors (and assurances that the doors were big - from some who had been there before), the main door to the room was just 75cm. The horn diameter was 90cm! No way to get it through. To be continued...
|
|
|
Post by Eduardo Wobblechops on Oct 13, 2014 15:39:59 GMT
Fantastic Steve. When's the pocket version coming out? :-)
|
|
|
Post by speedysteve on Oct 13, 2014 16:18:40 GMT
Fantastic Steve. When's the pocket version coming out? :-) Do you mean of the horn system or the write up? As for the system. Scottie summed it up perfectly. "You cannae change the laws of physics Captain!"
|
|
|
Post by MikeMusic on Oct 13, 2014 16:24:52 GMT
Pocket horn system Could be interesting
|
|
|
Post by Sovereign on Oct 13, 2014 17:13:48 GMT
How much did you pick up the ribbon tweeters for Steve, I was looking at some RAAL tweeters, very impressive
|
|