Inductors , 44 crossover , new Cabinet , midrange driver.
Ferrite inductors are recommended for the Woofer Filter section of a crossover ,
NOT for the "bass frequencies" , but for the reason that ferrite cored inductors are much lower price than the very large air-core inductors
as the ferrites are much smaller size for the low DC resistance that is optimum in woofer filter inductors than air-cores' very large size for there.
The Ferrite cores will saturate when the music is loud , and the bass sound will become somewhat mushy ,
however if one has heard only higher DC resistance Air-cores , such as in the Celestion 66 ,
then one does not know just how good the sound can be with much lower DC resistance Air-cores.
I have read this entire thread , and given the amount of work you are going to do building new cabinets , etc ...
it really will be beneficial to buy , or wind your own , low DCR Aircore inductors.
I recommend that you do not buy Air-core inductors for the woofer filter from Falcon ,
because they do not manufacture any that are low enough in DCR to get the best from the Celestion woofers.
For the lowest priced low DCR aircores I recommend you buy from Solen in Canada , where it seems you are ...
Better aircores are manufactured by Jantzen , but the USA sellers do not have the lowest DCR sizes ,
and not one of the values that you will need.
Thus the Jantzens would have to be bought from Europe.
Better again can be wound for you by a specialist company in the USA ,
however the price will not be less than USD 50 each for the 3.5mH , and about USD 70 if you buy the even better much lower DCR they make.
The 2.2mH will not be as as expensive.
For the 2.2mH in the midrange filter low DCR is NOT what is needed ,
and Celestion got close to optimum there with their 1.3 ohm DCR 2.2mH ,
so to buy or wind for that one will not be as expensive.
For the smaller inductor in the mids' filter , the lower the DCR the better , and again in Air-core.
For the inductor in the tweeter filter very low DCR will not help , but will allow an audible resonance ,
{ same as with the Parallel connected 2.2mH in the mids' filter }.
About 0.5ohm DCR or very little less will be best for the tweeter filter's inductor.
As you are starting with , { presumably ? } , none of the original Celestion inductors ,
you do not have to copy the values in Celestion's crossover if you are using a SEAS 19TFF1.
Instead you can parallel the SEAS with a suitable resistor , and it does need that to enable a smooth roll-off with the simple crossover ,
and buy standard inductor and capacitor values , specifically:
0.15mH in as close to 0.5 ohm as possible , 3.3uF and 10uF , to cross at 5kHz.
If you want to do that I will calculate the value for the Parallel resistor for you.
SEAS 19TFF1 can be crossed at 3.5kHz , but at 4kHz or higher is simpler to do ,
because less parts needed in the crossover for smooth roll-off then.
Both Lucas , and a serious listener to some other 3-way loudspeakers , have found the 44 cone midrange to sound quite good to 5kHz ,
so you may as well copy the 44 crossover for the midrange filter , and for the woofer filter.
For your new cabinet:
(1)- design for the mid-cone to be at ear height for when you are seated listening , and for the tweeter to be higher.
(2)- design for no extended cabinet edges to protrude past the front of the baffle ,
and better will be to round off the front edges around the baffle.
(3)- the volume of the mid-cone enclosure is NOT critical , BUT it must be no less than the original.
It can be a bit larger , and then ideally rectangular and not a cube shape.
It will need to be filled with sound absorbing material , but not stuffed tight.
(4) - for a 44 make the woofer cabinet slightly larger Internal volume than Celestion's original ,
and ignore the small increase of the absence of the internal midrange enclosure.
{ That is design at least 5% larger , and as much as 10% may be better for a more even and extended response into the lowest bass frequencies.}
You can install the woofer at the same height in the baffle , but as you seem to be intending to put a separate mids' enclosure on top ,
then if you want the woofer up higher , divide the front baffle internal height in this ratio:
0.382 : 0.618
and centre the woofer at that height in the ratio.
(5)- design the cabinet's Internal Depth to be close to 80% of its Internal Width.
The above is for a 44 design , but if you decide to make a 66 or 25 ABR cabinet , then copy those for Internal volume and the ABR position ,
but use the above ratio for the woofer position.
Crossover capacitors and resistors:
If you want to use new Alcaps that is your business to decide ,
but until you have heard a good quality polypropylene cap with the optimum resistor value in Series with it you do not know how good that is ,
and it is much better than poly caps with no resistors when used in Parallel arms of the filter , such as for the woofer section and the upper midrange section of the 44 and 66 crossovers.
The Series cap in the mids' filter and both caps in the tweeter filter will not need resistors if you use low DCR inductors in the woofer filter for higher output of bass and lower midrange
so as to equal the outputs of the mid-cone and SEAS tweeter.
Solen polypropylene caps are OK to use in the woofer filter , but not audibly good enough for the mids and treble.
Better can be bought for not much higher price from another Canadian , and USA sellers.
I have posted about alternate caps , and much about the crossovers of the 44 and 66 in two other threads:
(1)- Crossover nightmare!!!!!!!
started by: lorienblack , who I see has posted here , 'ullo Lorien !
(2)- celestion 66 needs Mid-range
started by: Mr_White , and start on Page 21 of that long thread for the discussion about the resistors , etc ...
I will be posting in the 66 thread about suitable low DCR inductors for the Celestion woofer as soon as I have time available to do that comprehensively ,
as some 66 owners there are interested.
If you want to query any of what I have posted above , then do , and I will reply when I have time available ,
however I do not have time to repeat everything that I have posted in the two other threads I listed above.
