celestion 66 needs Mid-range - Page 94 - diyAudio
Go Back   Home > Forums > Loudspeakers > Multi-Way

Multi-Way Conventional loudspeakers with crossovers

Please consider donating to help us continue to serve you.

Ads on/off / Custom Title / More PMs / More album space / Advanced printing & mass image saving
Reply
 
Thread Tools Search this Thread
Old 8th January 2013, 07:34 PM   #931
diyAudio Member
 
Join Date: Apr 2006
Location: Midlands
Default Felt for Diffraction control , continued ...

Continuing from my Post #929 on Page 93:

SAE specification felts are known as Engineering Felts.

Industrial felts are known as Technical Felts.

Both types are available in various thicknesses ,
and in various Sheet or Roll sizes , and not all suppliers will have all options of sizes , so you will have to search the internet or a printed Trade Directory for your area.

I drew some options to scale for the 66 baffle with regard to the drivers' sizes and spacings.
I do not have access to a scanner to post my drawings here , so I will attempt a diagram using several keyboard key symbols , as follows:

# for the extended wood of the Top of Cabinet Lip ,

0 shows the inner edge of the vertical recessed strip -{ only Left side will be shown } ,

* for Felt covered area , including where the felt overlaps the chassis of the driver's mounting ,

X to form the circles that are the edge of the drivers' chassis ,
and for the clamps that hold the drivers to the baffle.

x for the flat sections of the drivers' chassis and the surfaces of the domes ,

O for the circumferences of the domes.

Note: there is no clamp shown at the top of the tweeter chassis because the felt is to cover over that clamp.

o is for uncovered areas of the baffle , because the blanks spacer bar does not save to completion of posting.

Edge details for the Right side are not shown , because the small typing box is not wide enough for me to calculate easily how to include the Right side ,
however the Right side is to be the mirror image of the shown Left side.

I do not know what caused the wavey line down through the Right area of the diagram ... ?

__________________________________________________ _______________
################################################## ##
################################################## ##
0************************************************* ********
0************************************************* ********
0**oooo******************************************* ********
0***oooooooooooo********************************** *********
0****ooooooooooooooooooooooo********************** **
0*****oooooooooooooooooooooooooooooooX*****X
0******oooooooooooooooooooooooooooXxxxxxx*xxxxxX
0*******ooooooooooooooooooooooooXxxxxxOxxxxxOxxxxX
0********ooooooooooooooooooooooXxxxxxOxxxxxxxOxxxx X
0*******ooooooooooooooooooooooooXxxxxxOxxxxxOxxxxX
0******oooooooooooooooooooooooooooXxxxxxO_OxxxxxX
0*****oooooooooooooooooooooooooooXoooXxxxxxxxXoooX
0****ooooooooooooooooooooooooooooooooooooX
0***oooooooooooooooooooooooooooooooooooooo
0***oooooooooooooooooooooooooooooooooooooo
0****oooooooooooooooooooooooooooooooooooXXX
0*****ooooooooooooooooooooXooooooooXxxxxxxxxxxxxxX oooooooX
0******oooooooooooooooooooooXooXxxxxxxxxxxxxxxxxxx xxxXooX
0*******oooooooooooooooooooXxxXxxxxxxxxxxxxxxxxxxx xxxxxX
0********ooooooooooooooooXxxxxxxxxxxxxxxxxxxxxxxxx xxxxxxxxxX
0*******ooooooooooooooooXxxxxxxxxxxxxxxxxxxxxxxxxx xxxxxxxxxxxX
0******ooooooooooooooooXxxxxxxxxxxxxxxxxxxxOOxxxxx xxxxxxxxxxxxX
0*****ooooooooooooooooXxxxxxxxxxxxxxxOxxxxxxxxxxxO xxxxxxxxxxxxxX
0****ooooooooooooooooXxxxxxxxxxxxxxOxxxxxxxxxxxxxx xxOxxxxxxxxxxxX
0***oooooooooooooooooXxxxxxxxxxxxxOxxxxxxxxxxxxxxx xxxOxxxxxxxxxxxX
0***oooooooooooooooooXxxxxxxxxxxxxOxxxxxxxxxxxxxxx xxxOxxxxxxxxxxxX
0****ooooooooooooooooXxxxxxxxxxxxxxOxxxxxxxxxxxxxx xxOxxxxxxxxxxxX
0*****ooooooooooooooooXxxxxxxxxxxxxxxOxxxxxxxxxxxx OxxxxxxxxxxxxX
0******ooooooooooooooooXxxxxxxxxxxxxxxxxxxxOOxxxxx xxxxxxxxxxxX
0*******oooooooooooooooooXxxxxxxxxxxxxxxxxxxxxxxxx xxxxxxxxxX
0********ooooooooooooooooooXxxxxxxxxxxxxxxxxxxxxxx xxxxxxxX
0*******ooooooooooooooooooXoooXxxxxxxxxxxxx*xxxxxx xxxxXoooX
0******ooooooooooooooooooXooooooooooooo*****oooooo oooooooX
0*****ooooooooooooooooooooooooooooo*************
0****oooooooooooooooooooooooo********************* **
0***ooooooooooooooooooooo************************* *******
0****ooooooooooooooo****************************** *********
0*****oooooooo************************************ *********
0******o****************************************** ********
0*************************************O*********** ********
0**********************Oxxxxxxxxxxxxxxxxxxxxxxxxxx xxxxxxxxxO*****
0*****************Oxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx xxxxxxxxxxxxxxxxxO
0**************O
0***********O
0*********O
0*******O
0*****O
0****O
0***O
0**O
0*O
0O
0<------------------------------this part circular space is the woofer--->
0O

