Go Back   Home > Forums > >
Home Forums Rules Articles diyAudio Store Blogs Gallery Wiki Register Donations FAQ Calendar Search Today's Posts Mark Forums Read

Multi-Way Conventional loudspeakers with crossovers

celestion 66 needs Mid-range
celestion 66 needs Mid-range
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
Thread Tools Search this Thread
Old 6th May 2007, 10:40 PM   #41
tonedef2 is offline tonedef2  United States
diyAudio Member
Join Date: May 2006
Confused as to the different components and a good replacement, brands? solen? daton? theta?$eek, what do you think Alan?
Also this is my week spot, are all those different types, black/green/yellow things just filter caps?

-3.3uf 160 ?
-2.2 uf/10/100?
24 mfd 50v i know this one...


  Reply With Quote
Old 8th May 2007, 04:12 PM   #42
alan-1-b is offline alan-1-b  United Kingdom
diyAudio Member
Join Date: Apr 2006
Location: Midlands
Default Ditton 66 - tweeters and cabinet top problem.

Given the reports by various owners about treble quality, and requests by some for recommendations for substitute tweeters, I will post here :-

First, look at the overhang of the top panel of the cabinet, yes it may look asthetically pleasing, but it is acoustically poor design in general, and more-so in this case as it is so close to the tweeter -{it looks about 2" away only in Grahame's photo in this thread}.

This will cause reflections of treble frequencies, delayed in time, thus various peaks and dips in the frequency response in the listening area.
Such would have been less of concern to Celestion when this speaker was first on the market because most buyers back then had not heard quality treble reproduction, but did expect loudspeakers to look good as furniture - especially at the price the Ditton 66 was.

Substituting a new dome tweeter will not solve the reflections problem, thus one should solve that first and listen to what the HF2000 tweeter is capable of, and then decide whether to try modern tweeters, or simply install better audio quality capacitors.

Obtain some double sided adhesive tape and some high-frequency sound absorbtive material - natural wool as in a felted woollen insulation batt is best, if not excessively compressed - there needs to be sufficient space between the fibres for sound to enter and vibrate the individual fibres.
Rockwool insulation batt or Fibreglass insulation batt of some grades is suitable also.
Next best is an Acoustic grade plastic foam, but only buy one for which the manufacturer has published absorbtion co-efficients, because not all expanded plastic foams absorb evenly, and some do not absorb much at all, and some partially reflect sound - particually part of the treble spectrum.
Bonded Acrylic or Acetate fibres, Dacron, etc ... do not absorb as well as the above listed materials, and there is only space for a thin strip between the tweeter's flange and the underside of the overhanging cabinet top.

Of course, for a simple trial one could roll up a woollen sock -{natural wool, not Polyester or Acrylic - those won't work well, if at all}- and use the double sided tape to secure it above the tweeter and to fully block the underside of that overhang, then listen !

For absorbtion the strip of a batt must be cut and placed so that an open cut edge is facing the tweeter, as the sealed or semi-sealed sides of most batts do not absorb treble frequencies well, and some sealed types reflect treble.
Remember, those batts are designed to absorb heat and/or midrange sound frequencies.

There looks to be space for only a about a 1" wide strip between the underhang and the egde of the tweeter flange.
For High Fidelity sound, it is not a good practice for a thick absorbtive to be placed at the edge of the tweeter dome itself, thus place it only to the edge of the tweeter's mounting plate, OR, cut a taper across about 1" of a 2" thick strip - that is, extend it to between 1/4" to 1/2" away from the dome but with it cut tapering away from the dome, but with still at least 1" thickness at outer part under the underside of that overhang to get sufficient absorbtion under the outer edge of the overhang.
Cut the strip to extend the full width of the baffle.

If you want this to work with the Grille frames on, then you will have to think of a way to acheive that - perhaps cut a taper under the top edge of the grille frame and stick the absorbtive to that, and also stick strips down the top parts of both insides of the sides of the grille frame down to a few inches below the level of the midrange dome.
2" thick would fit there, and absorb well enough to give less spikey or rough edged sound from treble and upper mid frequencies.