*** *** ***
'ullo Lucas ,
if you didn't receive a diyaudio message notifying , then I have replied to your post in the 66 thread , and again today to correct a mistake in the initial post.
Ferrite inductors are recommended for the Woofer Filter section of a crossover ,
NOT for the "bass frequencies" , but for the reason that ferrite cored inductors are much lower price than the very large air-core inductors
as the ferrites are much smaller size for the low DC resistance that is optimum in woofer filter inductors than air-cores' very large size for there.
The Ferrite cores will saturate when the music is loud , and the bass sound will become somewhat mushy ,
however if one has heard only higher DC resistance Air-cores , such as in the Celestion 66 ,
then one does not know just how good the sound can be with much lower DC resistance Air-cores.
I have read this entire thread , and given the amount of work you are going to do building new cabinets , etc ...
it really will be beneficial to buy , or wind your own , low DCR Aircore inductors.
I recommend that you do not buy Air-core inductors for the woofer filter from Falcon ,
because they do not manufacture any that are low enough in DCR to get the best from the Celestion woofers.
For the lowest priced low DCR aircores I recommend you buy from Solen in Canada , where it seems you are ...
Better aircores are manufactured by Jantzen , but the USA sellers do not have the lowest DCR sizes ,
and not one of the values that you will need.
Thus the Jantzens would have to be bought from Europe.
Better again can be wound for you by a specialist company in the USA ,
however the price will not be less than USD 50 each for the 3.5mH , and about USD 70 if you buy the even better much lower DCR they make.
The 2.2mH will not be as as expensive.
For the 2.2mH in the midrange filter low DCR is NOT what is needed ,
and Celestion got close to optimum there with their 1.3 ohm DCR 2.2mH ,
so to buy or wind for that one will not be as expensive.
For the smaller inductor in the mids' filter , the lower the DCR the better , and again in Air-core.
For the inductor in the tweeter filter very low DCR will not help , but will allow an audible resonance ,
{ same as with the Parallel connected 2.2mH in the mids' filter }.
About 0.5ohm DCR or very little less will be best for the tweeter filter's inductor.
As you are starting with , { presumably ? } , none of the original Celestion inductors ,
you do not have to copy the values in Celestion's crossover if you are using a SEAS 19TFF1.
Instead you can parallel the SEAS with a suitable resistor , and it does need that to enable a smooth roll-off with the simple crossover ,
and buy standard inductor and capacitor values , specifically:
0.15mH in as close to 0.5 ohm as possible , 3.3uF and 10uF , to cross at 5kHz.
If you want to do that I will calculate the value for the Parallel resistor for you.
SEAS 19TFF1 can be crossed at 3.5kHz , but at 4kHz or higher is simpler to do ,
because less parts needed in the crossover for smooth roll-off then.
Both Lucas , and a serious listener to some other 3-way loudspeakers , have found the 44 cone midrange to sound quite good to 5kHz ,
so you may as well copy the 44 crossover for the midrange filter , and for the woofer filter.
For your new cabinet:
(1)- design for the mid-cone to be at ear height for when you are seated listening , and for the tweeter to be higher.
(2)- design for no extended cabinet edges to protrude past the front of the baffle ,
and better will be to round off the front edges around the baffle.
(3)- the volume of the mid-cone enclosure is NOT critical , BUT it must be no less than the original.
It can be a bit larger , and then ideally rectangular and not a cube shape.
It will need to be filled with sound absorbing material , but not stuffed tight.
(4) - for a 44 make the woofer cabinet slightly larger Internal volume than Celestion's original ,
and ignore the small increase of the absence of the internal midrange enclosure.
{ That is design at least 5% larger , and as much as 10% may be better for a more even and extended response into the lowest bass frequencies.}
You can install the woofer at the same height in the baffle , but as you seem to be intending to put a separate mids' enclosure on top ,
then if you want the woofer up higher , divide the front baffle internal height in this ratio:
0.382 : 0.618
and centre the woofer at that height in the ratio.
(5)- design the cabinet's Internal Depth to be close to 80% of its Internal Width.
The above is for a 44 design , but if you decide to make a 66 or 25 ABR cabinet , then copy those for Internal volume and the ABR position ,
but use the above ratio for the woofer position.
Crossover capacitors and resistors:
If you want to use new Alcaps that is your business to decide ,
but until you have heard a good quality polypropylene cap with the optimum resistor value in Series with it you do not know how good that is ,
and it is much better than poly caps with no resistors when used in Parallel arms of the filter , such as for the woofer section and the upper midrange section of the 44 and 66 crossovers.
The Series cap in the mids' filter and both caps in the tweeter filter will not need resistors if you use low DCR inductors in the woofer filter for higher output of bass and lower midrange
so as to equal the outputs of the mid-cone and SEAS tweeter.
Solen polypropylene caps are OK to use in the woofer filter , but not audibly good enough for the mids and treble.
Better can be bought for not much higher price from another Canadian , and USA sellers.
I have posted about alternate caps , and much about the crossovers of the 44 and 66 in two other threads:
(1)- Crossover nightmare!!!!!!!
started by: lorienblack , who I see has posted here , 'ullo Lorien !
(2)- celestion 66 needs Mid-range
started by: Mr_White , and start on Page 21 of that long thread for the discussion about the resistors , etc ...
I will be posting in the 66 thread about suitable low DCR inductors for the Celestion woofer as soon as I have time available to do that comprehensively ,
as some 66 owners there are interested.