__________________________________________________ _______________

The diagram is not exactly to Scale , because it was too difficult to get such to work with the between lines spacings , etc ...

No felt is needed below the woofer , as it would not be effective there.

For most of the area to be covered quarter inch thick felt is sufficient if it is very soft and porous ,
though half inch thick will produce somewhat better results.

For the area above the tweeter to the top wooden lip use 1 inch thick felt ,
or cut 2 pieces from half inch and stack them , because ALL the underside of that wood must be covered ,
or there will be destructive reflections.
To stick these two pieces together apply Fabric Adhesive to the back half only of the surfaces ,
because if there is adhesive solid at the front edge the sound waves will not absorb as well as possible.
The felt here must be placed so that it has open cut edges facing downwards for there to be absorption of sound from the tweeter.

In #929 I stated to cut Acute angles for the zig-zag side of baffle felt.
That will not work well with this narrow baffle , so cut Obtuse angles , as shown.

The felt above the tweeter , and between the mid-dome and woofer , can be cut to an obtuse angle apex , or slightly curved around the apex.
Do whichever you prefer the look of.

For all the felt cuts observe the following three conditions , or there will be some destructive reflections because the felt is not completely absorptive of sound waves.

(1)- no 90 degree angles , nor near to 90 degrees angles , in any of the felt and spaces between.
(2)- no cut felt lines if extended to the driver domes are to form a Radius of the dome.
(3)- no cut felt lines are to form Tangents to circles that would be concentric with either of the domes.

It is essential to observe the above conditions if you want to improve on the performance of the asymetrically placed square cut felt pieces in David Ralph's photos.

Cut small Triangle shape pieces of felt if you want to cover the clamps that hold the drivers to the baffle.
Point an Apex of each triangle towards the centre of each dome.

You will need very sharp sizzors , or sharp serrated sizzors , to cut thick felt neatly.

Also , I recommend you first cut templates of the shapes you intend to use from thick paper or thin cardboard
and use those over the felt when cutting it to shape.


DennyG ,
you will get more useful measurement results if you cover the baffles as above.
__________________
Alan

Last edited by alan-1-b; 8th January 2013 at 08:00 PM. Reason: to correct two mistakes , and add a sentance
  Reply With Quote
Old 17th January 2013, 10:17 AM   #932
apb is offline apb  United Kingdom
diyAudio Member
 
Join Date: Apr 2012
Location: Lancashire
Hi everyone. As a newbie, firstly a big thanks to Alan, who's advice I've followed in restoring my recently purchased 66s. The crossovers have been recapped with Jantzen caps and compensating resistors as advised, internal wiring replaced, and the sound has been transformed. They used to be a bit woolly and lacking in detail, but now have a clear finely detailed sound with much tighter bass. Not a subtle improvement by any means.

They sound very impressive, and regularly raise a smile, but I find the balance just a touch forward in the mid. I do recall them being this way inclined before the recap. I can't tell if it's the overall mid driver level that needs to come down, or if there's a crossover resonance creating a bit of a glare. I've experimented with varying the value of the esr resistors (within the accepted margins) which seems to have tempered it slightly.

So, what to do next? I'm tempted to upgrade the LF inductors to thicker gauge, to increase the level of the overall band up to 500Hz and further tighten the bass, and maybe balance better with the mid and top, but I'm wary of making them bass-heavy. Or maybe try an L-pad on the mid driver? Any thoughts would be appreciated.

Andrzej
  Reply With Quote
Old 19th January 2013, 07:58 PM   #933
diyAudio Member
 
Join Date: Apr 2006
Location: Midlands
Default interim ...

Where are you apb ?
Attach a National flag to your user name ,
and that is useful if I advise on where to buy parts from.

Thankyou for your report about the changes to the sound ,
as that is useful to all owners of 66s here !

You stated:
"but I find the balance just a touch forward in the mid"

Is that over the entire midrange ?
or only the low mids , or the high mids , or in a narrow band somewhere in the midrange ?

I mentioned the narrow band option because you stated:
"if there's a crossover resonance creating a bit of a glare."

