compromises
There are some misbehaviors in those Celestion drivers.
In the 12" woofer is audible resonance in the lower midrange, though I do not know the exact frequency region.
Resonances in its upper midrange are mostly filtered by the crossover.
There is a resonance at the low end of the mid-dome, but cleverly damped by Celestion in that design, thus it will be interesting to see how well, if the PRAXIS measurements display its extent.
There will be some breakup at the treble end of the mid-dome, but it may be all above the crossover point.
You have changed your tweeters, but if you still have the original ones and measure them you will see at least one resonance in their audio band, but it is not a bad one for drivers of that era.
Notch filters often cause audible problems of similar magnitude to the other anomalies they are intended to correct, thus one has to compromise a bit here, and not perfect but reasonabaly well behaved drivers usually sound better with no notch filters, at least when used in the fairly well-behaved parts of their bandwidths.
The effect of the cabinet lip is audible, thus it will be interesting to see the degree of anomaly that relative measurements show of with versus without lip.
You may find that the characteristics of those electrolytic caps were taken into account when the crossover was designed and measured, and thus it will likely have crossed how and where it was intended to, {within the limits of the Specification for it}, when the caps were new. Celestion's engineers did understand those things, and how to compromise, etc ...
There are some misbehaviors in those Celestion drivers.
In the 12" woofer is audible resonance in the lower midrange, though I do not know the exact frequency region.
Resonances in its upper midrange are mostly filtered by the crossover.
There is a resonance at the low end of the mid-dome, but cleverly damped by Celestion in that design, thus it will be interesting to see how well, if the PRAXIS measurements display its extent.
There will be some breakup at the treble end of the mid-dome, but it may be all above the crossover point.
You have changed your tweeters, but if you still have the original ones and measure them you will see at least one resonance in their audio band, but it is not a bad one for drivers of that era.
Notch filters often cause audible problems of similar magnitude to the other anomalies they are intended to correct, thus one has to compromise a bit here, and not perfect but reasonabaly well behaved drivers usually sound better with no notch filters, at least when used in the fairly well-behaved parts of their bandwidths.
The effect of the cabinet lip is audible, thus it will be interesting to see the degree of anomaly that relative measurements show of with versus without lip.
You may find that the characteristics of those electrolytic caps were taken into account when the crossover was designed and measured, and thus it will likely have crossed how and where it was intended to, {within the limits of the Specification for it}, when the caps were new. Celestion's engineers did understand those things, and how to compromise, etc ...
Hi Alan - I kept the layout of the inductors much as they were before to try and avoid changing any magnetic interaction between them. If anything, they are all slightly further from the caps.
The 'thickening' I've noticed does indeed seem to emanate from the bass driver. I'm happy to replace the pairs of bass caps with the original values if this will help.
Stuart
Modeling... I'm curious about how it will turn out, too. I have early 1970's era JBLs --a 10" three-way and a 10" two-way-- that benefited greatly from redesigned crossovers (done by Troels Gravesen in Denmark). These speakers went from being unlistenable to being very enjoyable and non-fatiguing.
The Ditton 66s have halfway decent crossovers to begin with, so I suspect that the improvements will not be as great as with the old JBLs...but it will be interesting to watch a redesign of these.
For those of you here that don't yet have measuring capability but would like to make some audible improvements, I highly recommend the purchase of a microphone and phantom power supply. These can then be connected to most any computer that's running RTA software ( I use a free RTA software program ). You'll then be able to graphically see any changes made in speaker placement, EQing, etc. It can also help to identify various driver problems.
mic--
Parts-Express.com:*Behringer ECM8000 Measurement Microphone | measurement mic measurement microphone mic microphone RTA RTA mic RTA microphone special-11-2-behr
power supply--
Parts-Express.com:*Behringer MIC100 Tube Ultragain Tube Preamp with Limiter | Behringer MIC100 Tube Preamp Ultragain mic preamp microphone preamp NewBEHR08
free RTA software--
jDFT - Download
pink noise-- download from the internet
It's the very best $80 that I've spent on audio!
Here are a couple of RTA screen shots, with and without EQ applied. (My MAC 4100 receiver has a simple built-in equalizer ( +/- 12db at 30, 150, 500, 1500, and 10k Hz .) The EQ in the graph below has a boost at 1500 and at 10k.
The Ditton 66s have halfway decent crossovers to begin with, so I suspect that the improvements will not be as great as with the old JBLs...but it will be interesting to watch a redesign of these.