If you want to query any of what I have posted above , then do , and I will reply when I have time available ,
however I do not have time to repeat everything that I have posted in the two other threads I listed above.
*** *** ***
'ullo Lucas ,
if you didn't receive a diyaudio message notifying , then I have replied to your post in the 66 thread , and again today to correct a mistake in the initial post.
Last edited:
Wow what a wealth of information, thank you Alan!
A lot of what you have mentioned with regards to the cabinets is already happening, such as the midrange enclosure being rectangular and also increasing the volume of the bass cabinet.
I will definitely adhere to the midrange being at ear height and the tweeter being higher, with my experience with my big IMFs and Kefs, I had already figured this was most suitable.
As for the crossover, thank you for your advice here. A lot of what you spoke of I had no idea of, but now I will be more careful in my selection of parts. With the Seas tweeter, that sounds like a fairly easy thing to do, I'll give that a shot, thank you.
As for the capacitors, I may go for poly. I know they are better, another option is I could couple charge the circuit with electrolytics. I have this done in my IMF Super Compacts and it made a very nice improvement, but of course I guess it all depends on what is simpler and more cost-effective.
In the end, I don't want to spend too much money on this, but I want to spend the money in the right areas.
Thanks again for your help with this Alan!
A lot of what you have mentioned with regards to the cabinets is already happening, such as the midrange enclosure being rectangular and also increasing the volume of the bass cabinet.
I will definitely adhere to the midrange being at ear height and the tweeter being higher, with my experience with my big IMFs and Kefs, I had already figured this was most suitable.
As for the crossover, thank you for your advice here. A lot of what you spoke of I had no idea of, but now I will be more careful in my selection of parts. With the Seas tweeter, that sounds like a fairly easy thing to do, I'll give that a shot, thank you.
As for the capacitors, I may go for poly. I know they are better, another option is I could couple charge the circuit with electrolytics. I have this done in my IMF Super Compacts and it made a very nice improvement, but of course I guess it all depends on what is simpler and more cost-effective.
In the end, I don't want to spend too much money on this, but I want to spend the money in the right areas.
Thanks again for your help with this Alan!
some follow-up ...
The tweeter and mid-cone can be in the same enclosure if the tweeter is higher ,
because that will cause both to have same path length to seated ears for Time Alignment ,
however you can seal between them with an internal partition inside the enclosure to reduce vibration from the mids impacting on the tweeter.
The benefit may be small , but it may be audible.
Charge-coupling electrolytic caps can cause better sound than otherwise ,
however it does not remove the Dielectric Absorption of the signal from the electros.
Dielectric Absorption causes smearing of the sound ,
however I realize that some listeners like this effect as it can seem as if the music is a bit richer in Tone ,
and KEF know that , which is why they persist with electros even in their expensive Reference Series models ,
but remember that most KEF buyers are not true Audiophiles , but are more often listeners who own quite ordinary , even if high powered , amplifiers ,
and sometimes only average quality CD players/Turntables , etc ...
I have to go now , I'll check back for more here when I have time available.
The tweeter and mid-cone can be in the same enclosure if the tweeter is higher ,
because that will cause both to have same path length to seated ears for Time Alignment ,
however you can seal between them with an internal partition inside the enclosure to reduce vibration from the mids impacting on the tweeter.
The benefit may be small , but it may be audible.
Charge-coupling electrolytic caps can cause better sound than otherwise ,
however it does not remove the Dielectric Absorption of the signal from the electros.
Dielectric Absorption causes smearing of the sound ,
however I realize that some listeners like this effect as it can seem as if the music is a bit richer in Tone ,
and KEF know that , which is why they persist with electros even in their expensive Reference Series models ,
but remember that most KEF buyers are not true Audiophiles , but are more often listeners who own quite ordinary , even if high powered , amplifiers ,
and sometimes only average quality CD players/Turntables , etc ...
I have to go now , I'll check back for more here when I have time available.
Sounds good, that is the plan actually, to have the tweeter and mid in one enclosure but separated internally.
Interesting you mention the Kef reference models, as my main speakers currently are Kef 107s. And your description is spot on, I don't really consider myself an audiophile, more of a music lover. The Kefs are very pleasing to listen to and this is what I like in a speaker. I also own IMF RSPM Mk IVs which I would consider more "audiophile" but the Kefs are simply more enjoyable. And I am using Carver equipment to power the Kefs, so again, spot on there.
I guess that's why I'm not overly picky about the capacitors. I do understand and realize the difference, but I also do enjoy my other speakers which are littered with electrolytic caps. We'll see, I think I definitely want to spend the money on the inductors, but as for the capacitors we will see. My budget is not too huge, maybe around $1000 total, and that will have to include the purchase of measuring equipment. The midrange cost me $80, so I am left with around $700 for the rest including cabinets, materials and such.
Okay another sketch is coming along. I'll post it up when it's ready.
Question, for damping the cabinets, any recommendations there? Is polyfill good for the bass cabinet? I believe Solen sells some good stuff. I just know there is open and closed cell and am not sure what would be more suitable for which part of the enclosure.
Question, for damping the cabinets, any recommendations there? Is polyfill good for the bass cabinet? I believe Solen sells some good stuff. I just know there is open and closed cell and am not sure what would be more suitable for which part of the enclosure.