All Band-Pass filters , except for simple First Order of one inductor + one capacitor , are resonant to some degree.
The degree depends on the Order of the filter -{ 2nd , 3rd , 4th }- and how much Resistance is in the circuit.
In the 66 , the resistance of the 2.2mH inductor reduces the resonance a little ,
and the compensating resistor I advised to connect in Series with the 3.9 or 4uF capacitor reduces it a little more.

Some of the Celestion mid-domes become more resonant in themselves as they age -
- see the various plots of Impedance and Frequency Response that sba has posted in places in this thread.
If either of your domes has become excessively resonant there is nothing you can do easily about it , but only more complicated repairs.
I recommend you listen carefully to each speaker by itself -{ not the pair together }-
to hear if one or other of the mid-domes sounds more resonant than the other , and if one does , then post about that here ,
as well as stating more about the midrange as I asked above.

List the resistance values you installed , in mid and woofer circuits.
Depending on your reply about the type of midrange excess , I may advise swapping around some of those resistors.

If the entire midrange is slightly too loud only , then replacing the two inductors in the woofer filter only with significantly lower DC resistance types will allow a little more bass output to balance the mids ,
plus will tighten up the bass a bit , BUT the choice of inductor has to be done carefully or the result may not be better but only different ...
I will be posting more about that for other interested participants here soon.
What I will be advising will not cause the l'speakers to be "bass heavy".

*** *** ***

I will return with a bit more about placing Felt on the front baffle when I have time available ,
hopefully soon ...
__________________
Alan
  Reply With Quote
Old 20th January 2013, 12:11 PM   #934
apb is offline apb  United Kingdom
diyAudio Member
 
Join Date: Apr 2012
Location: Lancashire
Thanks for the reply. I've now edited my location.

I can't tell just yet if it's one particular frequency or the whole band, but will experiment with a parametric EQ in order to try to locate the offending frequencies and report back.

I will post up a circuit diagram with all component values I used when I get a chance to have a look.

I will also post links to some LF inductors I found and am considering.

Regards

Andrzej
  Reply With Quote
Old 29th January 2013, 04:34 PM   #935
diyAudio Member
 
Join Date: Apr 2006
Location: Midlands
Default Continuing from #931 above , and more about felts ...

I stated in #931:

Quote:
Originally Posted by alan-1-b View Post

For all the felt cuts observe the following three conditions , or there will be some destructive reflections because the felt is not completely absorptive of sound waves.

(1)- no 90 degree angles , nor near to 90 degrees angles , in any of the felt and spaces between.
(2)- no cut felt lines if extended to the driver domes are to form a Radius of the dome.
(3)- no cut felt lines are to form Tangents to circles that would be concentric with either of the domes.

It is essential to observe the above conditions if you want to improve on the performance of the asymetrically placed square cut felt pieces in David Ralph's photos.
(3)- above needs clarification , I should have included at the end of the sentance:
... if the cut felt extends as a Tangent past a Radius of the dome.

because obviously for several of the felt cuts above if a line was extended from , it would be a Tangent to a concentric circle.

There is no major acoustic problem caused by what I drew in #931 ,
however it can be improved further , as below:

H is now the symbol for the Clamps' positions around the tweeter and mid-dome in the below diagram ,
and as in the previous diagram the top clamp above the tweeter is covered by felt.

