Crossover nightmare!!!!!!!

[harnfield]

Caddock MP930 have been mentioned.
These are used by Tony Gee for his designs. Although to pay £3 is higher than £1, they are worth it.
In use with low impedance drivers that drew high current, these Caddock resistors out-performed the costly Duelund resistors.

[harnfield]

Rhopoint Components Ltd. stock the full range of values of Caddock MP930, unlike Farnell.

[alan-1-b]

Thankyou harnfield,

for the tip where one can obtain the Caddocks !

And whilst on crossovers, how is your project proceeding ?
Do please update us in the other Thread.

[moermusic]

Hi all: I hve studied both threads intensively. My project same as Lucas', except with the MK 2 printed circuit board. Have pretty much deciphered all the info I was looking for except for 2 things. What was ever decided on as the best *compromis* resister values for the mid filter? bearing in mind I am dealing with a 24 vs 33 cap and a 4 vs 6 cap. Are the 2.7 and 2.2 r's that Lucas used applicable to my xover? Should they be higher or lower?
Sorry if I missed seeing this if it exists. So many pages so many times.
Comments. Dont need to start all the whys & howcomes over
Thanks

[alan-1-b]

Are you telling us that the 2 capacitors in the midrange filter's circuit are different in your 44s than in Lucas' ?

I am aware of an original and a MkII version of the 44, but rather than me speculate, please Post the actual capacitor values in your 44s ... ?

Values for ESR compensation resistors are a Compromise.
Post whether you prefer to have the upper-midrange response prominent,
or the lower midrange response prominent for your listening ... ?

Note, this does not mean that part of the midrange has to be recessive - of lower output than the other areas of the frequency spectrum.

Describe the current balance of the frequency spectrum in your 44s,
of treble versus upper midrange versus lower midrange versus bass,
and what you would like to change in that balance ... ?

[moermusic]

Sorry Alan. I have Sept/1978 66's. the printed circuit board is labelled 44/66. The in series C 7 is 24 uf, the C 5 is 4 uf. Think I want/need a touch more in the upper mid. Posted here as I wasnt sure if anyone *monitoring* Mr. White thrd anymore.Thanks.
DM

[LucasAdamson]

Hi all,

The Mk2 (PCB crossover) Midrange caps are different in that mine were 25uF with an 8uF in parallel. I have a friend whose Mk1 (black front, p2p wired x-over) 25uF caps had no paralleled 8uF, so I guess at some point they decided to go to 25uF. Which is the better solution, however, is not so clear. I am happy with my 33uF caps.

I think you'll be very pleased you've done this. Replacing caps is not as simple as all that. The suggestion of not using ESR resistors at all was quickly put to bed at first listen - the sound was simply excruciatingly awful!

My current set-up for ESR compensating resistors is as follows:

In the treble circuit, replacing film caps with polypropylene film caps, almost no compensation is technically required. I found, however, that a 0.5 Ohm 2 watt wire-wound resistor in series after the last cap attenuated the tweeters which are now much much cleaner and sweeter and seem brighter. The 0.5 helps keep the tonal balance I feel.

In the mids, I have a 2.2 Ohm 4W wire-wound in series immediately after 33uF and a 2.7 Ohm 4W wire-wound in series immediately after 6uF to earth.

For the bass I replaced old electros with new electros, so no ESR compensation here. The cost of 4x huge polyprops is hard to justify here, especially for the bass circuits, where the sensible money would first go on new air-core inductors if anything.

This set-up works well if your amp can produce good bass. I have a cheap Onkyo amp that I originally tested the amendments to ESR with and they made the speakers sound very bass light, which is a shame because the big easy bass is a big part of the appeal with these speakers. I considered what to do with ESR to improve bass, but fortunately it resolved itself as the issue was with the bass-shy amp. My newly built amp reintroduced the bass to give a lovely balanced sound and so I recommend the values I have now. They do sound quite special now - cleaner and crisper.

I hope that helps. All praise to Alan though - not me!
Lucas

[alan-1-b]

Originally Posted by moermusic

Sorry Alan. I have Sept/1978 66's. the printed circuit board is labelled 44/66. The in series C 7 is 24 uf, the C 5 is 4 uf. Think I want/need a touch more in the upper mid. Posted here as I wasnt sure if anyone *monitoring* Mr. White thrd anymore.Thanks.
DM

Hi DM,

Celestion used the same PCB for both 44 and 66 because both use the same circuit configuration, even though some different capacitor values in each.

For maximising upper mids, use 3.9ohm in Series with the 4uF cap -{and that is likely the maximum ESR that would have been in the original cap in the upper mids}- as that will reduce the filtering roll-off of upper mids.