Jeffrey, why not time delay the speakers, so the faster moving treble has further to travel etc. Offsetting driver to ear distance, where the bass driver is nearer and the tweeter further away is common these days in high end 3 ways. I don't know how important it is, but I will be doing that as it's' so easy to implement by copying the principle if not the execution in high end designs like the B+W 801, for example.
http://www.ebay.co.uk/itm/B-W-801-Series-3-Matrix-Floor-Standing-HiFi-Speaker-Rare-Concept-B-W-Kef-N-R-/330894487070?pt=UK_AudioVideoElectronics_HomeAudioHiFi_HiFiSpeakers&hash=item4d0ad92e1e&nma=true&si=vWOmVJ54rp9z8XSHnfR9t8G9%252F1w%253D&orig_cvip=true&rt=nc&_trksid=p2047675.l2557
http://www.ebay.co.uk/itm/B-W-801-Series-3-Matrix-Floor-Standing-HiFi-Speaker-Rare-Concept-B-W-Kef-N-R-/330894487070?pt=UK_AudioVideoElectronics_HomeAudioHiFi_HiFiSpeakers&hash=item4d0ad92e1e&nma=true&si=vWOmVJ54rp9z8XSHnfR9t8G9%252F1w%253D&orig_cvip=true&rt=nc&_trksid=p2047675.l2557
Last edited:
It's a good suggestion but I don't think I will be taking that on this go around. I think I would rather keep it the way it is, and then perhaps later down the road it wouldn't be difficult to simply remake the top enclosures. Nice thing about this design is its flexible so if I want to change something later on I can. Plus, doing it this way will give me an idea of the benefit and effect of doing a vertical alignment of the drivers. Some manufacturers do it but also many don't, so it's hard to say if it really is worth it.
Do you think spikes are the ideal thing to isolate the two cabinets? I would have though that the purpose of isolation was to prevent bass vibrations moving up to rattle the upper cabinet. Would sorbothane or something similar be a better option? I realise that there are 2 schools of thought on this, but spikes would seem almost to be like a transducer to transfer the energy, instead of suppress it!
My research into defraction is leading me to completely redesign the mid and treble units, in a manner not unlike these:
BLUEROOM MINIPOD
Maybe I'm going to end up using new paper drivers too :/ Maybe my 44 rebuild plans are leading me to design a completely new pair of speakers.
Oh the wife WILL be pleased, hehe
BLUEROOM MINIPOD
Maybe I'm going to end up using new paper drivers too :/ Maybe my 44 rebuild plans are leading me to design a completely new pair of speakers.
Oh the wife WILL be pleased, hehe
Last edited:
Hmm that's a fair assessment you make there, the point of the spikes was to raise the midrange enclosure up a bit to give it the separated look. So more cosmetics than practical application. What could be done is to put some rubber feet where the spikes would sit in.
on several points in others' Posts since my last
Lucas ,
there is NO "faster moving treble" !
All sound frequencies travel at exactly the same speed.
Wavelength changes and Frequency changes ,
but both by the same Ratio such that when one increases the other decreases as follows:
Wavelength x Frequency = Speed
I think what you are refering to is path length difference and the subsequent Time Delay.
With Jeffrey's 2-box design the top box can be moved back to the point where the distance from the centre of the mid-cone = the distance from the centre of the woofer cone to the seated listener's ears.
There will then be equal arrival times of all sound from bass up through all the midrange ,
and may be through all the treble IF the height of the tweeter is such that the distance from its dome centre = distance from centre of mid-cone from seated listener's ears.
This will work IF all the filters in the crossover
-{ including the effects of the non-linear Impedance and not quite flat response from all drivers in the octave around each nominal crossover point }-
have complementary Phase differences.
The above condition is difficult to achieve perfectly , and especially if accurate measurements are not made ,
however we can get close enough if apply some practical estimates.
One compromise is for the centre of woofer cone to be not further than one eighth of the crossover frequency wavelength further distance than the centre of the mid-cone.
That is achievable , because at 500 Hz , one eighth is somewhere between 3" <--> 4".
At 5kHz it is one tenth that length , so get the equal path lengths for mid-cone and tweeter , and get near enough for the woofer ,
then the sound is usually acceptable.
For the top mounted box ,
if the spikes are Flat-end under the bottom of that box , and spike point down into the lower box , then more of the vibration is directed down than up.
OK , but there will be greater amplitude vibration from the bass box , however there has to be a compromise if one is building within a budget.
To further reduce upward vibration , postion the spikes so that they direct into the walls of the bottom box and not into the vibrating top area of it.
Use floor spikes also , pointed down and positioned under the walls of the bottom box.
Ideally use 3 spikes.
2 in corner positions and 1 in a centre position.
The centre spike can be under centre of the baffle wall or centre of the back wall ,
and the other 2 under the opposite.
The purpose for the spikes is to collect energy from the cabinet's vertical walls ,
because it cannot be usefully collected from vibrating horizontal panels ,
and then direct it all down into solid Ground - in most cases into a mass loaded floor.
If a floor is bouncy , then the spikes' remedy does not work well , and neither does much else !
I recommend that you do not use Rubber feet , because the box will vibrate - bounce - on those.
The top box does not need to be a complete box.
Better for Diffraction reduction is to use a wide radius curved baffle to form a Semi-circle ,
and attach a flat bottom panel to it.
Behind the centre of the semi-circle attach a sealed enclosure to the flat bottom.
This can be in rectangular shape , or it can be a cylinder on its side.
For a cylinder , or a small bucket shape , choose one that has an inside diameter at least 2" , and preferable 4" , larger than the cut-out hole diameter for the mid-cone.
