Custom Celestion Ditton project - need expert opinions

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Hi Allan regarding to your questions,

I replace old caps with modern ones and used the same values.

You are right mundorf doesnt do 50uf - I placed there 47uf
as 6uf I used 2.7+3.3uf
and as 16 = 15+1uf

all caps are Mundorf MKP Mcap

Oryginal sound on old crossovers is very lush with beautiful midrange and sound is very wide. after replacing sensation is like tweeter and midrange volumes went up and level of bass is lacking now, they sound good in terms of detail ,there is no distortion or anything like that , just balance between drivers is messed up a bit, I would say that tweeter is the loudest , midrange got some extra output and bass is behind everything.

drivers measures: ( 0.7 resistance of my meeter not included))
tweeters - 1) 4.3 ohm 2) 4.8ohm
midrange- 1) 6 ohm 2) 7.3 ohm
bass - 1) 6 ohm 2) 6ohm

resistors got aprox 12.7 and 41ohm

Regards
 
Leak sandwich 600

The original Leak sandwich was a 2-way , and later a 3-way 600 was developed.
If your 600 has a small dome tweeter with a small flat plate covering the centre of the tweeter's dome , then you have an original of the 3-way.
If the tweeter is different to the above , you may have a user-modified version of the original 2-way.

The Resistances you found for the tweeter are close enough to the same for normal production samples' variations for that era ,
however it is the mid-cones' resistances that indicate a possible problem.

There are 3 problems causing the sound you have now , but first we need to find if the mid-cones are both working well.

To test the mid-cones:

Disconnect the crossover from the back-panel amplifier terminals.
Disconnect the mid-cones from the crossover.
Connect the mid-cones directly to the back-panel amplifier terminals
with NO crossover components , and NO other drivers connected .
The mid-cones only are to be connected to the amplifier.

Test carefully as follows so that the mid-cones are not overdriven and damaged.

Use music with treble and midrange , but no bass instruments.
Best is solo Violin music , such as Bach's Partitas and Sonatas for solo Violin.
That is what I use to test small mid-cones { and tweeters }.
You can use solo Cello music , such as Bach's Suites for Cello { violoncello } if you keep the volume level fairly low.
Other Composers' violin and cello music will work , and even if some Harpsichord accompaniment if you keep the Volume fairly low.
Solo Piano music will work if it is lightly played , and does not feature the bass notes of the piano loudly.
Gut or Nylon string guitar music will work if not played too loudly.

If you do not have any Classical Music CDs , then Folk Music or light Pop Music is OK if:
Solo Acoustic guitar , or Flute , or Acoustic Guitar with Female vocal , or high-pitched Male vocal { like a Boy's voice }.
NOT heavy deep Male voice , and not heavy Electric guitar with distortion ,
because these will overdrive the mid-cones and damage them.
Electric guitar is OK only if clean sound and trebley , NOT if lots of Low-midrange because that small mid-cone will not play the full sound of it.
That is , you will not hear it but it will be still driving the cone hard , and that can damage it.
Not Electric keyboards , because the sound is too dense , { unless it is only trebley tones }.
Not Band music , and no Drums.
Only Solo or Duo or Trio groups if all upper midrange and treble sounds.

First try each cone separately - that is one at a time.
Start with low volume and listen carefully.
If no distortion then turn up the volume a little.
Listen for distortion , or buzzing sounds , or scraping sounds as the cone moves.

If both cones work OK separately , then listen with both playing together.
Are both the same volume ?
Is one a bit anemic in sound ?
Is one a bit too resonant in sound ?
Is one more treble than the other ?
Is one more low-midrange than the other ?

If you are listening to Stereo music , remember that one Channel may have a different Tonal Spectrum to the other channel.
If that is , then swap the two speakers around , each to the other channel , to compare again.
Better is to use a Mono recording , or a Stereo recording that has a strong Centre image , and not a lot of things in the extreme Left and Right channels only.

Identify any different sound of one cone to the other ,
and post a report here with Identifiers' Midrange 6 and Midrange 7.3 , so I can know which is causing which sound.
Depending on the type of difference it might be possible to get each to sound the same as the other , but to do that I need to know which is which.

As it is with 6 ohms to one crossover and 7.3 ohms to the other crossover
there will be an audible problem , so one will need another resistor ,
which I will decide after you report the sounds.
 
