View Full Version : Choosing drivers for a given enclosure and desired sound.
I've a little project to get on with.
Been enjoying my Celestion SL700s - nice natural sound, great feeling for soundstage - but my Acoustic Energy AE2 Pros have a tactility (speed of driver?) that I miss..
So, in the vein of the SL700s, I saw some Celestion Kingston cabinets for sale and bought them! The Kingstons were supposed to improve upon the SL700s in many areas whilst keeping some of their strengths.
Anyway...
How does one go about choosing drivers for a given cabinet design, shape, construction and volume?
Does it end up as a case of taking a chance on some that have worked in other designs and the rest is down to the cabinet design (fixed) and crossover? And a lot of trial and error?
Or is there a way of working out what will work?
When I read threads on say DIYaudio.com, it seems people have some idea of a pair of drivers they want to use together first, then come up with a cabinet and cross-over design afterwards - like geek gear lust over the drivers happens first.
Also, what kind of driver property will give a "tactile" sound? What I mean by this is that something sounds so real in terms of immediacy of sound that it could be there in the room - kind of no veils infront. Is it it a dynamic ability, an ability to push the air quickly from a standstill without flapping about - stiff cones and big magnets/coils? Or simply it's down to everything - the overall result and can't be attributed to drivers alone?
Having said that, it would be best to find a mid at least that fits the mounting holes that are already there - might restrict things quite a lot.
Anyway, any help would be appreciated.
Easy way out would be to buy some Celestion A100s as I think they have the same drivers.. but I'd like to push them further, treat the cabinets to modern drivers.
How does one go about choosing drivers for a given cabinet design, shape, construction and volume?
It aint as simple as that.
First you will need to get to grips with the likes of WinISD or similar free download speaker design software. http://www.linearteam.dk/ Or you could start here: http://www.mh-audio.nl/spk_calc.asp You will need to input all the theile parameters into the program for your drivers to get a suitable cabinet volume after choosing which type of cabinet alignment you want.
If you are intending on a multi-way design you'll then have to get to grips with crossover designing. A minefield in it's own right. Again there are free software download programs on the web. try here http://www.diyaudioandvideo.com/Calculator/XOver/ or for a 3-way design here: http://www.diyaudioandvideo.com/Calculator/APCXOver/
Good luck. It has taken me about 3-years to get confident with speaker design and then I'm still a relative novice at it.
http://www.speakerworkshop.com/
http://www.linearteam.dk/default.aspx?pageid=winisd
my preference
http://www.gedlee.com/speak.htm
Good quality drivers are expensive and yes you can notice the difference.
For Bass and mid-range my preference is still for paper cone with fairly steep angle, long pole, narrow coil to pole gap with minimum voids and powerful magnets.
For high range, I much prefer silk domes with ferro fluid cooling, they just seem easier on the ears.
You can’t just swap drivers around in a pre built enclosure and expect to get a decent sound. If you change the driver, the probability is you’ll need to adjust the crossover.
You can’t do this by ear, you have to measure and there is a lot to measure, including the impedance curve of your driver.
Once you’ve built the speaker and achieved the best measurable performance then comes the listening and adjustment bit. A completely flat response does not necessarily make for the best listening experience.
Proper cabinet construction will make a much larger difference to a driver’s performance than the average loudspeaker design would have one believe. For drivers in a box you just can’t have an over build enclosure. Yep, I know what those who are fans of the floppy BBC style resonating cabs believe. In the vast majority of builds I think they are wrong; the physics just doesn’t make sense.
Daniel Quinn
26-06-2012, 10:48
Whilst I wouldnt disagree with anything said above , I do think there is a tendancy to overcomplicate , part of the fun is trust trying things . There are plent of reasonably priced drivers on ebay you could look at and not so inexpensive that a mistake would matter much . I did exactly like you bought a cabinent and then set about placing drivers in . The crossover was a ball ache to my none technical brain , again I bought completed crossovers and parts of ebay and just experiemented by ripping then apart and swapping and changing , in the end -thanks to a suggestion from Mr Dunn , I foresake a crossovers except capacitor and resistor on the tweeter and blocked up the port and wired in series not parrallel the results were excellant .