Length of cylinder must be sufficient for the total Volume to be not less than that of Celestion's mid-enclosure.
There are large thick-walled cardboard cylinders used for packaging various substances.
Thickness of cardboard would need to be at least 1/2" for the mids' enclosure.
If a cylinder is found with a minimum 10" , 12" better , diameter it can be used to form the curved baffle.
A curved baffle can also be made using layers of flexible plywood ,
or glue such type of ply to both sides of the cardboard curve.
If the bottom panel is cut to a shape of quarter circle curve of 3" radius + 6" flat + another 3" radius ,
then the cardboard can be forced in shape around its edge for a rounded edges flat front baffle ,
and plywood forced around that and tied in place whilst the adhesive sets.
Search for " bendy ply " or " bendy-ply ".
That is used in furniture manufacture for some types of chairs.
One spike can be attached under the baffle centre , and one under each end of its curved section.
For the mids' enclosure one can use 2 small buckets of suitable volume.
Roughen , then cover the internal surface of one and the external surface of the other with glue ,
and cut 1/2 thick packaging foam or paper-mache to fit between the 2 buckets.
Slightly thicker stuffing will be needed between the flat bottom surfaces ,
and use a more compressable foam there so that there is no air gap or unglued area of bucket base free to vibrate.
A stiffening circle of wood can be glued to the outer base of the outer bucket to reduce amplitude of vibration which could otherwise be reflected back to the mid-cone rear ,
unless a fortuitiously very absorptive substance at all frequencies is used inside the inner bucket.
The tweeter does not need a sealed section behind it , but needs only to be rebated into the flattened front of the baffle ,
or have a piece of absorptive Felt placed around it , including down as far the the top of the mid-cone.
If the top-box is pushed back , then thicker absorptive felt will be very beneficial from its front extending to the top edge of the bottom box.
Even if not pushed back , the gap between the boxes as result of the spikes should be filled with a sound absorptive substance.
If you are going to put the woofer near to the top of the baffle , as shown in your diagrams Jeffrey ,
then allow at least 2" internal space between the edge of driver cut-out and underside of top panel ,
because there will need to be sufficient space for sufficient absorptive material otherwise there will be lower midrange colouration reflected back into the cone to interfere with the sound.
Solen's web-site will not open at the time I am typing this , thus I do not know what acoustic damping materials they sell ,
but if it is a plastic foam , then it needs to be Open-cell to absorb sound.
Closed-cell foam will only reduce vibration transmission , and basically that type is packaging foam , or for use between 2 rigid outer skins.
Jeffrey ,
what is the "measuring equiptment" you are planning to buy ?
Carver amplifiers are quite good , and some models were very good ,
and unless there are some models that I do not know about , then all are above average in sound quality ,
and are good enough for differences between poly caps and electro caps in crossovers to be distinguished.
That amplifiers have electro caps in various sections of their circuits does not prevent one's ability to hear differences between cap types in crossovers ,
because the Back-EMF of the drivers in the loudspeaker is also Signal through the crossover caps back into the Feedback loop of the amplifier ,
and thereby all non-linear aspects of crossover capacitors , especially Dielectric Absorption , is pushed into the amplifier and there-in changes the Signal through the amplifier.
Similar occurs with any non-linear effects from the Dielectric -{ that is the Insulation }- of Cables ,
thus one reason why different types of cables cause different sounds with different amplifiers.
Cables with parallel conductors , regardless of whether straight or twisted or woven , are both capacitors and inductors ,
{and are RF receiving aerials sending signal into amplifier feedback loops }.
Hey , why am I typing all this instead of doing my own DIY !
Lucas ,
there is NO "faster moving treble" !
All sound frequencies travel at exactly the same speed.
Wavelength changes and Frequency changes ,
but both by the same Ratio such that when one increases the other decreases as follows:
Wavelength x Frequency = Speed
I think what you are refering to is path length difference and the subsequent Time Delay.
With Jeffrey's 2-box design the top box can be moved back to the point where the distance from the centre of the mid-cone = the distance from the centre of the woofer cone to the seated listener's ears.
There will then be equal arrival times of all sound from bass up through all the midrange ,
and may be through all the treble IF the height of the tweeter is such that the distance from its dome centre = distance from centre of mid-cone from seated listener's ears.
This will work IF all the filters in the crossover
-{ including the effects of the non-linear Impedance and not quite flat response from all drivers in the octave around each nominal crossover point }-
have complementary Phase differences.
The above condition is difficult to achieve perfectly , and especially if accurate measurements are not made ,
however we can get close enough if apply some practical estimates.
One compromise is for the centre of woofer cone to be not further than one eighth of the crossover frequency wavelength further distance than the centre of the mid-cone.
That is achievable , because at 500 Hz , one eighth is somewhere between 3" <--> 4".
At 5kHz it is one tenth that length , so get the equal path lengths for mid-cone and tweeter , and get near enough for the woofer ,
then the sound is usually acceptable.
For the top mounted box ,
if the spikes are Flat-end under the bottom of that box , and spike point down into the lower box , then more of the vibration is directed down than up.
OK , but there will be greater amplitude vibration from the bass box , however there has to be a compromise if one is building within a budget.
To further reduce upward vibration , postion the spikes so that they direct into the walls of the bottom box and not into the vibrating top area of it.
Use floor spikes also , pointed down and positioned under the walls of the bottom box.
Ideally use 3 spikes.