__________________________________________________ __

For substitute tweeters, if still wanted after the above, and new capacitors have been installed, do you want tweeters with 3 screw holes such as the originals are ?
Or, will you tolerate the now more common 4 or 5, etc ... screw hole tweeters and drill new holes in your 66s baffles ?

Yes, there are some 3 hole tweeters still available, but you will have to measure the spacings of the Celestion tweeters' holes and look at the current manufacturers' data to see if their spacings match.

Look at the data for the Hiquphon OW1 tweeter, the SEAS 19 series tweeters as these are good quality 3 hole types.
I do not know the exact Voltage Sensitivity{Efficiency} of the Celestion HF2000 -{does any reader of this ? - please post if you do}.
I estimate it would be about 89dB/w/m, and that is the figure for the OW1 and a few of the SEAS 19 series models.
__________________________________________________ __

by "daton", I think you meant to type "Dayton" ?
and by $eek, I think that is your reaction to the price of the Thetas !

I will email you to discuss capacitors. I have posted the other information above for any other Forum members who may be following this thread now or in the future.

  Reply With Quote
Old 8th May 2007, 08:44 PM   #43
Grahame is offline Grahame  United States
diyAudio Member
Grahame's Avatar
Join Date: Mar 2007
Just a quick update guys,

I had a closer look at the installed crossover, even managed to remove it a get a picture of the traces on the back which I can post when I have time.

I'll also try and make a schematic using EAGLE (light) software, and spend some time working out if the topology is the same as for the point-to-point wired version.

At least I know the build date now. Undid 4 hex nuts at corners of crossover to remove and get view of back of circuit board.
Behind board was stamped "Nov 22 1977" so ~30 Years then.

New text values for the "Green" capacitors (C1, C2, C3, C4) are

C1+2 and C3+4 are in parallel on the traces.

160 (V)

1.5 uF

C1+C2 = 4.7 + 1.5 = 6.2 uF



C3+C4 = 3.3 + .68 = 3.98uF

Which looks like someone was trying to approximate these values (6, 4) with newer(?) capacitors on the point-to-point crossover.

I'm getting up to speed, learning all about crossovers + electronics.

If I've pattern matched (it looks the same as ...) the crossover schematic correctly with here

The Tweeter has a Third order high pass ( 2 C's in series , 1 L in parallel )

The Sub has a fourth order Low pass ( 2L's in series , 2 C's in parallel)

Which I guess leaves a band pass filter for the Mid - I might need some help on that.

I plotted the schematic by hand - matching it to the layout on the point to point board - I'll do the same for the Circuit board version.

There is a local electronics shop - if they are friendly, I may be able to get them to measure the inductance of the inductors - so we'll know what their values are - and see how the capacitors have changed from their printed values.

Given the values - we should be able to derive / simulate the cross over points / slopes.

I've been loaned some (sound) measurement equipment - so with a bit of luck I may also be able to plot some frequency response graphs!

It will be interesting to see the before + after results.

Thanks for all your help guys - we should be able to get some definitive info together for all the other '66 owners out there.

tonedef2 - you mentioned spikes - what type did you use, and how did you affix them?
  Reply With Quote
Old 9th May 2007, 01:18 AM   #44
tonedef2 is offline tonedef2  United States
diyAudio Member
Join Date: May 2006
Grahame you are all over this crossover, i just dont have the time to learn it right now....new job lots there to learn.

It will be very interesting to see your results.

Spikes were just the standard wood anchore type...drill hole, slip in threaded anchore and screw in spiked feet, picked em up at Parts Express.

Alan, I'm using the tweeter rings that surround the whole tweeter, do you think i will gain by using your defraction method you mentioned above?

  Reply With Quote
Old 9th May 2007, 02:02 PM   #45
alan-1-b is offline alan-1-b  United Kingdom
diyAudio Member
Join Date: Apr 2006
Location: Midlands
Default Thanks and more ...

Thankyou for posting all that information Grahame ! , and whatever else you can post and have measured may be helpful also for getting the best from 66s with modifications , etc ...