I do however note the cabinent you have bought is a special polymer [alpha crystal ] and I wouldnt have thought there was to much scope for additional bracing or modificiations . So you couild contact Wilmslow Audio with the dimensions of the cabinent , internal volume etc and they will offer a suggestion of drivers and appropriate crossovers at various price points or a design schemaitc for a charge . Drivers are fairly standard in size so I wouldnt have thought you would have to mcuh trouble find appropraite sizes .
Whilst I wouldnt disagree with anything said above , I do think there is a tendancy to overcomplicate , part of the fun is trust trying things . There are plent of reasonably priced drivers on ebay you could look at and not so inexpensive that a mistake would matter much . I did exactly like you bought a cabinent and then set about placing drivers in . The crossover was a ball ache to my none technical brain , again I bought completed crossovers and parts of ebay and just experiemented by ripping then apart and swapping and changing , in the end -thanks to a suggestion from Mr Dunn , I foresake a crossovers except capacitor and resistor on the tweeter and blocked up the port and wired in series not parrallel the results were excellant .
I do however note the cabinent you have bought is a special polymer [alpha crystal ] and I wouldnt have thought there was to much scope for additional bracing or modificiations . So you couild contact Wilmslow Audio with the dimensions of the cabinent , internal volume etc and they will offer a suggestion of drivers and appropriate crossovers at various price points or a design schemaitc for a charge . Drivers are fairly standard in size so I wouldnt have thought you would have to mcuh trouble find appropraite sizes .
This is my take but comes from my technical ignorance but being practically minded and being bogged down in thoughts too often that just make you freeze - "am I making the right choice here?" etc.
This 'Alpha Crystal' is kind of Celestion's way of making it sound like they'd invented it... In the real world it is called Jesmonite (the trade name) and can be bought from art and theatre suppliers - good for making sets look like marble for example and is non-toxic too. I bought some cos I wanted to make both a turntable plinth from it and design a speaker like the Kingston.
Thanks for the tip about Wilmslow Audio - they might have something they can run the dimensions into and advise drivers.
http://www.speakerworkshop.com/
http://www.linearteam.dk/default.aspx?pageid=winisd
my preference
http://www.gedlee.com/speak.htm
Good quality drivers are expensive and yes you can notice the difference.
For Bass and mid-range my preference is still for paper cone with fairly steep angle, long pole, narrow coil to pole gap with minimum voids and powerful magnets.
For high range, I much prefer silk domes with ferro fluid cooling, they just seem easier on the ears.
You can’t just swap drivers around in a pre built enclosure and expect to get a decent sound. If you change the driver, the probability is you’ll need to adjust the crossover.
You can’t do this by ear, you have to measure and there is a lot to measure, including the impedance curve of your driver.
Once you’ve built the speaker and achieved the best measurable performance then comes the listening and adjustment bit. A completely flat response does not necessarily make for the best listening experience.
Proper cabinet construction will make a much larger difference to a driver’s performance than the average loudspeaker design would have one believe. For drivers in a box you just can’t have an over build enclosure. Yep, I know what those who are fans of the floppy BBC style resonating cabs believe. In the vast majority of builds I think they are wrong; the physics just doesn’t make sense.
I do actually have some tweeters - Scanspeak 97000s. I picked up 3 for £80, used on ebay about 4 years ago to replace the tweeters on some Kef 107s as they have the same mounting as a Kef T33 and similar impedance plots (as if they were designed as a direct replacement). I also have a pair of ribbon tweeters, same as would be on a pair of ProAc Future 1s. Both I think may be too deep for mounting area on the cabinets - the original speakers would have had a slightly curved front baffle (not on these empty cabs) so perhaps I could make something to accomodate the Scanspeaks.
Cabinet is already built but despite the claims of innertness by Celestion, it is still a uniform material and bongs - slightly like a damped ceramic pot. There is room for improvement and I'm sure bracing could be set up inside.
I do imagine that the cross-over/driver interaction will be the most fiddly bit.
I could cop out and find a pair of Celestion A100s - I believe they use the same drivers in a wooden cabinet, like SL6s were to the SL600. Wanted to improve things though if possible.
It aint as simple as that.