Hi Alan

Thank you for your response

model of leak 600 i got look exactly like those:

http://images8.fotosik.pl/2619/abb7e237ecd43cfc.jpg

I did a test today as you have suggested ,played music directly from amp to middriver

I have 2 pairs of those mid drivers. I had to read something wrong last time, I measured them again and thats what they reads:

1st pair measure: 6.3oh and 5.8ohm
2nd pair: 6.2ohm and 6.1ohm

all those measures includes 0.6ohm resistance of digital meeter cable.

to my ears pair which measure very close to each other sounds most similar and i didnt notice any distortion , sometimes small resonants but that happened to all those drivers.
 
One Mistake in Schematic for Leak 600 , plus sound's problems.

'ullo Mikolaj ,

I received your message.
I get very little computer access time , and have so many things currently to do
that it has taken me till now to get back here.

There seems to be at least 1 mistake in the Schematic you posted in #119 on Page 12 for your Leak 600 crossover.
Where you have the 39 ohm resistor connected in the midrange filter
it is not likely to be connected at that junction of the 16uF cap and the 2 inductors ,
because that would reduce the effects of both the inductors.
It is likely to be connected at the other end of the 1mH inductor
so that it is effectively in Parallel only with the mid-cone and the 6uF cap ,
same as the 12 ohm resistor is in the tweeter filter.
I have used this technique myself. It is to electrically dampen the high Impedance peak at the Fs of the mid-cone.
Have another look at the crossover to check.
If you have not taken these resistors out , then they should be in the correct electrical positions where the Leak designer wanted them.


Your Leak 600 , as in the photo you posted , has the original tweeter ,
so let's get it sounding right , and you may find you do not need to change the tweeters.

The problems with the sound in the midrange and treble are because the new poly caps have no internal resistance ,
AND in the places where you have connected caps in Parallel there is now a distortion of the signal's wave-form
because the caps in each Parallel pair have different Pulse Rise times.
There will be very little difference with the 2.7uF//3.3uF pair , because both those caps are the same type and the same length ,
though two of 3uF would give slightly better result , and one of 6uF or 6.2uF would give the best.
A single capacitor usually causes better sound to a tweeter than a Parallel pair ,
but for now leave that caps' pair there and connect your 2.2ohm resistors ,
one for each crossover ,
in Series with that pair of caps.
That should bring the treble down to correct level to match the bass , or fairly close to.

In the midrange filter the 1uF cap is causing a distortion of the Transient Response of the music signal.
Take it out , and use the 15uF cap only.
The difference there between 15uF and 16uF is very small.
Later , after you decide the correct resistor value by listening ,
you can buy a 16uF cap or a pair of 8.2uF caps if you want to ,
but it may not be necessary.

For initial listening put 1.8 ohm in Series with the 15uF cap.
Direct electrical Series with the cap , to simulate the effect of a resistance inside a capacitor.
It does not matter which end of the cap you connect it to , but use the same end for both crossovers.
This resistance should reduce the level of midrange to same or close to that of the bass ,
but final decision will have to be made by listening.

In the midrange filter also ,
connect 2.7 ohms in direct electrical Series with the Parallel pair of caps that sum to 6uF so that it simulates a resistance inside a capacitor.
As for the other resistors , it does not matter which end of the caps' pair you connect the resistor at ,
but do it at same end for both crossovers.
This resistor reduces the onset and magnitude of resonance in the upper midrange.

In the woofer filter , connect 1.8 ohms or 2 ohms in direct electrical Series with the 47uF cap.
This resistor reduces the onset and magnitude of resonance in the lower midrange.
The 600 design uses the woofer for the low mids quite well ,
because that is a very good design of woofer ,
so put the resistor in to improve low midrange quality of sound there.
It will NOT reduce level of bass - that will stay the same as now.

If you want to read more about the theory of the above ,
go to the Celestion 66 needs Midrange thread , to Page 8 , #74 ,
and start reading from there.
There is more on later pages , and possibly some later corrections for if I made any mistakes that I found
when I next logged in and re-read my previous posts -{ I usually later check my earlier posts }.

The resistance values I stated above are estimates.
I cannot guarantee they will be exactly right.
You will have to listen to hear if Tone is close to what you want ,
and if outputs from each driver balance.


Alternate values of resistance:

Buy 1.5ohm as an alternate for the 1.8 in the midrange in case output is now a bit too low there with 1.8 .

Buy 3.9 ohm to put into the midrange of your Celestion crossover where you now have the 3.3 ohm.
You can then try the 3.3 ohm in the Leak where the 2.7 ohm is AFTER all other resistances have been decided by ear.
This location is a fine-tuning point for integration between mids and treble.
It will not be less than 2.7 ohm , but it might need 3 ohm or 3.3 ohm.