All other symbols are the same as in the previous diagram ,
and the centre of both domes are now marked with O .
__________________________________________________ _______________
################################################## ##
################################################## ##
0************************************************* ********
0************************************************* ********
0*****oo****************************************** *******
0******ooooooooo********************************** ********
0*******oooooooooooooooooooo********************** **
0********ooooooooooooooooooooooooooooX*****X
0*********ooooooooooooooooooooooooXxxxxxx*xxxxxxX
0**********oooooooooooooooooooooXxxxxxOxxxxxOxxxx X
0***********oooooooooooooooooooXxxxxxOxxxOxxxOxxxx X
0**********oooooooooooooooooooooXxxHxxOxxxxxOxxHx X
0*********oooooooooooooooooooooooHxxxxxxO_OxxxxxxH
0********ooooooooooooooooooooooHoooXxxxxxxxxxxxXoo H
0*******oooooooooooooooooooooooooooooooooX
0******ooooooooooooooooooooooooooooooooooo
0*****oooooooooooooooooooooooooooooooooooo
0****oooooooooooooooooooooooooooooooooooXXX
0***ooooooooooooooooooooooHoooooooXxxxxxxxxxxxxxxx XooooooH
0****oooooooooooooooooooooooHXxxxxxxxxxxxxxxxxxxxx xxxxxXH
0*****ooooooooooooooooooooXxxxHxxxxxxxxxxxxxxxxxxx xxxxHxxxX
0******oooooooooooooooooXxxxxxxxxxxxxxxxxxxxxxxxxx xxxxxxxxxxxX
0*******oooooooooooooooXxxxxxxxxxxxxxxxxxxxxxxxxx xxxxxxxxxxxxxxX
0********oooooooooooooXxxxxxxxxxxxxxxxxxxxxOOxxxxx xxxxxxxxxxxxxX
0*********oooooooooooXxxxxxxxxxxxxxxxOxxxxxxxxxxxO xxxxxxxxxxxxxxX
0**********oooooooooXxxxxxxxxxxxxxxOxxxxxxxxxxxxxx xOxxxxxxxxxxxxxX
0***********oooooooXxxxxxxxxxxxxxOxxxxxxxxxxxxxxx xxxOxxxxxxxxxxxxxX
0************ooooooXxxxxxxxxxxxxxIxxxxxxxxxxOxxxxx xxxIxxxxxxxxxxxxxX
0***********oooooooXxxxxxxxxxxxxxOxxxxxxxxxxxxxxx xxxOxxxxxxxxxxxxxX
0**********oooooooooXxxxxxxxxxxxxxOxxxxxxxxxxxxxx xxOxxxxxxxxxxxxxX
0*********oooooooooooXxxxxxxxxxxxxxxOxxxxxxxxxxxx OxxxxxxxxxxxxxxX
0********oooooooooooooXxxxxxxxxxxxxxxxxxxxxOOxxxx xxxxxxxxxxxxxxX
0*******ooooooooooooooooXxxxxHxxxxxxxxxxxxxxxxxxxx xxxxxxHxxxxX
0******ooooooooooooooooooooHXxxxxxxxxxxxxxxxxxxxxx xxxxxxxXH
0*****ooooooooooooooooooooHoooooooXxxxxxxxxxxxxxx xxXooooooH
0****ooooooooooooooooooooooooooooooooooooXXXoooooo ooo ooooooo
0***ooooooooooooooooooooooooooooooooooooo***oooooo ooo ooooooo
0****oooooooooooooooooooooooooooooooo********** oooooooooooo
0*****ooooooooooooooooooooooooooo***************** oooooooo
0******oooooooooooooooooooooo********************* **ooooo
0*******ooooooooooooooooo************************* *****oo
0********oooooooooooo***************************** ********
0*********ooooooo**********************O********** * *******
0**********oo***********Oxxxxxxxxxxxxxxxxxxxxxxxxx xxxxxxxxxxO*****
0*****************Oxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx xxxxxxxxxxxxxxxxxxx
0**************O
0***********O
0*********O
0*******O
0*****O
0****O
0***O
0**O
0*O
0O
0<------------------------------this part circular space is the woofer--->
0

__________________________________________________ _______________

The diagram is not exactly to Scale , because it was too difficult to get such to work with the between lines spacings , etc ...

Note for this version there is an Apex of felt pointing towards the centre of the mid-dome , { same as for the tweeter } ,
and that will cause better mids absorption , along with whatever residual reflection being diverted to further absorptive cut edges of felt.

***

Alternate:

If you cannot get thick soft SAE spec Engineering felt , or thick soft Technical/Industrial felt ,
then for the thinner Craft felts buy a type that is equal thickness to the flange of the mid-dome ,
and cut precise circular holes in it for tweeter and mid-dome , and cover the entire front baffle from above the woofer , and to the inner edges of the vertical recesses.
Roll some up loosely and stick it under the overhanging lip at the top of the cabinet to cover the entire under surface of that lip.

***

Felt Colour:

Pure Wool and 95% Wool felts are White colour.
If you try to dye it a colour the later dryed dye will cause stiffness within the felt , thus less sound absorption , so do not dye white felt.

If you do not want white colour , then in the SAE specification felts F-11 is pale grey , and F-13 is slightly darker grey.
See the photos of the F-11 and F-13 grey felts on David Ralph's web-site.
At least one photo has both greys - the colour difference is small , but is noticeable.
These are not quite as absorptive as F-10 , but are not bad , and are better than home-dyed felt for sound absorption.

In Technical/Industrial felts there are some pale grey versions ,
which same as for Engineering/SAE , the grey ones are less than 95% Wool ,
but there are some in the very soft , low density specification.

Craft felts are manufactured from various colour dyed strands of Wool , and Rayon , etc ... { and from synthetic fibres }.
Because the fibre is pre-dyed , the felt made from it is softer than if dyed after the felt was formed ,
so if you want coloured felt buy Craft Felt , but find a Pure Wool type if you can ,
or at least a high wool/low rayon composite , and not a fully synthetic fibre felt , such as Acrylic , etc ...

***

I think I have posted sufficient about Felt , but do ask if something is not sufficiently explained.