Aarrgh! Don't say that... :lol:
First you will need to get to grips with the likes of WinISD or similar free download speaker design software. http://www.linearteam.dk/ Or you could start here: http://www.mh-audio.nl/spk_calc.asp You will need to input all the theile parameters into the program for your drivers to get a suitable cabinet volume after choosing which type of cabinet alignment you want.
If you are intending on a multi-way design you'll then have to get to grips with crossover designing. A minefield in it's own right. Again there are free software download programs on the web. try here http://www.diyaudioandvideo.com/Calculator/XOver/ or for a 3-way design here: http://www.diyaudioandvideo.com/Calculator/APCXOver/
Good luck. It has taken me about 3-years to get confident with speaker design and then I'm still a relative novice at it.
Fortunately the cabinet is already built... That would involve too many variables otherwise.. although I would like to make some curved floorstanders one day.
Thing is, I see people like Hart Audio finding a brass speaker enclosure (I started a thread about it) on ebay, ex-B&W design, taking just a few weeks to stick some drivers in with a cross-over and then ask £10,000 for them. Now that could just be BS but there must be a certain level of sound quality to even consider the possibility of selling them at a price higher than others in your range. Therefore, perhaps you don't need such complexity and over-thinking?
Will the programs you've mentioned above do things the other way around? Enter enclosure details and get a range of Theile / Small parameters (I added in the "Small" as it rang a bell in this context - a bluff!)
As an additional question - on the revelation that I alread have some Scan Speak 97000 tweeters (should they fit) will Scan Speak already have a mid/bass driver that suits their own tweeter and so much easier to integrate?
Are drivers from the same manufacturer the best bet?
Reid Malenfant
26-06-2012, 17:34
As an additional question - on the revelation that I alread have some Scan Speak 97000 tweeters (should they fit) will Scan Speak already have a mid/bass driver that suits their own tweeter and so much easier to integrate?
Are drivers from the same manufacturer the best bet?
You could go that route, some manufacturers go to reasonable lengths to get a mid/bass unit to integrate with a tweeter using as simple a crossover as possible by building it to roll off it's upper response in a controlled way.
Dynaudio drivers spring to mind in that category.
Overall it sounds like you want to take a few factors into account when you choose a driver. You say you want "immediacy" so I'd tend to steer clear of polypropylene coned drivers. Your best bet if you like what the AEs do is to look for a good quality metal coned driver as that is what the AEs use.
Scanspeak make some very high quality Magnesium coned bass/mid drivers, though they aren't exactly cheap. Another alternative & cheaper, but still rather good may be take a look at some of the aluminium coned drivers made by either Visaton or dare I say it HiVi.
The problem you may well have using a metal cone is you'll probably need to use a steep low pass filter for the crossover to avoid the effects of cone breakup & the resultant resonances that tend to be produced.
At the bass end don't consider any bass/mid with a total driver Q (QTS) of above 0.4 :) 0.4 & below will mean the driver can be used in either a sealed box or a reflex without too much overhang. It'll be well damped, though if you do intend to make some kind of reflex enclosure I'd suggest you look for a driver with a QTS of no higher than 0.33 in reality. The lower the QTS, the more control the amplifier will have over the cone & it'll start & stop faster.
As it's likely to be small, 5" to 7" or maybe 8" at a push with a two way speaker I'd suggest you don't go below a QTS of 0.3 as you'll lose too much deeper bass due to it rolling off.
Clever enclosure design can help out here though, you can use room boundaries to incease the deep bass you'd get from the port to create a flatter in room response.
In brief though, could I choose any tweeter/woofer pairing (that work well together) and adjust the crossover for the cabinet/driver combo or do I have to match the drivers to the cabinet?
Is it only bass resonance that I need to match up with the drivers?
The Kingston cabinets are sealed designs with a tapered rear.
http://www.audiocostruzioni.com/r_s/diffusori/diffusori-3/celestion-kingston/fianco_small.jpg
You have to match the bass/mid driver to the internal air volume of the cabinet - that is why the cabinet is designed after the drive units have been chosen.
Reid Malenfant
27-06-2012, 16:28
In brief though, could I choose any tweeter/woofer pairing (that work well together) and adjust the crossover for the cabinet/driver combo or do I have to match the drivers to the cabinet?
Is it only bass resonance that I need to match up with the drivers?