The Inductors are probably the values that are printed on the layout you posted previously, and given there in microhenries, = uH, instead of the more usual millihenries, = mH.
Eg:- 547uH = .547mH
and 1163uH = 1.163mH
etc .... , as this does calculate OK for the 4th order filter on the woofer, for what seems to be a 4ohm woofer with 1.359mH ; 1.358mH ; 72uF ; 72uF , BUT, do not expect to get these values from a Standard 3-way x-over calculator for any textbook filter name, because this specific electrical filter has been adjusted to take into account the woofer's rising impedance at 500 Hz, and the possible rising frequency response from about 400 to 800 Hz - typical of 12" drivers.

Similarly for the mids high-pass and low-pass filter arms. Those are electrically 2nd order, but there will be roll-off at both ends of the MD500's response, thus the actual filter slopes in the loudspeaker could be 3rd order or 4th order - I'll guess at 4th order Hi-pass and 3rd order Lo-pass, but I could be mistaken, and they couild both be in-between slopes !
Especially the Lo-pass, because the Hi-pass to the tweeter is slightly unusual.
Yes, it is electrically 3rd order, but 1.163mH is much larger than would be used into a simple resistive load for 5kHz with those capacitor values.
The HF2000 will have a higher impedance in the band a little below its cross-over point, and much higher impedance at its Fs, which is probably about 1 1/2 or 2 octaves below cross-over, but does affect the electrical filter, thus the larger inductor may be intended to compensate for that, IF my estimate is correct that those numbers are the inductor values.

You posted the MD500s measured 6.7 and 6.9 ohms DC resistance, and that is about right for a nominal 8 ohm driver.
The woofers are apparently 4 ohm - if so they'd measure about 3.5 ohm DC.
I think the tweeters are 8 ohm, thus would measure about the same resistance as the MD500s - if you have time, measure the DC resistance of both tweeters and both woofers and post.

If you can disconnect and measure the individual driver's frequency responses with no cross-over connected, as well as each with cross-over connected and post both the without and the with, then we can solve this cross-over, and maybe make a mod. to it that could improve its High-Fidelity.

EIRE is an old brand of capacitor, and those are probably Polyester or Metalized Polyester types - not great for sound, and can start to leak when very old, especially if simple Polyester, but they're not as risky as the black electrolytic caps which should be replaced now to avoid undue risk.
The inductors are unlikely to be faulty. Measure each for DC resistance when you have the drivers all disconnected, and see if they are close to the ohm values shown on that schematic you posted.

As to whether to use 6uF or 6.2uF depends on whether one intends to change the tweeters, and what the frequency response plots ot the HF2000s look like without the cross-overs connected.
Try a measurement of the HF2000 with the 6uF x-over and another with the 6.2uF x-over.

I suggest leave those 6 and 6.2 as is and try the anti-reflection ideas I posted, then you can substitute your x-overs and hear the effect of the different value caps to the tweeters more clearly.
__________________________________________________ __

the tweeter rings you are using, how far in front of the baffle do they extend - thickness above baffle ?
And, how wide are they from inside of ring to outside of ring ?

  Reply With Quote
Old 10th May 2007, 12:36 AM   #46
tonedef2 is offline tonedef2  United States
diyAudio Member
Join Date: May 2006
Hi Alan,

This is where i picked them up...

AIG System Tweaks
AIG Imagers:


Andrew Marshall's original Imager design was quite straightforward: a circular neoprene ring that surrounded the tweeter and absorbed energy that would otherwise have been radiated along the speaker baffle and re-radiated milliseconds later at the listener when it reached the edge. The design was sound, and 5000 pairs were sold with not one consumer complaint or return, and a plethora of praise for the improvements wrought with normal direct radiating box speaker designs. Speakers just imaged better, with more depth! (At right you'll see the new, oval Imager)

However, it was pointed out by a major Canadian speaker designer that reflections off the inside of the Neoprene ring caused slight peaks and dips in frequency response (2 dB or less) around 10 kHz. The solution? Make the new Imagers oval in shape, so that the frequency of these minor reflections would be staggered between 8 and 12 kHz, and therefore unmeasurable in their interaction with the directly radiated sound of the tweeter