For those of you here that don't yet have measuring capability but would like to make some audible improvements, I highly recommend the purchase of a microphone and phantom power supply. These can then be connected to most any computer that's running RTA software ( I use a free RTA software program ). You'll then be able to graphically see any changes made in speaker placement, EQing, etc. It can also help to identify various driver problems.
mic--
Parts-Express.com:*Behringer ECM8000 Measurement Microphone | measurement mic measurement microphone mic microphone RTA RTA mic RTA microphone special-11-2-behr
power supply--
Parts-Express.com:*Behringer MIC100 Tube Ultragain Tube Preamp with Limiter | Behringer MIC100 Tube Preamp Ultragain mic preamp microphone preamp NewBEHR08
free RTA software--
jDFT - Download
pink noise-- download from the internet
It's the very best $80 that I've spent on audio!
Here are a couple of RTA screen shots, with and without EQ applied. (My MAC 4100 receiver has a simple built-in equalizer ( +/- 12db at 30, 150, 500, 1500, and 10k Hz .) The EQ in the graph below has a boost at 1500 and at 10k.
Vertical axis of measurement ?
Hi sba,
its great that you are still with us !
Thankyou for Posting the equipment list - hopefully it will be of use to some here.
The two plots you Posted - what vertical axis did you have the mic on:-
Centre of tweeter ? -{though I think that was unlikely}
Midway between tweeter and mid-dome ?
Centre of mid-dome ?
Midway between mid-dome and woofer ?
Another axis ?
And which of your 66s ? - Blacky or Woody ?
I ask, because of the different mid-dome in each.
Did you use Solen caps for your JBL mods' crossover, or one of the Jantzen types that T.Gravesen uses ?
From Posts in several places on the web, and including at least one recent user in this forum, it seems the recently introduced Jantzen "cross caps" sound good, thus recommendable depending on local price versus whatever else in one's buying area.
********************************************************
Stuart,
I'll reply to yours when I have time to do so at the length it will require.
Hi sba,
its great that you are still with us !
Thankyou for Posting the equipment list - hopefully it will be of use to some here.
The two plots you Posted - what vertical axis did you have the mic on:-
Centre of tweeter ? -{though I think that was unlikely}
Midway between tweeter and mid-dome ?
Centre of mid-dome ?
Midway between mid-dome and woofer ?
Another axis ?
And which of your 66s ? - Blacky or Woody ?
I ask, because of the different mid-dome in each.
Did you use Solen caps for your JBL mods' crossover, or one of the Jantzen types that T.Gravesen uses ?
From Posts in several places on the web, and including at least one recent user in this forum, it seems the recently introduced Jantzen "cross caps" sound good, thus recommendable depending on local price versus whatever else in one's buying area.
********************************************************
Stuart,
I'll reply to yours when I have time to do so at the length it will require.
Inductive coupling and effects on the midrange
Hi Stuart,
If you haven't seen already, then look at Page 5, Post#47, for a photo of the last 66 series' crossover boards.
On Page 10, Posts' #96 & #97 photos are the earliest boards, and #98 & #99 are the mid period boards.
Which are yours ?
Notice that the spacing between the inductors is a bit different for each period, and the final period was the worst,
but all periods' boards have the larger of the mid-filter's inductor close to the largest bass inductor.
Fortunately the treble filter's inductor is sufficiently distant from all the others to be almost not coupled - so don't worry about that one.
Mostly what I don't like is the various amounts of coupling between the mids' coils and the bass coils.
Is your new board large enough to allow placing of any of the four largest coils differently ?
The old axial capacitors would all have had some inductance, {as do most new axial electrolytic caps, but not most new axial poly caps}, but see how they were placed - their axis at 90 degrees to the large coils - thus vey little inductive coupling into the caps.
Are your new caps, especially the 80uF electros, aligned in the same way as on the old boards ?
Basically, if axial caps are lying flat on the boards there should be very minimal inductive coupling because their windings will not detect magnetic fields which are at 90 degrees all the possible field lines through them.
The caps field lines are all horizontal.
The coils field lines are all vertical.
{except for a very small amount of field at each end of the caps and over top and bottom of the coils}
If you have not turned any coils up-side down then you will not have changed any degree of inter-coil coupling, and it is better not to.
There is some lower midrange resonance in the woofers, and now with the clearer sound to the mid-dome and tweeter through the new caps that resonance lower in the midrange will be a bit more audibly obvious than previously.
It cannot be entirely removed, but it can be reduced a little.
I will explain how next time as I have to go now,
but do Post which Photo # is closest to your boards, and how much extra spacing you have done.
Hi Stuart,
If you haven't seen already, then look at Page 5, Post#47, for a photo of the last 66 series' crossover boards.