The Kingston cabinets are sealed designs with a tapered rear.
http://www.audiocostruzioni.com/r_s/diffusori/diffusori-3/celestion-kingston/fianco_small.jpg
If you were to pick a pair of drivers that already happen to have a crossover designed to pair them up then you won't need to mess about with the crossover at all, in theory. In practise you may need to alter the DC resistance of the choke feeding the bass/mid driver, depending on if you are going to change the bass loading of the enclosure. If you aren't then you won't need to do a thing, if you are you still may well not need to, it just depends on how anal you want to be :D
You have to match the bass/mid driver to the internal air volume of the cabinet - that is why the cabinet is designed after the drive units have been chosen.
I agree, you do have to match the driver to work with the enclosure.
However, I think you'll find that it best to decide on what size enclosure you can live with & then look for a driver that'll work with the allotted volume.
Doing it arse about face & buyng a driver first can often end in tears, when the designer realises that he'll need to build something the size of a coffin to work with the driver, isn't a good idea :eyebrows:
You have to match the bass/mid driver to the internal air volume of the cabinet - that is why the cabinet is designed after the drive units have been chosen.
Yet most of the speakers in the world use generic drivers and the designs are all completely different...
Volume of a set cabinet can always be reduced too (I even have some jesmonite/alpha crystal to use) so as long as the driver doesn't require a minimum volume that is greater than these then there should still be a good choice.
These share exactly the same drivers as the wooden, square box of the Celestion 100 so if it comes to it I can at least resort to devouring some 100s - normally about £150 sometimes less, just not many around.. Probably not much different to the A1's drivers either.
Having had a brief look, the choice might be mostly restricted to driver size and mounting arrangement. The hole for the mid/bass runs very close the edge which makes the common 5 or 6 hole mountings difficult to use and most four hole mounting drivers with square basket edges are the cheap crappy ones... :(
Unless there are some top drivers of the last 10 years or so now out of production I could find?
Yet most of the speakers in the world use generic drivers and the designs are all completely different...
But they don't - that may be your perception, but it is a gross over simplification.
However, I quite agree that the drive unit must be chosen with a target volume in mind - you can't set out to make a small stand mount and then find that your chosen drivers need a cabinet the size of a wardrobe!
What you can't do is choose two random drivers you like the look of and put them in a box with a random internal volume (well you can, but the result will be very coloured and uneven). If you look at some of the software that was recommended to you earlier in this thread, you will quickly see that the volume and loading of the box has a profound effect on the speakers output, depending on the Thiele-Small parameters.
But they don't - that may be your perception, but it is a gross over simplification.
However, I quite agree that the drive unit must be chosen with a target volume in mind - you can't set out to make a small stand mount and then find that your chosen drivers need a cabinet the size of a wardrobe!
What you can't do is choose two random drivers you like the look of and put them in a box with a random internal volume (well you can, but the result will be very coloured and uneven). If you look at some of the software that was recommended to you earlier in this thread, you will quickly see that the volume and loading of the box has a profound effect on the speakers output, depending on the Thiele-Small parameters.
Probably is a gross over simplification I agree - was thinking of all the Kef drivers of yore and scanspeaks in different guises, accutons recently, focals, volts, ATCs, blah blah. Appear to be drivers created as OEM and then integrated into manufacturer's designs.
I think what I was interpretting from your post was that matching is very specific and that there is only one driver to match any one cabinet size.
Of course there will be many designs that match a driver, dependant on the construction and how much of the volume the driver 'sees'.
I would only need a driver to match the volume in a general ball park and the cabinet can be adjusted, stuffed differently (not sure what would have been in an original Kingston). As I say, I can reduce the volume easily.
My SL700s have had the internal foam removed and that kind of circular-ribbed matting placed on the internal walls instead. Cross-over has been re-worked and placed on the outside. Someone else did it and quite well and I'm loving their sound.. I wonder if they just re-built the xover based on the standard values or adjusted to account for the change in internal properties..
hi Nat,
stuffing a speaker doesn't alter the internal volume , on the face of it making speakers is simple BUT as others have said theres a lot more that goes on than most realise.
nat8808 It seems to me from your posted replies you have made up your mind to go your own way with these speakers and ignoring the vast knowledge of those who have tried to help you regardless the advise given . If you want go ahead, buy some drivers that you like the look of and are of a size that fits your box but don't expect them to sound good.