2 in corner positions and 1 in a centre position.
The centre spike can be under centre of the baffle wall or centre of the back wall ,
and the other 2 under the opposite.
The purpose for the spikes is to collect energy from the cabinet's vertical walls ,
because it cannot be usefully collected from vibrating horizontal panels ,
and then direct it all down into solid Ground - in most cases into a mass loaded floor.
If a floor is bouncy , then the spikes' remedy does not work well , and neither does much else !
I recommend that you do not use Rubber feet , because the box will vibrate - bounce - on those.
The top box does not need to be a complete box.
Better for Diffraction reduction is to use a wide radius curved baffle to form a Semi-circle ,
and attach a flat bottom panel to it.
Behind the centre of the semi-circle attach a sealed enclosure to the flat bottom.
This can be in rectangular shape , or it can be a cylinder on its side.
For a cylinder , or a small bucket shape , choose one that has an inside diameter at least 2" , and preferable 4" , larger than the cut-out hole diameter for the mid-cone.
Length of cylinder must be sufficient for the total Volume to be not less than that of Celestion's mid-enclosure.
There are large thick-walled cardboard cylinders used for packaging various substances.
Thickness of cardboard would need to be at least 1/2" for the mids' enclosure.
If a cylinder is found with a minimum 10" , 12" better , diameter it can be used to form the curved baffle.
A curved baffle can also be made using layers of flexible plywood ,
or glue such type of ply to both sides of the cardboard curve.
If the bottom panel is cut to a shape of quarter circle curve of 3" radius + 6" flat + another 3" radius ,
then the cardboard can be forced in shape around its edge for a rounded edges flat front baffle ,
and plywood forced around that and tied in place whilst the adhesive sets.
Search for " bendy ply " or " bendy-ply ".
That is used in furniture manufacture for some types of chairs.
One spike can be attached under the baffle centre , and one under each end of its curved section.
For the mids' enclosure one can use 2 small buckets of suitable volume.
Roughen , then cover the internal surface of one and the external surface of the other with glue ,
and cut 1/2 thick packaging foam or paper-mache to fit between the 2 buckets.
Slightly thicker stuffing will be needed between the flat bottom surfaces ,
and use a more compressable foam there so that there is no air gap or unglued area of bucket base free to vibrate.
A stiffening circle of wood can be glued to the outer base of the outer bucket to reduce amplitude of vibration which could otherwise be reflected back to the mid-cone rear ,
unless a fortuitiously very absorptive substance at all frequencies is used inside the inner bucket.
The tweeter does not need a sealed section behind it , but needs only to be rebated into the flattened front of the baffle ,
or have a piece of absorptive Felt placed around it , including down as far the the top of the mid-cone.
If the top-box is pushed back , then thicker absorptive felt will be very beneficial from its front extending to the top edge of the bottom box.
Even if not pushed back , the gap between the boxes as result of the spikes should be filled with a sound absorptive substance.
If you are going to put the woofer near to the top of the baffle , as shown in your diagrams Jeffrey ,
then allow at least 2" internal space between the edge of driver cut-out and underside of top panel ,
because there will need to be sufficient space for sufficient absorptive material otherwise there will be lower midrange colouration reflected back into the cone to interfere with the sound.
Solen's web-site will not open at the time I am typing this , thus I do not know what acoustic damping materials they sell ,
but if it is a plastic foam , then it needs to be Open-cell to absorb sound.
Closed-cell foam will only reduce vibration transmission , and basically that type is packaging foam , or for use between 2 rigid outer skins.
Jeffrey ,
what is the "measuring equiptment" you are planning to buy ?
Carver amplifiers are quite good , and some models were very good ,
and unless there are some models that I do not know about , then all are above average in sound quality ,
and are good enough for differences between poly caps and electro caps in crossovers to be distinguished.
That amplifiers have electro caps in various sections of their circuits does not prevent one's ability to hear differences between cap types in crossovers ,
because the Back-EMF of the drivers in the loudspeaker is also Signal through the crossover caps back into the Feedback loop of the amplifier ,
and thereby all non-linear aspects of crossover capacitors , especially Dielectric Absorption , is pushed into the amplifier and there-in changes the Signal through the amplifier.
Similar occurs with any non-linear effects from the Dielectric -{ that is the Insulation }- of Cables ,
thus one reason why different types of cables cause different sounds with different amplifiers.
Cables with parallel conductors , regardless of whether straight or twisted or woven , are both capacitors and inductors ,
{and are RF receiving aerials sending signal into amplifier feedback loops }.
Hey , why am I typing all this instead of doing my own DIY !
Last edited:
Hey Alan,
Again thank you for all the information.
With regards to the top enclosure - is it possible you could draw a picture? I'm sorry, you've provided some very detailed information, but a picture is worth a thousand words If you don't mind that is.
With regards to the time delay my main problem with that is indeed the crossover. As it is I am not changing a whole lot here, crossovers are not something I am too good with just yet.
As for the measuring equipment it is the EMM-6 and Tascam US-122mkII combo. And some other bits and pieces for measuring impedance.
Another idea - what do you think of doing a d'appolito design? I have the ability to get another pair of 44 mids, allowing me to do a Mid-Tweeter-Mid combination. Would this be difficult to implement besides cabinetry?
Again thank you for all the information.
With regards to the top enclosure - is it possible you could draw a picture? I'm sorry, you've provided some very detailed information, but a picture is worth a thousand words
If you don't mind that is.With regards to the time delay my main problem with that is indeed the crossover. As it is I am not changing a whole lot here, crossovers are not something I am too good with just yet.