On Page 10, Posts' #96 & #97 photos are the earliest boards, and #98 & #99 are the mid period boards.
Which are yours ?
Notice that the spacing between the inductors is a bit different for each period, and the final period was the worst,
but all periods' boards have the larger of the mid-filter's inductor close to the largest bass inductor.
Fortunately the treble filter's inductor is sufficiently distant from all the others to be almost not coupled - so don't worry about that one.
Mostly what I don't like is the various amounts of coupling between the mids' coils and the bass coils.
Is your new board large enough to allow placing of any of the four largest coils differently ?
The old axial capacitors would all have had some inductance, {as do most new axial electrolytic caps, but not most new axial poly caps}, but see how they were placed - their axis at 90 degrees to the large coils - thus vey little inductive coupling into the caps.
Are your new caps, especially the 80uF electros, aligned in the same way as on the old boards ?
Basically, if axial caps are lying flat on the boards there should be very minimal inductive coupling because their windings will not detect magnetic fields which are at 90 degrees all the possible field lines through them.
The caps field lines are all horizontal.
The coils field lines are all vertical.
{except for a very small amount of field at each end of the caps and over top and bottom of the coils}
If you have not turned any coils up-side down then you will not have changed any degree of inter-coil coupling, and it is better not to.
There is some lower midrange resonance in the woofers, and now with the clearer sound to the mid-dome and tweeter through the new caps that resonance lower in the midrange will be a bit more audibly obvious than previously.
It cannot be entirely removed, but it can be reduced a little.
I will explain how next time as I have to go now,
but do Post which Photo # is closest to your boards, and how much extra spacing you have done.
audible room modes' effects
I was half asleep yesterday, and re-reading my then Post now I see my choice of words for some of it is not optimum,
but I think most readers will understand what I meant there.
If not, then do ask !
And now to continue ...
All rooms have standing waves at various audio frequencies relevant to the Height; Width; Length of the room.
These are usually more audible in the bass, but can be significantly audible in the lower midrange, and in particular the standing waves between the ceiling and floor because those are in most rooms the shortest wave-lengths, and the least damped or dispersed.
Thus it is possible Stuart, given your description of audibility at different heights, that there may be lower midrange amplification by your room at one height and not at another in it.
With the Dynaudio Contour 1.8 loudspeakers do you hear any significant lower midrange thickening similar at the same height as you did for the Celestions, and not at the higher axis where it was not with the Celestions ?
I do not remember if I've heard Dynaudio 1.8, but looking at a photo I see the two bass-mid drivers in each cabinet -{if I've found the right model}- and the combined effect of those radiating from different heights may in part counter-act any resonant build-up in your room, but do comment on this because no change of capacitors can reduce a room caused effect.
Is there any similar audible effect from your KEF 104 loudspeakers in the same room position ?
*********************************************************************
Hah, looking again for a photo of the earlier Contour 1.8, what do I find but your Thread in hifiwigwam !
I doubt your Classe electronics will be the cause of any major audible problem.
Dynaudio loudspeakers in general do reveal any brashness in recordings,
and to me it seems that is result of the forward upper-midrange in Dynaudio's mid-bass cone drivers of that size,
and can also be result if microphonic caps are used with that very revealing Dynaudio tweeter.
I do not know the sound of AE509 loudspeakers, thus would have to guess from the types of drivers used,
and some other aspects of their design.
I was half asleep yesterday, and re-reading my then Post now I see my choice of words for some of it is not optimum,
but I think most readers will understand what I meant there.
If not, then do ask !
And now to continue ...
All rooms have standing waves at various audio frequencies relevant to the Height; Width; Length of the room.
These are usually more audible in the bass, but can be significantly audible in the lower midrange, and in particular the standing waves between the ceiling and floor because those are in most rooms the shortest wave-lengths, and the least damped or dispersed.
Thus it is possible Stuart, given your description of audibility at different heights, that there may be lower midrange amplification by your room at one height and not at another in it.
With the Dynaudio Contour 1.8 loudspeakers do you hear any significant lower midrange thickening similar at the same height as you did for the Celestions, and not at the higher axis where it was not with the Celestions ?
I do not remember if I've heard Dynaudio 1.8, but looking at a photo I see the two bass-mid drivers in each cabinet -{if I've found the right model}- and the combined effect of those radiating from different heights may in part counter-act any resonant build-up in your room, but do comment on this because no change of capacitors can reduce a room caused effect.
Is there any similar audible effect from your KEF 104 loudspeakers in the same room position ?
*********************************************************************
Hah, looking again for a photo of the earlier Contour 1.8, what do I find but your Thread in hifiwigwam !