I find such folk annoying in the least. You ask for advice, are given it, then post up arguing that advise and knowledge of those that know. So why ask the question(s) in the first place when you have no intention of taking up that advise?
Sorry you feel that way.
The cabinet design is FIXED.
Your advice Toppsy was:
You will need to input all the theile parameters into the program for your drivers to get a suitable cabinet volume after choosing which type of cabinet alignment you want.
What drivers? I don't have any drivers! That is the whole point.
If you want to ramble on about best speaker deisgn practice then add your own blog to list of hundreds around the web. I guess in asking for advice I'm seeking true response rather than monologue - maybe that mismatch has caused the annoyance you've experienced previously? Certainly it's a bit annoying for people to ignore the original post content and go off on one about something else and then complain I'm not taking their advice..
To be clear, I have these cabinets already and I would like to turn them into a working pair of speakers best I can.
I can find the original drivers (same drivers are used in different Celestion design too, yes different design/same drivers) but hoping I might be able to improve on the original design given more modern technology and commercial cost constraints of the original.
I guess people's own blogs/threads (real experiences) of upgrading/replacing drivers to improve things gives me the belief I can do it too! Hoping someone might have previous experience and be able to give advice..
So far I've taken the infered advice to investigate what parameters I need to go for for the given volume. I might cop out a bit in that I've asked Wilmlow Audio too and so they might be able to come up with a selection.
I was also wondering about how wide a range of parameters values I can choose from to then hone in the match via the crossover design, if you see what I mean.
hi Nat,
stuffing a speaker doesn't alter the internal volume , on the face of it making speakers is simple BUT as others have said theres a lot more that goes on than most realise.
Depends on accoustic density/properties surely? Stuffing a cabinet with something with the same properties as the enclosure would be the same as making the cabinet smaller for example.
Maybe it can be equally considered mathematically as an integral of all the different 'seen' cabinet sizes by each frequency when finding the overall response of a speaker by using internal stuffing?
So is there a way to or will those programs mentioned before be able to guide me to the right driver parameters for the given/fixed enclosure volume of these infinite baffle cabinets?
I'll work out the volume of them by filling them with water I suppose. One of the benefits of a resin cabinet!
or just use simple maths to work the volume out
Reid Malenfant
02-07-2012, 18:59
So is there a way to or will those programs mentioned before be able to guide me to the right driver parameters for the given/fixed enclosure volume of these infinite baffle cabinets?
I'll work out the volume of them by filling them with water I suppose. One of the benefits of a resin cabinet!
Download winISD (it's free), you should be able to find out the driver parameters from any decent manufacturer.
Don't use water, use your head :lol: Go get yourself a load of styrofoam balls used to fill beanbags, it might cost a bit more, but it weighs a damn site less & is re-usable. I think I have about 4ft³ of the stuff here, which I used to accurately size a chamber of a 4th order bandpass or three :eyebrows:
If you are making a sealed box then you have quite a few more drivers to play with. You can range closed box Q from 0.5 to just over 0.7 & still get a clean response. If you don't mind stuffing a box then I'd recommend you use rockwool insulation, I haven't found anything better, even the holy grail - long haired wool isn't as good. You'll be able to use an enclosure with 20% less internal volume & get the same response as an unstuffed box.
Or if you like, a driver that might have been a little out of the question could now be used :cool:
or just use simple maths to work the volume out
Dunno.. Volume calculations of curved shapes aren't my forté, and I don't know the type of curve or equation for the curve..
Would seem quicker to hold in the gap of workmate and fill with water, litre by litre.
Download winISD (it's free), you should be able to find out the driver parameters from any decent manufacturer.
Don't use water, use your head :lol: Go get yourself a load of styrofoam balls used to fill beanbags, it might cost a bit more, but it weighs a damn site less & is re-usable. I think I have about 4ft³ of the stuff here, which I used to accurately size a chamber of a 4th order bandpass or three :eyebrows:
If you are making a sealed box then you have quite a few more drivers to play with. You can range closed box Q from 0.5 to just over 0.7 & still get a clean response. If you don't mind stuffing a box then I'd recommend you use rockwool insulation, I haven't found anything better, even the holy grail - long haired wool isn't as good. You'll be able to use an enclosure with 20% less internal volume & get the same response as an unstuffed box.