I would start with whatever resistors I had on-hand.
If you have a 3.3ohm then try that with the 4uF cap.
If you have a 1ohm or 1.2ohm then try either in Series with the 24uF cap,
but I would start with 1.5ohm there,
so if you are buying, buy 1.5ohm and 3.9ohm.
Perhaps buy cheap 5watt wirewounds from a local shop to try, to hear if those values give close to the frequency balance you want.
If a little too much upper mids, then reduce the 3.9 to 3.3 or even to 2.7ohm, but if you go lower than about 2.7 ohm then you may start to hear a narrow-band midrange resonance via the mid-dome.
I posted why that occurs ... some where in the 66 thread ...

To reduce mids overall, increase the 1.5ohm to 1.8 or 2.2 or 2.7 until you hear close to what you are wanting - but with 3.9ohm to the 4uF cap to retain upper mids, and then reduce to 3.3 or 2.7 there if necessary after listening.

After that, then buy a good quality 5watt non-inductive resistor for the 24uF application.
The 4uF application does not need a non-inductive BUT it does need a quality manufactured resistor so that voltage annomolies resulting from poorly constructed resistors do not couple into the mid-dome's signal.
The sound should be clearer or cleaner with the better resistors than with the cheap ones you experiment first with,
BUT, please do Post your listening results about resistors' quality - in the 66 Thread - because that is useful to other restorers.

Post also the brand and type of capacitors you are using for each filter section of the crossover.

From where you seem to be located you will likely be able to buy Mills MRA-5 non-inductive wirewounds via Mail-order easily,
and those are fine for both resistor locations.

The Mr_White , Celestion 66 thread is alive and well.
I am currently discussing other capacitors, and the eventual ESR resistor to his 25uF Series midrange cap, with a critical listener there.

After you have decided your preferred resistor values, go to Post #349 on Page 35 of that Thread and try the EXPERIMENT I described there-in.
Then tell us there what you prefer !

[alan-1-b]

Originally Posted by LucasAdamson

Hi all,

The Mk2 (PCB crossover) Midrange caps are different in that mine were 25uF with an 8uF in parallel. I have a friend whose Mk1 (black front, p2p wired x-over) 25uF caps had no paralleled 8uF, so I guess at some point they decided to go to 25uF. Which is the better solution, however, is not so clear. I am happy with my 33uF caps.

************************************************** ***********

In the treble circuit, replacing film caps with polypropylene film caps, almost no compensation is technically required. I found, however, that a 0.5 Ohm 2 watt wire-wound resistor in series after the last cap attenuated the tweeters which are now much much cleaner and sweeter and seem brighter. The 0.5 helps keep the tonal balance I feel.

In the mids, I have a 2.2 Ohm 4W wire-wound in series immediately after 33uF and a 2.7 Ohm 4W wire-wound in series immediately after 6uF to earth.

************************************************** ***********

For the bass I replaced old electros with new electros, so no ESR compensation here. The cost of 4x huge polyprops is hard to justify here, especially for the bass circuits, where the sensible money would first go on new air-core inductors if anything.

************************************************** ***********

This set-up works well if your amp can produce good bass. I have a cheap Onkyo amp that I originally tested the amendments to ESR with and they made the speakers sound very bass light, which is a shame because the big easy bass is a big part of the appeal with these speakers. I considered what to do with ESR to improve bass, but fortunately it resolved itself as the issue was with the bass-shy amp. My newly built amp reintroduced the bass to give a lovely balanced sound and so I recommend the values I have now. They do sound quite special now - cleaner and crisper.

************************************************** ***********

I hope that helps. All praise to Alan though - not me!
Lucas

Hi Lucas,

If your friend's board has still the old capacitors, look very closely to see if any evidence of a parallel connected cap having been de-soldered from the original 24uF cap.

Another possibility:-

take a midrange cone driver out of his and compare it visually to one of yours.

If the cones themselves are visually different, or with different termination material surrounds, then a different value capacitor may have been required for the correct cross-over frequency,
and/or,
if the sizes of the magnets on the back of the drivers' chassis are different, then almost certainly a different cap value would have been required.

But, if his mid-drivers are identical to yours he can try 33uf there if he wants to, but if that worries him then simply use a new 25uF cap there, or use a 27uF cap for a least a little additional low mids' weight, unless he prefers the 25uF sound.

Post and tell us about the actual midrange driver in each version, as this will be useful to other restorers.

************************************************** ***********

The "seem brighter" -{in addition to "cleaner and sweeter"}- from the 0.5ohm resistor terminated tweeters is likely a psychoacoustic effect resulting from closer matching of the transient response of the tweeter to the midrange as result of the resistor ... and possibly better displaying the quality of the capacitors you used.

The "4W" wirewounds you have installed in the mids, are those Welwyn's WP series ?
or another Welwyn series ?
or another brand ?