I doubt your Classe electronics will be the cause of any major audible problem.
Dynaudio loudspeakers in general do reveal any brashness in recordings,
and to me it seems that is result of the forward upper-midrange in Dynaudio's mid-bass cone drivers of that size,
and can also be result if microphonic caps are used with that very revealing Dynaudio tweeter.
I do not know the sound of AE509 loudspeakers, thus would have to guess from the types of drivers used,
and some other aspects of their design.
Last edited:
Hi Alan - my boards were originally as per the JPG on page 10, Post 99, although my 66s are 'blackies'. I kept the component layout close to original, and the spacing and positioning of the coils is much the same. The caps are oriented the same as previously but (apart from the pairs of 80uF) I moved them around 2 -3 cm towards the edge of my new, slightly larger board, away from the coils.
My Contour 1.8 mk1's are a two way design, the bottom 'driver' is a passive radiator. I don't notice the 'thickening' I described with the 66's but the tonal balance of these speakers in my room is very lightweight in comparison. They do give more of a sense of depth and space to recordings, and the treble is generally better, apart from a slight sibilance. Overall, they sound too mid-range biased for my taste. There are certainly some aspects of the sound which better the 66's, but there is not one recording where I prefer listening to the Contours over the 66's. If I could take some of their open character and apply it to my 66's I'd be on to a winner.
Going back to the Kef 104's, I do notice the same 'thickening' again. It seems to be the warmer sounding speakers that provoke it, as my AE 509's don't.
I've kept the speaker position exactly the same through all the mods I've made. I'll experiment with placement as well. I know this can make a big difference in my room. Perhaps the x-over mods have also changed the optimum position?
Last edited:
replies, and an experiment
Hi Stuart,
With that version of the 66 board layout the two closest magnetically coupled inductors are the 3.5mH input to the bass filter which is dropping a lot of 500Hz and above and those frequencies will couple into that close 2.2mH in the mids' filter which is highish impedance from about 500Hz and up, thus will pick up those unwanted frequencies from the bass filter's 3.5mH, but the 27uF cap in mids' filter is lowish impedance for those frequencies compared to the next section of the mids' filter and the mid-dome, thus perhaps sufficient of the unwanted lows is grounded through the input section of the mids' filter to not cause too much problem, but there will be a small amount of intermodulation with the same frequency band presented electrically from the amplifier through the 27uF cap, though with its phase shifted versus the magnetically coupled component therefore I think likely more-so mutual cancellation of that frequency band than reinforcement.
I hope the reason Celestion engineers placed the inductors in that particular layout was for the inevitable magnetic couplings to do the least harm, and that seems to be the case within my current thinking about the relative functions of each inductor there.
I've never measured to confirm my thinking, because I never place inductors in ways that will significantly magnetically couple !
As you have made your own boards, if sufficient space move the 3.5mH coil further into the corner - further away from both the 2.2mH coils - and you should get a slightly clearer mids' sound - but don't move either 2.2mH coil any closer to the two smaller coils, nor those closer to each other.
Mids and treble caps slightly further from the coils, as you've said you've done, is good so long as the horizontal axis for each is still in the direction of the originals.
*****************************************************
I've definitely not heard Dynaudio Contour 1.8Mk1, because I'd remember hearing an ABR design even if its weighted differently to the Celestion and KEF designs, as all ABRs have part-similar characteristic sound.
I've heard a mid 80s' era Monitor Audio ABR design in similar cabinet shape to that 1.8 , and another brand I've forgotten in a smaller cabinet.
All Dynaudios I've heard have a prominent midrange - it is caused by the frequency characteristic of their bass-mid cone, particually in the upper mids.
As you don't hear the lower mids' thickening with Contour 1.8 -{with their single midrange source}- it is reasonable to presume your room is not the main cause of such.
The slight sibilance could be in part from the upper mids of the cone - what is the specified crossover frequency ?
Sibilance from that tweeter is often reported, and the reason given by a few users is microphonic capacitors in the tweeter's filter.
Have a look at the Brand, and type of capacitor(s) in both the tweeter circuit and in the bass-mids' circuit, and Post here ... ?
Look also and Post the specific model number of the tweeter if it is printed on the back of its magnet or chassis - there were at least 3 versions of that tweeter, and all look much the same from the front.
EXPERIMENT:-
As for the Dynaudios' "open character" - try reversing the Polarity of the tweeter connections to the crossover in the Celestion 66s.
Reverse it at either the board output OR the tweeter terminals - whichever is easier, though at the board output is safer because there one is not applying a hot soldering iron to tweeter leads which if prolonged can conduct sufficient heat into the voice-coil to melt a part of it.