Or if you like, a driver that might have been a little out of the question could now be used :cool:
Using my head says that styrofoam balls will be sticking to things for ever and finding themselves in cracks and stuck to plastic surfaces! Aaargh! And that water is readily available from the tap..
Thanks for the Q advice - I'll download the bit of software soon too. Having to put this on hold for a week or so as I'm going to be away from home for a few days.
StanleyB
05-07-2012, 23:45
When I used to live in South America in the 70's the father of a mate of mine used to have a subscription to a US HIFI magazine. I remember seeing a speaker in there that was filled with sand and another one that had a balloon inside it that could be inflated or deflated with a bicycle pump. I assume those two techniques were developed by the manufacturers in order to tune the speaker cabinets somehow.
walpurgis
06-07-2012, 09:01
If you want the simplest approach, maybe its worth experimenting with some well chosen, cheap drive units of the right size for your cabinets and of similar sensitivity to each other. Don't be put off by low cost speaker drivers as they can sound very good, while some expensive stuff can be a let down. There are always drive units to be found on eBay along with possibly suitable crossovers. Try using drivers of similar impedance as it makes the choice of crossover easier. Impedance ratings are generally pretty nominal mind you, as impedance curves can swing all over the place, all you tend to get is a reasonable match, hopefully around the crossover point. With say a 6.5 inch bass driver, you'd be likely looking at using a crossover frequency of around 3 to 4 kHz. If, as often happens, the tweeter output is higher than the bass output, you can use an 'L' pad (variable resistor/potentiometer) to adjust this as a simple option. In a medium sized, sealed cabinet, avoid using bass units with huge magnets as you could end up with an overdamped system. People have been making speakers for years just using their ears and a bit of careful judgement and achieving surprisingly decent results (and sometimes horrible ones). A bit of experience helps of course, but you get that by having a bash.
Gordon Steadman
20-07-2012, 20:44
Interesting thread. I am, by training, a guitar maker. I have no technical knowledge of electronics / connected acoustics whatsoever.
However, all my life I have played about with hi fi in order to make my system sound EXACTLY like the guitars I make. This sounds simple but guitars are actually very subtle instruments with many overtones and resonances that get lost in many systems I have heard. The basic sound of the guitar seems OK but the bits that make each one individual get lost.
So, my main system is and is likely to remain, Stereo 20 and ESL57's. However, it hasn't stopped me experimenting. In my case, this means doing something that feels right, putting it all together and see what it sounds like. Very scientific.
I was given a pair of Ditton 15's. Sadly, the cabinets were beyond hope. However, all the drive units and crossovers were fine so they were obvious candidates for a bit of playing.
My wife's system was that of a typical head banger, HUGE Pioneer amp with JBL speakers that used to skin the ears of the cats. I always like the base response but that tweeter had to go. Insert one Celestion tweeter. Unbelievable difference. I can even listen to classical music on them now and enjoy it.
We are decorating bedrooms at the moment and there is all sorts of stuff that sleeps with us in the second room. I have an 8 inch celestion, a Kef T27 and the celestion crossover. Bung 'em in a box (its a sort of quarter wave/horny sort of thing!@! and see what happens.
Driven by any of the amps you see here, Quad34, Quad50ds, Musical Fidelity A100, even the Pioneer, they sound fantastic. The normally quite sharp T27 seems much smoother in this combination, maybe a sensitivity thing. No idea at all if they measure well but they move the feet. Guitar sounds like guitar and they make music. I don't care how they measure. I have no intention of comparing them directly with the Quads because they serve a different purpose.
I suppose all this means that I am suggesting you can achieve a result purely by giving it a go and seeing how it turns out.
Half the fun I reckon.
I'm sure the objectivists will be horrified but I don't much care about that either:ner:
Measure your internal volume. You will need to match your new woofer to this volume (VB)
Search drivers on the web, many sellers will offer suggested volume and alignments for their drivers. e.g.
http://www.solen.ca/pub/index.php?content&s1=9&page=13&nocol&nodiapo
Once you have a desired driver search the web again for a kit based on this woofer.