As for the measuring equipment it is the EMM-6 and Tascam US-122mkII combo. And some other bits and pieces for measuring impedance.
Another idea - what do you think of doing a d'appolito design? I have the ability to get another pair of 44 mids, allowing me to do a Mid-Tweeter-Mid combination. Would this be difficult to implement besides cabinetry?
Last edited:
replies to the above
Hey Jeffrey ,
my problem is that I do not know how to load drawn diagrams onto forum pages.
I do not have a Scanner to upload any diagrams , and there is no-one here who can show me how to upload even if I could get use of a Scanner.
Yes , I could learn from some-one some-where how to do those things , but I have no spare time available to fit that in ...
unless you want to wait for ages till I reply to your posts ...
Instead , I refer you to the picture you posted in #50 on Page 5.
Imagine a Cylinder standing vertically on top of the woofer cabinet.
Cut the back half off the cylinder.
Draw that and post it here.
The half-cylinder does not have to be a perfect half-circle shape.
It can have another type of curve , and better is for it to have a flattened front
and then the curves around both sides.
The curves should extend to the full 90 degrees side around each side , because that will allow most of the Diffracted sound to disperse in backwards' directions.
If the curves stop at less than 90 degrees , more of the diffracted sound will be directed in forward-sides angles and reflect forwards off walls of room and sooner towards the listener ,
which will cause a greater degree of colouration and image diffusion than occurs when the diffraction is mostly to side-back angles and then much later reflected arrivals at listener's ears.
Where you have the two sides connected to your mids + tweeter half-box in #50 ,
replace the sharp 90 degrees corners with 3 inch or larger Radius curves.
Draw that and post here , and I will comment further if necessary.
D'Appolito mid-range will not work close to the ideal with only two upper mid-range cones.
For D'Appolito to be worthwhile it needs also two woofers/lower midrange , one at top and one at bottom.
Also , the Sensitivities/Efficiencies of two Celestion mid-cones will cause noticeably louder upper mids than will be from a single woofer/lower mids ,
and , to get Time Alignment you will have to recess the tweeter into behind the baffle to align it to the plane of two separate mid-cones ,
and that will raise another diffraction problem to have to be addressed.
Better is to keep it as you planned with a single mid-cone ,
however do buy the two additional mid-cones whilst they are available ,
because the old Celestion mid-cones are not as robust as the old woofers ,
so you should obtain spares for the mids now.
After you have assembled the loudspeakers you can try all four mid-cones and pick the two best condition by ear ,
then keep the other two as spares.
For the EMM-6 microphone , it is worth buying the Dayton EMM-6 Premium+ calibrated version , here:
Cross·Spectrum - Calibrated Dayton Audio EMM-6 Microphones for Sale
as that will be more useful than an uncalibrated and thus untested sample.
If that mic will work into the Tascam , then I suppose the Tascam is a good buy ,
however I do not know much about the Tascam , except that their products are usually better value than other brands at the same price.
I have read on Solen's web-site now about the sound absorbing materials they sell ,
however I will post more about that another time as I have to go now ,
and you have plenty to do before you need to buy any absorbents.
Hey Jeffrey ,
my problem is that I do not know how to load drawn diagrams onto forum pages.
I do not have a Scanner to upload any diagrams , and there is no-one here who can show me how to upload even if I could get use of a Scanner.
Yes , I could learn from some-one some-where how to do those things , but I have no spare time available to fit that in ...
unless you want to wait for ages till I reply to your posts ...
Instead , I refer you to the picture you posted in #50 on Page 5.
Imagine a Cylinder standing vertically on top of the woofer cabinet.
Cut the back half off the cylinder.
Draw that and post it here.
The half-cylinder does not have to be a perfect half-circle shape.
It can have another type of curve , and better is for it to have a flattened front
and then the curves around both sides.
The curves should extend to the full 90 degrees side around each side , because that will allow most of the Diffracted sound to disperse in backwards' directions.
If the curves stop at less than 90 degrees , more of the diffracted sound will be directed in forward-sides angles and reflect forwards off walls of room and sooner towards the listener ,
which will cause a greater degree of colouration and image diffusion than occurs when the diffraction is mostly to side-back angles and then much later reflected arrivals at listener's ears.
Where you have the two sides connected to your mids + tweeter half-box in #50 ,
replace the sharp 90 degrees corners with 3 inch or larger Radius curves.
Draw that and post here , and I will comment further if necessary.
D'Appolito mid-range will not work close to the ideal with only two upper mid-range cones.
For D'Appolito to be worthwhile it needs also two woofers/lower midrange , one at top and one at bottom.
Also , the Sensitivities/Efficiencies of two Celestion mid-cones will cause noticeably louder upper mids than will be from a single woofer/lower mids ,
and , to get Time Alignment you will have to recess the tweeter into behind the baffle to align it to the plane of two separate mid-cones ,
and that will raise another diffraction problem to have to be addressed.
Better is to keep it as you planned with a single mid-cone ,
however do buy the two additional mid-cones whilst they are available ,
because the old Celestion mid-cones are not as robust as the old woofers ,
so you should obtain spares for the mids now.
After you have assembled the loudspeakers you can try all four mid-cones and pick the two best condition by ear ,
then keep the other two as spares.