Celestion 66 listeners have found 3.9 ohm works best there when modern dome tweeters are used ,
though 2.7 ohm there may be better when the original Celestion HF2000 is used ,
because there is a difference in the electrical Phase Angle at the low end of the HF2000 than is with most modern dome tweeters ,
thus it needs a different resistance in the upper midrange filter for optimum integration.


I recommend you buy Mills MRA-05 wirewounds for all the resistors ,
and if you choose those then buy their 2 ohm for the woofer filter instead of 1.8 ohm ,
because I think that will cause slightly better result.

If you do not want to spend Mills' prices during this deciding stage ,
then buy the lower priced Mundorf M-Resist in their 5 watt series ,
or buy the 5 watt versions of those Superes resistors.
You do not need the large 10 watt resistors for these loudspeakers.

When you install the resistors , connect one end to the circuit board where you take out the relevant capacitor's lead ,
then position the body of the resistor so that it does not touch anything.
These resistors need to be placed so there is free air circulation around them to keep them cool.
They will be carrying higher currents than the two original resistors in their circuit locations of that crossover ,
so do not put the new resistors close to the surface of the board.

With one end of each capacitor then not fixed to the board , the capacitors will need to be secured
so that their weight does not cause movement and subsequent breakage of the other end of their leads where those are fixed to the board.
Buy some Nylon cable ties - Nylon not Plastic - because Plastic will stretch over time and allow movement.
Drill small holes through the board , carefully at places where the drill will not go into the metal circuit tracks under the board.
Tie the capacitors tightly to the board through these hole with the Nylon cable ties.
Most Electronic Components sellers also sell Nylon cable ties.

When you have all the capacitors disconnected at one end to put the resistors in ,
you can listen-test the mid-cones through upper mids to treble with no tweeters , { but with woofers connected }.
To hear that , connect the 1.8 ohm resistor and 15uF cap back into the circuit but leave BOTH pairs of 2.7uF//3.3uF caps disconnected at one end ,
so that no signal can flow to tweeter , and no signal to Ground past the mid-cone ,
AND connect a short piece of wire in Parallel with the 1mH inductor in the mids filter
to bypass signal around it so all the treble will go to the mid-cone.
This may assist you to better evaluate all 4 of your mid-cones
because they will not be distorted by bass signals then , so can be played to higher volume levels.
The one that measures 6.3 ohm may be a better match to one of the other 6+ ohms ones than they are to each other ,
and the 5.8 ohm may match close to the 6.1 ohm.

Also , while you have things disconnected , disconnect one end of both tweeters and measure their DC resistance again.
As you found a mistake when you second time measure the mid-cones ,
you may find another mistake when you re-measure the tweeters.

Try all the above , then post listening results here , and I will advise further.


Which tweeter have you decided for the Celestions ?
 
Last edited:
Hi Alan

Thanks for all the expert advice posted here.

I have both Ditton 44s and 66s and I note that both the midrange unit and the filter is slightly different in each, although the topology is the same.

For the 66 the high side of the filter has a 4uF cap whilst for the 44 it's 6uF in the same position. Do you advocate using a 2R7 series resistor with the replacement for both of these, or should the value of the resistor be different depending on the capacitor value? I notice that you recommend up to 3R9 here for the 66 when using a replacement tweeter.

I've also found that when running a pair of 44s, one modified and one not, the unmodified one sounds slightly louder in the mids. Since I'm using 1R8 in series with 24uF for the low side of the midrange filter, I'm assuming I would be better off using 1R5 there instead?

Finally, am I correct in concluding that you are advocating 27R in parallel with the Seas TTF1 and 47R in parallel with the Hiquophon OW1 with no series resistor in either case, or have I got confused somewhere?

All resistors are Welwyn WP4S and Caps are Clarity Cap PWAs (3 x 24uF) for the bass and PX elsewhere (3.9uF and 12uF for the tweeter). I'm currently running the 44s with the standard tweeter and modified crossovers and although I'm pleased with the results I'd like to do even better.

Thanks.
 
Hello there.
I have all the drivers from the ditton 66 and have recreated the crossovers. but without the enclosures. I am trying to recreate them.. does anybody have the baffle dimensions, ie the spacing between driver's handy ???? It would be very much appreciated as I can't find it anywhere and have been after the dimensions for some years now. many thanks..
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.