Next time here I will return to finishing another of the topics we have been discussing - maybe for "reggie" , if he is reading ...
unless something else is asked.
__________________
Alan

Last edited by alan-1-b; 29th January 2013 at 04:43 PM. Reason: to add three words
  Reply With Quote
Old 30th January 2013, 10:28 AM   #936
reggie is offline reggie  Australia
diyAudio Member
 
Join Date: Feb 2010
Location: melbourne australia
I'm still here. Attempting to set my printer up to scan the Celestion 66 speaker review for the Celestion "library". Yes, I'll need your advice on parts to order for my new cross-over soon - before the Australian dollar falls out of favour (haha).

Reggie

Reggie
  Reply With Quote
Old 5th February 2013, 04:04 PM   #937
diyAudio Member
 
Join Date: Apr 2006
Location: Midlands
Default Hiquphon OWII tweeter , and continuing some other back-log

G'day Reggie ,

Looking at data for Hiquphon OWII tweeter here:
Hiquphon OWII (OW2) dome tweeter, technical data
and comparing it to the SEAS 19TFF1 and Hiquphon OWI ,
note that it is about 1dB more efficient in its lower treble , and 3dB more in its upper treble.
See the step in its frequency response between 4kHz and 5kHz.
That is a good co-incidence for this Celestion 66 project , because the upper crossover point is nominally 5kHz ,
thus for owners who have lost a bit of upper treble hearing the OWII can supply extra treble ,
including too much as Reggie found !

The Impedance plot is close to flat in the region around the crossover point , thus close to a standard 3rd order filter will work ,
and will work better with a resistor connected in Parallel with the tweeter as that will flatten the Impedance a little more.
The resistance can be chosen so that it in parallel with the tweeter reduces the total impedance to a suitable value to enable standard E Series capacitors to be used ,
and hopefully an available inductor.

Reggie has an L-Pad of 15 ohms parallel + 1 ohm series around his OWII.
That attenuates by about -2dB , and seems to have resulted in a suitable treble level for Reggie ... -{ yes or no Reggie ? }.

It will still be slightly brighter than treble would be with the SEAS 19TFF1 or OWI , but some listeners may need this slight additional amount of treble.



Quote:
Originally Posted by reggie View Post
Hello Alan,

Some responses to your questions and suggestions.

6. Solens 400v. Mcap 400v. As for the size well, the present crossover is in 2 parts, bass and mid/tweeter but my intention is to have the crossovers outside the cabinets (at least initially).
7. If you think it is a waste to get the big Obbligatoís then we will retain the Solens. My concern was with the disparity between the values of the original capacitors (144) and my current ones (136) so your suggestion to use the 8.2uf is welcome.
8. Alan, unless something unusual occurs, I will probably keep these speakers (Iíve had them for 36 years now). So Iím up for spending money on totally renewing the crossovers. Thatís new capacitors, coils and resistors. You tell me what I need and Iíll get them.


Regards
Reggie
Dealing with all the above , and incorporating adapting the OWII into the 66 crossover:

Unless there is something wrong with the original Celestion treble inductor it may as well be retained.
The only other reason to change it would be if one's amplifier had difficulty driving the lower impedance load in the treble than is in the midrange ,
however as the impedance is lower than the midrange in the bass , and similar in the bass to what it is in the treble ,
then most amplifiers that drive its bass well will drive its treble well ... but not all , so decide that before proceeding.

For the L-Pad ,
keep the 15 ohms in parallel , in at least 5 watt if a good quality resistor , or higher power rating if a dubious quality resistor.
Buy 1.2 ohms for the series resistor -{ power ratings same as above }.
This will attenuate by slightly more than 2dB , and will set the total impedance to match what is required for the 0.14mH Celestion inductor.

For the capacitors use 3.6uF for the input cap , and either 11uF or 12uF for the output cap.
11 or 12 - which will work better ?
I do not know , however I think for shaping the treble to be slightly higher in output in the upper treble to compensate for hearing loss ,
along with the specific L-Pad resistors above in conjunction with an OWII that 11uF is likely to be more suitable ,
however the final result will depend on the upper end of the frequency response of the mid-domes ,
which as we have seen from sba's plots varies quite a lot between samples in the upper mids.
Some of them have sufficient upper mids output to not need any help from the tweeter, but for others it would assist upper mids' level if the treble sloped off a bit more gently ,
and for such cases use at least 12uF.

To simplify the choice , I recommend starting with whichever cap value is available from the seller of the 3.6uF , and listen to the result.

Given what else Reggie wants I think it best he uses SoniCap Gen I series
for which there is 3.6uF and 11uF available.

For UK and European residents there is 3.6uF and 12uF available in Jantzen Cross Caps for lower price than Sonicap { from USA } ,
however these may not be quite as good in audio quality , though still better than Solen ,
and are physically larger size than Sonicaps.