For the EMM-6 microphone , it is worth buying the Dayton EMM-6 Premium+ calibrated version , here:
Cross·Spectrum - Calibrated Dayton Audio EMM-6 Microphones for Sale
as that will be more useful than an uncalibrated and thus untested sample.
If that mic will work into the Tascam , then I suppose the Tascam is a good buy ,
however I do not know much about the Tascam , except that their products are usually better value than other brands at the same price.
I have read on Solen's web-site now about the sound absorbing materials they sell ,
however I will post more about that another time as I have to go now ,
and you have plenty to do before you need to buy any absorbents.
Last edited:
Hello everyone,
Small update. I have acquired a set of 44 crossovers and 66 crossovers. Alan, is there anything I should worry about with using the 66 crossover with the 44 mid? Any recommendations on how to mod the crossover to implement the Seas tweeter?
Next step is to make the cabinets and put them together. Just have a couple more parts to order and this is a go.
Small update. I have acquired a set of 44 crossovers and 66 crossovers. Alan, is there anything I should worry about with using the 66 crossover with the 44 mid? Any recommendations on how to mod the crossover to implement the Seas tweeter?
Next step is to make the cabinets and put them together. Just have a couple more parts to order and this is a go.
a reply to #59
Hi Jeffrey ,
I am puzzled by what you are buying , as it seems to not be in accord with your original design intention of getting the best possible from the Celestion drivers.
Both Celestion crossovers , 44 and 66 , were part compromises to achieve as good a performance that could be got from moderately priced components that were available in that design era.
Better can be bought now.
The 66 crossover has only one useful component for your project - the air-core 2.2mH inductor in the midrange filter section ,
because that has a suitable resistance for such application , and as air-core it will not saturate.
Use that inductor for your cone-midrange.
The 0.35mH air-core in the midrange filter is not suitable to use with the 44's mid-cone -
- use the 0.66mH air-core from the midrange filter of 44's crossover.
The two air-core inductors in the 66 crossover's woofer filter are too high in DCR to get the best from the woofer.
There is 0.14mH air-core in the tweeter filter of both 44 and 66 crossovers.
You can use it with the SEAS tweeter if an approximately 30 ohm resistor is connected in Parallel with the tweeter to reduce its Impedance in the crossover region to suitable to match the 0.14mH inductor.
30 ohms is not a common value , but there may be one in Mills MRA-5 , 5 watt wire-wound.
If no exact 30 ohms , then buy a 27 ohm that measures on the + side of its +/- Tolerance , or a 33 ohm that measures on the - side of its +/- Tolerance ,
and that will be possible with +/- 5% Tolerance resistors , but not with the +/- 1% Tolerance Mills MRA-5.
Buy in a minimum of 5 watts power rating.
As you have 0.14mH you are now stuck having to buy the not common 3.6uF and 11uF capacitors so as to work correctly with that value of inductor.
I hope those old crossovers did not cost you much , because if they were expensive you would have been better off spending the money on new inductors ,
including of sufficiently low DCR for the woofer filter , and a 0.15mH for the tweeter filter which would have allowed you to use the common available 3.3uF and 10uF capacitors with it ,
-{ with larger resistance in Parallel ... at least 39 ohms , or 43 ohms if available in 5 watt }.
I posted most of the above in my #43 on Page 5 of this thread .
I am currently posting in the Celestion 66 midrange thread about suitable inductors for a new parts' crossover.
Hi Jeffrey ,
I am puzzled by what you are buying , as it seems to not be in accord with your original design intention of getting the best possible from the Celestion drivers.
Both Celestion crossovers , 44 and 66 , were part compromises to achieve as good a performance that could be got from moderately priced components that were available in that design era.
Better can be bought now.
The 66 crossover has only one useful component for your project - the air-core 2.2mH inductor in the midrange filter section ,
because that has a suitable resistance for such application , and as air-core it will not saturate.
Use that inductor for your cone-midrange.
The 0.35mH air-core in the midrange filter is not suitable to use with the 44's mid-cone -
- use the 0.66mH air-core from the midrange filter of 44's crossover.
The two air-core inductors in the 66 crossover's woofer filter are too high in DCR to get the best from the woofer.
There is 0.14mH air-core in the tweeter filter of both 44 and 66 crossovers.
You can use it with the SEAS tweeter if an approximately 30 ohm resistor is connected in Parallel with the tweeter to reduce its Impedance in the crossover region to suitable to match the 0.14mH inductor.
30 ohms is not a common value , but there may be one in Mills MRA-5 , 5 watt wire-wound.
If no exact 30 ohms , then buy a 27 ohm that measures on the + side of its +/- Tolerance , or a 33 ohm that measures on the - side of its +/- Tolerance ,
and that will be possible with +/- 5% Tolerance resistors , but not with the +/- 1% Tolerance Mills MRA-5.
Buy in a minimum of 5 watts power rating.
As you have 0.14mH you are now stuck having to buy the not common 3.6uF and 11uF capacitors so as to work correctly with that value of inductor.
I hope those old crossovers did not cost you much , because if they were expensive you would have been better off spending the money on new inductors ,
including of sufficiently low DCR for the woofer filter , and a 0.15mH for the tweeter filter which would have allowed you to use the common available 3.3uF and 10uF capacitors with it ,
-{ with larger resistance in Parallel ... at least 39 ohms , or 43 ohms if available in 5 watt }.
I posted most of the above in my #43 on Page 5 of this thread .
I am currently posting in the Celestion 66 midrange thread about suitable inductors for a new parts' crossover.
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.