You more then likely will have to massage your front baffle. I have done this with excellent results.
best of luck
walpurgis
21-07-2012, 14:13
Gordon,
I suspect that although your combination of the T27 and Ditton 15 bass unit and crossover may sound OK, you have a speaker that is not operating ideally. The T27 is an 8ohm unit, whilst the Celestion bass driver from the Ditton 15 is usually a 6ohm unit and so is the HF1300. This means that your crossover behaviour will not be right, as it was designed for two 6ohm loads. You might do better to search out crossover designs online that can match differing impedances. Also, the Ditton 15 crossover point is around 2.8kHz with 6db/oct and 12db/oct curves if I recall, but the T27 should crossover above 3.5kHz on a steep 18db/oct curve to avoid the diaphragm fundamental resonance and also to avoid power handling problems, i.e., you could blow it!
I like your pairing of the Leak Stereo 20 with the Quad ESL57 by the way. That's a combination I used to use and very nice it was too.
Geoff.
Gordon Steadman
21-07-2012, 14:46
Gordon,
I suspect that although your combination of the T27 and Ditton 15 bass unit and crossover may sound OK, you have a speaker that is not operating ideally. The T27 is an 8ohm unit, whilst the Celestion bass driver from the Ditton 15 is usually a 6ohm unit and so is the HF1300. This means that your crossover behaviour will not be right, as it was designed for two 6ohm loads. You might do better to search out crossover designs online that can match differing impedances. Also, the Ditton 15 crossover point is around 2.8kHz with 6db/oct and 12db/oct curves if I recall, but the T27 should crossover above 3.5kHz on a steep 18db/oct curve to avoid the diaphragm fundamental resonance and also to avoid power handling problems, i.e., you could blow it!
I like your pairing of the Leak Stereo 20 with the Quad ESL57 by the way. That's a combination I used to use and very nice it was too.
Geoff.
Its a study / bedroom system Geoff, so its very unlikely to get much power through it, unless of course the wife decides to use her Pioneer and dance a bit. Hopefully, she will use the JBLs, I think the HF1300s will take the power As I said, the bodge ups sound fine for what they are so I will probably leave well alone apart from making the speakers look nice - much later, I have bedrooms to decorate:(
The theory is beyond me but whilst the figures might not add up, the variables of mismatches seem to make a good sound so I'm happy enough with it.
The Leak and Quads have been my main system for 25 years and I have never found anything that provides more sheer musical enjoyment. The 50ds with the ESLs sounds pretty good too but lacks the 3Dness of the valves. Vinyl is still tops for me but the CDs sound fine through it too, even the Mac doesn't sound as bad as expected as, whilst it shines with a really good source, it seems to be sympathetic to lesser ones as well.
Mr Kipling
22-07-2012, 14:04
Some interesting stuff here. Bit surprised by comments regarding MDF.
http://tnt-audio.com/clinica/diyloudspeakers.html
Kind Regards
Stephen
Gordon,
I suspect that although your combination of the T27 and Ditton 15 bass unit and crossover may sound OK, you have a speaker that is not operating ideally. The T27 is an 8ohm unit, whilst the Celestion bass driver from the Ditton 15 is usually a 6ohm unit and so is the HF1300. This means that your crossover behaviour will not be right, as it was designed for two 6ohm loads. You might do better to search out crossover designs online that can match differing impedances. Also, the Ditton 15 crossover point is around 2.8kHz with 6db/oct and 12db/oct curves if I recall, but the T27 should crossover above 3.5kHz on a steep 18db/oct curve to avoid the diaphragm fundamental resonance and also to avoid power handling problems, i.e., you could blow it!
I like your pairing of the Leak Stereo 20 with the Quad ESL57 by the way. That's a combination I used to use and very nice it was too.
Geoff.
Should he bother to do this if it sounds good to him?
Or is it a geeky ' Even though it sounds fine, it is playing on my mind that I've not done things properly' kind of problem?
For my speakers, I would like to get things working I guess roughly and then slowly tweek over a long time to get it better and better.
Hence I personally just want to get a ball park area of drivers that COULD be made to be perfect when tweeked.
It's all too tempting for people to gush into the details on how to get something perfect from the outset but many people will want to simply try things out first and tweek and design later, much like upgrading hifi.
Problem is I'm starting to look at all the really expensive drivers now like Accuton..