For the midrange filter Reggie ,
you have a Mundorf 22uF for the input cap.
This will cause slightly less level at the lowest midrange frequencies than the Celestion spec. 24uF ,
however , with the specific samples of the mid-domes and woofers in your 66s ,
if you have a pleasing amount of mids around the lower crossover region then leave the 22uF there , and the Mundorf is quite a good quality capacitor.
You can leave the Solen 3.9uF as output cap of the midrange filter , or upgrade it to a 3.9 or 4uF Sonicap
-{ both are available , and I'd choose the 4uF if it will fit the board - note it is shorter but fatter than the 3.9uF }.
If you want a little more of the lowest mids , then buy a 24uF Sonicap.

You can fine-tune the upper midrange/lower treble output by changing the resistor that is in series with the 4uF midrange output cap.
2.7 ohms , 3.3 ohms , 3.9 ohms , I recommend you leave in whichever you have there for now till you hear the result with correct L-pad and new caps to the tweeter.

If you want to buy a Mills MRA-5 for mids at same time as you are buying for treble , { as these Mills are available from Sonicap's web-site via Mail-order } ,
then buy 3.9 ohms/5 watt , as I think you have 2.7 ohms currently if I am remembering correctly ... ?

Sonicaps and Mills MRA-5 are available Mail-order from:
Sonic Craft High-End DIY Audio Parts

Other resistors:

Either 1.8 ohm or 1.5 ohm , 5 watt , in series with the 22uF/24uF mids' input cap.
Buy both values and start with 1.8 ohms in midrange and put the 1.5 ohms in series with the output cap of the woofer filter.
If you want a little more midrange level then swap those two resistors positions.
{ If I remember correctly you already have 1.5 ohms , but 1.8 ... ? }

For the between the inductors cap of the woofer filter use 1 ohm in series -{ which I think you already have ... ? }

Buy two 8 ohm , or 8.2 ohm , also these can be Mills 5 watt , but there will be very little current through these , thus lower power is OK ,
so if you want to buy cheaper resistors for these then buy in 1 watt Carbon film from a local shop.

In the woofer filter you currently have 68uF Solen caps in both positions.

Option 1 - leave 68uF + 1.5 ohms as the output capacitor and resistor.
Connect 1 ohms in Series with the other 68uF cap.
Connect the 8.2 ohms resistor in Series with your Solen 8.2uF cap that you are taking out of the tweeter filter.
Connect the 8.2 + 8.2 series in Parallel with the 68 + 1 series.
Do NOT connect the pairs together at their central junction capacitor to resistor joints.
Connect only at the end points of each series pair.
This option should give sufficient filtering , and it may be the optimum alternate if you are deciding to keep the 22uF in the midrange filter.

Option 2 - If you are not happy with the lower midrange sound after listening to the above ,
then remove the 8.2uF cap and install the Solen 3.9uF cap that you are taking out of your tweeter filter.
Leave the 8.2 ohm resistor there.
Now you will need another pair of 3.9uF caps , which you will have if you upgrade the output caps of the midrange filter to Sonicaps.
Buy two 15 ohm , 1 watt carbon film resistors from a local shop and connect these , one each in Series with the two additional Solen 3.9 uF caps.
Connect each of these series pairs in Parallel with the 68uF + 1.5 ohm series pair at the output of the woofer filter.
{ Connect as described in Option 1 , not short-circuiting the centre joins of each series pair to the other .}
You will then have effectively 72uF in each capacitor position , as is the Celestion specification ,
however try Option 1 first , because I think it may work a little better.
There were not as many capacitance values economically available when Celestion designed the 66 , thus they used the closest that was available , 72uF for both positions.
Obviously that worked well enough , but I think 76uF and 68uf could work a little better in the positions I described above.


---> DennyG - you have effectively 74uF in the central cap's position.
Ideally the 6.2uF cap you connected there needs its own separate resistor.
Use 8.2 ohm or 10 ohm in 1 watt Carbon film from a local shop.
If you have some 1 watt resistors already , then so long as not less than 4.7 ohms or larger than 12 ohms , try those you have.


Reggie ,
further about your worries of useable life of the old drivers in the 66s.
For a modern loudspeaker which will give similar scale of sound to a 66 , and will last longer than you ,
look on your local ebay , and in B&W stockists for Trade-ins of:
B&W 800S - that is the S version of the model 800.
These may be available now at affordable prices since B&W replaced them with the new 800D series.
D is Diamond tweeter version , and very expensive.
S have a good quality metal dome tweeter.
There was an interim Signature 800 , which was an 800S with deluxe finish , but owners may want a lot of money for those.

800 has a pair of 10" woofers , and have faster sound than 801 with its 12" woofer.
Both have a good quality mid-cone.

802 has a pair of 8" woofers , plus mid-cone and tweeter , but these do not have the weight of sound of the 800.

I do not recommend 803 and 804 for Celestion 66 lovers.
{ 805 is good for a stand mounted 2-way.}

Before the S series there was the Nautilus series , but these do not sound as good as the S series ,
however are not bad , but listen first because they do sound quite different to the S series.