That's just what we did, made them oval in shape, but there's more! Marshall researched the available Neoprene compounds, and found one that was softer and more absorptive in the audio frequency range, and therefore did not reflect as much energy. The result is a much improved version of our famous Imagers: Imagers II. You can have the wider part of the oval at the sides, or top, it doesn't matter. Self-adhesive, Imagers II are easy to fit to your speakers, and need never be removed (if you do decide to remove them the special adhesive WILL NOT leave any kind of residue on your speakers). They'll deepen the soundstage and control side-wall reflections while also eliminating baffle diffraction.

Bafflegab? Hey, it's good physics: the more your speakers are a point source, the better they'll image. And that's the truth! Imagers II are priced at $9.99 a set, but if you order 2 sets, they're only $15, and 3 sets (which should cover a home theatre system with a spare) are only $19.99


  Reply With Quote
Old 10th May 2007, 05:48 AM   #47
Grahame is offline Grahame  United States
diyAudio Member
Grahame's Avatar
Join Date: Mar 2007
Ok, to recap - here is the PCB Crossover
Attached Images
File Type: jpg crossover_front.jpg (79.3 KB, 658 views)
  Reply With Quote
Old 10th May 2007, 05:50 AM   #48
Grahame is offline Grahame  United States
diyAudio Member
Grahame's Avatar
Join Date: Mar 2007
PCB Crossover from behind

Sticky label =
Rola Celestion Ltd

Text on PCB =
SP 2271
ISS. 1
Made in England
Attached Images
File Type: jpg crossover_back.jpg (76.5 KB, 607 views)
  Reply With Quote
Old 10th May 2007, 06:04 AM   #49
Grahame is offline Grahame  United States
diyAudio Member
Grahame's Avatar
Join Date: Mar 2007
Mirroring the back of the PCB, comparing it to the front, and some work in MS Paint gives us the following

Green = Green Capacitors C1, C2, C3, C4
Black = Black Capacitors C5, C6, C7, C8
Red = Inductors L1, L2, L3, L4, L5

Post= inputs 1-7
Tweeter = inputs 2-4
Mid = inputs 8-5
Sub = inputs 9-6

Using Alan's units/values we have


Hopefully we should get the real values for the inductors soon.

Alan - are there any issues I should be aware of when running the frequency response tests - I think it runs sweep tones / impulses via the inputs - 20-20KHz - I think - so that should be OK with the crossover in place.

If I run the sweeps directly to the drivers without the crossovers ( if I can wire it up like that) should I restrict the frequency range I subject each of the drivers to?

e.g. MD 500 says .5-5KHz on back

so I'd guess 20-500 for the sub
500-5KHz for the MD500
5KHz - 20KHz for the HF2000's

I won't try anything until I hear from you and have read the manuals on the measurement gear.
Attached Images
File Type: jpg crossover.jpg (83.3 KB, 516 views)
  Reply With Quote
Old 10th May 2007, 09:56 PM   #50
Grahame is offline Grahame  United States
diyAudio Member
Grahame's Avatar
Join Date: Mar 2007
I Have Measurements from the point to point crossover.

Thanks to those nice folks at Electronics Plus
and an LRC meter we have the following measurements of the inductors - taken when the inductors where isolated from the rest of the circuit - by means of cutting the wires

The Names are from the PCB layout - The number in brackets is the text on the outside of the inductor - and the measurement is from the LRC meter.


Alan - does this look right? - now we have all the values can we determine the crossover points + slope values?

Is there any software where I can plug in the numbers?
  Reply With Quote


celestion 66 needs Mid-rangeHide 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

Forum Jump

New To Site? Need Help?

All times are GMT. The time now is 05:47 PM.

Search Engine Optimisation provided by DragonByte SEO (Pro) - vBulletin Mods & Addons Copyright © 2017 DragonByte Technologies Ltd.
Resources saved on this page: MySQL 15.79%
vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2017 DragonByte Technologies Ltd.
Copyright ©1999-2017 diyAudio