Have kind of realised that I can make my own front baffle as long as the magnet will fit the hole already there - don't have to worry about exact mountings on the driver.
walpurgis
30-07-2012, 20:14
Well...
I think people should do what they want!
I've probably built fifty pairs of speakers over the years, generally with good results, but also the occasional disgrace.
Keep it simple, especially if you haven't attempted a build before.
Very simple choices are maybe starting with a full range driver. There are some decent 5" and 6.5" units about and they can sound very good. Just use a ported cabinet of 15 or 20 litres and a 2" port tube about 3" to 4" long and a bit of wall damping and away you go (a bit of experimenting with different port lengths may benefit).
Alternatively, try a similar sized bass/mid driver with a tweeter, just rolling the tweeter in with an audio quality capacitor at around 5kHz (assuming an 8ohm tweeter experiment with capacitors around 4mFd to 6mFd). An 'L' pad (potentiometer) can be used to adjust treble levels. Your ears will tell you a lot!
These are simple basic options. The sky is the limit though when it comes to complexity, designing multi driver systems with matching complicated steep crossovers, exotic cabinet forms, etc. etc.
I'd say regard it as a bit of fun and don't take the first few attempts too seriously.
Think this little project is a little limited... The size and shape of the driver to fit is a bit limiting.
Might have to resort to finding some original drivers afterall (from a pair of Celestion 100s).
Problem is that the basket needs to be square or curved with corners (if you see what I mean, similar to a Kef B110 shape but more so) like the original. Can't fit a round basketed driver on there without making a front baffle that is too thick/far forward of the cabinet making it look daft.
Can anyone think of perhaps discontinued 6" mid-bass drivers with flat basket sides? The higher-end the better!
The Grand Wazoo
09-08-2012, 07:25
Any chance you could get set to with a router? You could make them more accommodating with a little work.
Due to the mess at my place, this project kind of took a back seat....
Routing the cabinets would probably make them crumble away due to the nature of the resin composite they're make of. Has ground up white granite in it to create the flecks..
Anyway! I happened apon some Celestion 100s which where released a couple of years later and used up the remaining Kingston drivers.. so bought them!
Not even tried them yet but have confirmed to myself that they are the same drivers at least (the SL6s didn't use the same drivers as the 600/700s - different basket - so it was a slight risk with the 100s).
So.... now, I need to make front baffles that mount the tweeter and cover up the mid baskets.
But more importantly, I need crossovers.. I can start with those from the square-boxed 100s but no doubt they will be different. Might see if I can get a schematic from Celestion.
Well, unforatunately/fortunately this is becoming much less of a DIY project than a re-construction project..
I emailed Celestion and via some internal email forwarding, Kef replied with the original schematic and a photo of the cross-over.. said they had some in stock! At £106 each + VAT I think I'll pass.
So, it's now a case of copying the cross-over with some fancy components, making an external box for them and sorting out a decent front baffle to moung the tweeter and hide the mid-driver.
What's the best way to choose components? Do the expensive coils and caps really make a big difference? A matter of tweaking?
Reid Malenfant
05-12-2012, 19:57
If you are going to change the inductors make sure you fit new ones which have the same DC resistance as the ones already fitted ;)
The crossover was designed to obviously work with the drivers & enclosure, if you use inductors of different DC resistance you'll find the bass response altering. If you fit one with a significantly lower resistance feeding the bass unit, you'll get less deep bass as the driver Q will be lowered.
While the same can't be said for the other inductors, the crossover response will be altered if they are not to spec.
Inductors don't tend to age, they are just windings of insulated copper wire, with or without an iron or ferrite based core. There is nothing to age in reality.
I'd be looking at the capacitors if I was you :)
Thanks for the advice about their DC resistance.
When I say copy the crossovers with fancy components, I mean copy from the schematic that I was sent as a .pdf - I don't actually have the originals at all.. i.e. I have no crossovers at all for the Kingstons.
I'll have a look inside the Celestion 100s which are the driver doners as they may well use the same make/brand and even share some values of component with the Kingston. I was also sent a photo of one they had in stock so I can visually check if they are the same or not..
Powered by vBulletin® Version 4.2.3 Copyright © 2024 vBulletin Solutions, Inc. All rights reserved.