I recommend you listen to one of the versions of the 800 at a B&W dealer , and with a high-end , high power amplifier ,
or take in your own amplifier and ask for demo in a small room to suit your amplifier's power.

Their vent/port can be blocked if the bass output is excessive , and with the 800 the bass goes low enough in frequency to not need the port output.
I think they sound better with the ports blocked.
Stuff the ports tightly with closed cell plastic foam or with any tightly rolled soft fabric - wool would be best.

***

I will return to post about upgrades to the Inductors in the woofer filter for higher and faster bass output when I have time available.
These will be from different sellers than the Sonicaps and Mills resistors ,
so you can buy those and start now , but do post here if you have interim queries.
__________________
Alan
  Reply With Quote
Old 8th February 2013, 05:30 PM   #938
diyAudio Member
 
Join Date: Apr 2006
Location: Midlands
Default a correction , and some additions to my #937

Quote:
Originally Posted by alan-1-b View Post

Buy two 8 ohm , or 8.2 ohm , also these can be Mills 5 watt , but there will be very little current through these , thus lower power is OK ,
so if you want to buy cheaper resistors for these then buy in 1 watt Carbon film from a local shop.

In the woofer filter you currently have 68uF Solen caps in both positions.

Option 1 - leave 68uF + 1.5 ohms as the output capacitor and resistor.
Connect 1 ohms in Series with the other 68uF cap.
Connect the 8.2 ohms resistor in Series with your Solen 8.2uF cap that you are taking out of the tweeter filter.
Connect the 8.2 + 8.2 series in Parallel with the 68 + 1 series.
Do NOT connect the pairs together at their central junction capacitor to resistor joints.
Connect only at the end points of each series pair.
This option should give sufficient filtering , and it may be the optimum alternate if you are deciding to keep the 22uF in the midrange filter.

Option 2 - If you are not happy with the lower midrange sound after listening to the above ,
then remove the 8.2uF cap and install the Solen 3.9uF cap that you are taking out of your tweeter filter.
Leave the 8.2 ohm resistor there.
Now you will need another pair of 3.9uF caps , which you will have if you upgrade the output caps of the midrange filter to Sonicaps.
Buy two 15 ohm , 1 watt carbon film resistors from a local shop and connect these , one each in Series with the two additional Solen 3.9 uF caps.
Connect each of these series pairs in Parallel with the 68uF + 1.5 ohm series pair at the output of the woofer filter.
{ Connect as described in Option 1 , not short-circuiting the centre joins of each series pair to the other .}
You will then have effectively 72uF in each capacitor position , as is the Celestion specification ,
however try Option 1 first , because I think it may work a little better.

B&W - Bowers & Wilkins:

800 has a pair of 10" woofers , and have faster sound than 801 with its 12" woofer.
Both have a good quality mid-cone.

802 has a pair of 8" woofers , plus mid-cone and tweeter , but these do not have the weight of sound of the 800.
Correction:
About B&W 801 , it was the older versions that had the 12" woofer.
Recent versions had a 15" woofer, though there seems to be no 801D for the new Diamond tweeter series ,
probably because it was found that most listeners to the S series preferred the 800S with its two 10" than the 801S with its 15".

Reggie ,
if shops in your city have only the 802 , it is still worth hearing , as it will give you some idea of the large sound , and it does have same mid-cone as 800.

*

I have looked again at the Soniccraft web-site.
The Sonicap GEN 1 caps that are on the Specials page at reduced price are their older +/- 10% tolerance versions.
I recommend you do not buy those because all will be distant in value to their nominal ,
because the close to nominal samples will have been sold to buyers who pay for sample measured capacitors.
Buy the new versions from the GEN 1 page , and in the 200 volt rating , as you do not need the larger 400 volt size for Celestion 66
- 66 will never accept such voltage , nor even 200 volts , even on peaks.


I see Sonic Craft sell some audio quality 1 watt resistors:

There is no 8.2 ohms , but if you want to buy a quality 1 watt for the 8.2uF cap addition to your woofer filter , the Takman 4R7 -{ 4.7 ohms }- will be OK there.

For the Option 2 , 15 ohms may be a little high , I now think 12 ohms may be better.