************************************************** ***********

Electros in the bass filter cause part of the bass and low-mids' sound, and if you like that then it is fine.
I have discussed this with dloper in the 30s' pages of the Celestion 66 thread.

The types, and the DCRs, of the inductors also cause part of the bass sound,
thus if you decide to spend significant money to try to get another specific sound in the bass later Post here if you want to discuss possibilities.

If you currently have low resistance in the cored inductors, but buy air-core inductors of higher resistance the sound will change but perhaps not in the way you may want.

The old cored inductors are fine for low to medium volume levels,
but will cause a noticeable change of sound if driven harder.

************************************************** ***********

Regardless of what Onkyo may achieve with some of their premium models of amplifiers, their cheap models are not usually wonderful, as you can hear !

For anyone wanting reasonably good quality sound bass extension from a low priced amplifier, try the basic Rotel current or recent models.

Given the general Rotel sound of recent years, though it is a bit different between models, I think they are a suitable minimum amplifier to use with the classic Celestion 44s and 66s, if one cannot build or afford substantially better.

Some past NAD cheap models were OK in the bass, and perhaps their current model is, however I have not heard a sample.

Lucas's description of the differences in bass quality when using different amplifiers is very good - particually as relevant for these loudspeakers.

************************************************** ***********

Thankyou for your compliment Lucas !
and, your reports of changes to the sound and the descriptions of your modifications are also worthy of note because this is one of the ways we learn things.

With Celestion 44, the "EXPERIMENT" I describe in Post #349 on Page 35 of the Celestion 66 midrange Thread can also be tried.

[LucasAdamson]

Originally Posted by alan-1-b

Hi Lucas,

If your friend's board has still the old capacitors, look very closely to see if any evidence of a parallel connected cap having been de-soldered from the original 24uF cap.

No 8uF removed. They came like that. I've seen other 44 x-overs that were a single 25uF too. My 8uF parallel was factory fitted too, as it is the same brand as other 72uFs fitted elsewhere in the circuits.

Another possibility:-

take a midrange cone driver out of his and compare it visually to one of yours.

If the cones themselves are visually different, or with different termination material surrounds, then a different value capacitor may have been required for the correct cross-over frequency,
and/or,
if the sizes of the magnets on the back of the drivers' chassis are different, then almost certainly a different cap value would have been required.

But, if his mid-drivers are identical to yours he can try 33uf there if he wants to, but if that worries him then simply use a new 25uF cap there, or use a 27uF cap for a least a little additional low mids' weight, unless he prefers the 25uF sound.

Post and tell us about the actual midrange driver in each version, as this will be useful to other restorers.

I won't be able to remove them to compare, but they certainly look identical. they may have different magnet sizes - it's possible. I reckon that they just responded to reviews or feedback, and changed caps in response to give a different sound.

The "seem brighter" -{in addition to "cleaner and sweeter"}- from the 0.5ohm resistor terminated tweeters is likely a psychoacoustic effect resulting from closer matching of the transient response of the tweeter to the midrange as result of the resistor ... and possibly better displaying the quality of the capacitors you used.

I meant that they seemed brighter and cleaner/sweeter than the old original caps. The resistor did nothing I could hear but attenuate them a touch.

The "4W" wirewounds you have installed in the mids, are those Welwyn's WP series ?
or another Welwyn series ?
or another brand ?

They're the Welwyn WP series, as you recommended. My previously stated preference for MOX resistors was just nonsense from me in response to multiple alterations and my making assumptions about what caused what effect. The Welwyns are good.

Regardless of what Onkyo may achieve with some of their premium models of amplifiers, their cheap models are not usually wonderful, as you can hear !

For anyone wanting reasonably good quality sound bass extension from a low priced amplifier, try the basic Rotel current or recent models.

Given the general Rotel sound of recent years, though it is a bit different between models, I think they are a suitable minimum amplifier to use with the classic Celestion 44s and 66s, if one cannot build or afford substantially better.

Some past NAD cheap models were OK in the bass, and perhaps their current model is, however I have not heard a sample.

Lucas's description of the differences in bass quality when using different amplifiers is very good - particually as relevant for these loudspeakers.

I'd certainly never buy an Onkyo again, that's for sure - this one is 20 years old - I was a teenager! In fact, I'd only ever recommend a Pass Labs B1 pre and F5 power combination these days. It seems that for about £150 you can make an amp that is almost flawless. I have the B1 already. It is superb! The F5 is my next project for January 2010

Thankyou for your compliment Lucas !
and, your reports of changes to the sound and the descriptions of your modifications are also worthy of note because this is one of the ways we learn things.

With Celestion 44, the "EXPERIMENT" I describe in Post #349 on Page 35 of the Celestion 66 midrange Thread can also be tried.

I'll check that out. Thanks Alan. As always, your commitment is compelling!
Lucas