There is no 12 ohms in 1 watt from Sonic Craft , but the Takman 11R will suffice , and the PRP 11R may be a little better.
{ 11R is 11 ohms.}

*

Reggie ,
the combined effect of the Frequency response and the Impedance variation over the bandwidth of the Hiquphon OWII with the particular L-Pad resistors
and capacitor values + inductor you have currently is such that there will be a bit of a dip in output in the low end of the treble , and a slope up to a small peak in the upper treble.
You may not hear the upper treble peak , { but perhaps your wife can ...} ,
however the lower treble dip will cause the average of treble level to sound lower output than will be the case after you install the 11uF cap which will fill in the lower treble dip.
Thus do not worry that increasing the Series resistor to 1.2 ohms will reduce treble to lower than is now audible - it won't after the effect of 11uF cap.
If you are happy with the average level of treble now , then proceed as I advised in #937 ,
but if you think you may want more treble overall , then buy 1.1 ohms and 18 ohms in Mills MRA-5 , 5 watt , for the L-Pad.
If you think you may want less treble than you have now , post here before you buy and describe the degree of treble difference you want.
__________________
Alan
  Reply With Quote
Old 12th February 2013, 04:48 AM   #939
reggie is offline reggie  Australia
diyAudio Member
 
Join Date: Feb 2010
Location: melbourne australia
Thank you Alan. I'm away for a week (at the beach) so I've printed this off and will sit and read it (while others are frolicking). I tried to post the Celestion review but the machine tells me it's too long for a post (679,000 bits while only 30,000 permitted - funny because it's only one page in length). I'll have to find a circumvention (?)

Reggie
  Reply With Quote
Old 12th February 2013, 03:40 PM   #940
diyAudio Member
 
Join Date: Apr 2006
Location: Midlands
Default to Reggie ; DennyG ; and whoever is following these topics

G'day Reggie ,
there was reported to have been a cyclone , or similar , along the East coast of Australia ,
and there you are "at the beach" !
Well , I hope your prints did not get blown away ...

As for posting the Celestion review , given the almost 23:1 ratio of bits versus bits permitted ,
you would need to substantially reduce Resolution ... but that would likely be irrelevant to a direct Posting here.

I do not have expertise in these matters , however it seems that one posts the specific page onto a file saving web-page ,
and then posts a Link from here to that file saving page.
Example: have a look on Page 87 of this thread at #866.
See how much DennyG has managed to get stored on:
home.exetel.com.au
and linked to here.
As that seems to be an Australian web-site , you may be able to easily use it.


Perhaps some other readers here can advise Reggie on how to store a lot of data and link it to here ..?


Quote:
Originally Posted by DennyG View Post
I now have a 2nd hand M-Audio 'USB Audiophile' audio interface for future spkr testing. Not tried it yet but it has reasonably good specs and zero latency apparently.

A Sydney, Australia pair of 66's just went for AUD1225 on ebay: Item 221147740783
G'day DennyG ,

hopefully those Celestion 66s work !
... or maybe their buyer will find us here ...

I don't know how there could be "zero latency" through a digital device ,
but perhaps the reference is to zero added latency when adding new data to stored data within the device ...

As you may be about to start measuring again , I refer you back to what I posted on Page 89 in #890 ,
and also to the baffle treatment I posted above , as that will give you results with minimal diffraction and reflection errors.

What I forgot to advise you last time , particually as you used your Sugden A28 to drive the speakers whilst measuring:
those old capacitors in the Sugden would have deteriorated in a non-linear manner ,
and differently between samples , thus one channel of the Sugden itself would have measured differently to the other - particually in the Treble.
Even with the new caps installed , as they are Electrolytics they will not be identical sample to sample for caps of the same value ,
thus , do all the measuring using ONE channel of the amplifier for all drivers ,
or at least for all drivers of the same type.
Eg: you may find your HF2000 tweeters are not as different to each other after you measure both using the same channel of the amplfier.
One needs to ensure one is measuring the differences between the drivers.
Later one can measure to find if any degree of difference between channels of the ampifier ...
and for that I recommend using a 10 watt 6.8 ohm wirewound resistor for the basic measurement for reference to Celestion 66 impedances.
Use same sample of the resistor to measure each channel of the amp.
Use sufficient power from the amp for the resistor to get slightly warm to touch - not hot , nor cold.
If you want to measure for higher power output level differences you will have to buy 20 or 25 watt or higher power resistors ,
or Parallel connect several 10 watters - which is what I do.
Amplifiers that measure close to identical channel to channel into Resistive loads sometimes do not into Reactive loads.
To test into a reactive load , connect the 6.8 ohm resistor to the output of one Celestion midrange filter , and 4.7 ohm/10 watts to outputs of its woofer and tweeter filters ,
and use that one test filter to measure each channel of the amplifier.
Here you will not get equal output across the full frequency bandwidth.
What one needs to do here is look at the relative output of each channel versus the other at the same frequency for all the measured frequencies.
__________________
Alan

Last edited by alan-1-b; 12th February 2013 at 03:47 PM. Reason: to correct spelling and add a word
  Reply With Quote

Reply


Hide this!Advertise here!
Thread Tools Search this Thread
Search this Thread:

Advanced Search

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are Off
Pingbacks are Off
Refbacks are Off



New To Site? Need Help?

All times are GMT. The time now is 05:20 PM.


vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2014 DragonByte Technologies Ltd.
Copyright ©1999-2014 diyAudio

Content Relevant URLs by vBSEO 3.3.2