The filter to the Celestion tweeter is 3rd Order.
3rd Order crossovers can be used in Same Polarity or in Reverse Polarity with respect to the two drivers.
Celestion have used Reverse Polarity of tweeter versus mid-dome.
Some listeners prefer the audio character of such, but others, including myself, prefer Same Polarity.
There are technical arguments why either one is supposedly more correct than the other, however I think it is better to decide by listening to the sound.
{Do NOT reverse the mids to bass Polarities, because that is basically a 4th Order crossover there and as such it must be connected as it is in Same Polarity of each driver.}
You may find you want to change the L-pad setting again after Polarity change, but if you do not like the sound after the Polarity change then no amount of L-pad resetting will improve it comprehensively, and if such is the case then change the Polarity back again.
Polarity change should not require any difference to lateral room positioning unless you have a very low ceiling or a very reflective floor and are sitting greater than about 2 metres away from the speakers.
What is often very noticeable in one polarity versus the other is the change of sound in the Vertical axis.
With the 66s the sound should be at its best either on mid-dome axis, or on an axis somewhere in between the mid-dome and the tweeter - regardless of which polarity connection.
Listening from both above tweeter and below mid-dome, sound will be quite different depending on polarity connection of the tweeter.
Stay with your current lateral positions of the speakers in the room when you listen after the polarity change.
Lateral position should only cause audible effects in the bass and low mids,
unless speakers are very close to sidewalls.
I very much doubt the recent crossover changes, even to the bass <--> mids crossover, will cause a problem with lateral positioning,
unless that position was not the optimum in your room prior and you subsequently have heard such better.
*****************************************************
I doubt the "warmer sounding speakers ..... provoke it" {the "thickening"}- I think the "thickening" is mostly in those two loudspeakers, as it definitely in KEF 104s I've heard, and it is to some degree in the 12" cone of the 66 bass driver, even if no longer noticeable in the cabinet walls after your tensioning modifications.
{Don't sell your KEFs yet - I'll Post some ideas to improve those when I have time to.}
Hi Stuart,
With that version of the 66 board layout the two closest magnetically coupled inductors are the 3.5mH input to the bass filter which is dropping a lot of 500Hz and above and those frequencies will couple into that close 2.2mH in the mids' filter which is highish impedance from about 500Hz and up, thus will pick up those unwanted frequencies from the bass filter's 3.5mH, but the 27uF cap in mids' filter is lowish impedance for those frequencies compared to the next section of the mids' filter and the mid-dome, thus perhaps sufficient of the unwanted lows is grounded through the input section of the mids' filter to not cause too much problem, but there will be a small amount of intermodulation with the same frequency band presented electrically from the amplifier through the 27uF cap, though with its phase shifted versus the magnetically coupled component therefore I think likely more-so mutual cancellation of that frequency band than reinforcement.
I hope the reason Celestion engineers placed the inductors in that particular layout was for the inevitable magnetic couplings to do the least harm, and that seems to be the case within my current thinking about the relative functions of each inductor there.
I've never measured to confirm my thinking, because I never place inductors in ways that will significantly magnetically couple !
As you have made your own boards, if sufficient space move the 3.5mH coil further into the corner - further away from both the 2.2mH coils - and you should get a slightly clearer mids' sound - but don't move either 2.2mH coil any closer to the two smaller coils, nor those closer to each other.
Mids and treble caps slightly further from the coils, as you've said you've done, is good so long as the horizontal axis for each is still in the direction of the originals.
*****************************************************
I've definitely not heard Dynaudio Contour 1.8Mk1, because I'd remember hearing an ABR design even if its weighted differently to the Celestion and KEF designs, as all ABRs have part-similar characteristic sound.
I've heard a mid 80s' era Monitor Audio ABR design in similar cabinet shape to that 1.8 , and another brand I've forgotten in a smaller cabinet.
All Dynaudios I've heard have a prominent midrange - it is caused by the frequency characteristic of their bass-mid cone, particually in the upper mids.
As you don't hear the lower mids' thickening with Contour 1.8 -{with their single midrange source}- it is reasonable to presume your room is not the main cause of such.
The slight sibilance could be in part from the upper mids of the cone - what is the specified crossover frequency ?
Sibilance from that tweeter is often reported, and the reason given by a few users is microphonic capacitors in the tweeter's filter.
Have a look at the Brand, and type of capacitor(s) in both the tweeter circuit and in the bass-mids' circuit, and Post here ... ?
Look also and Post the specific model number of the tweeter if it is printed on the back of its magnet or chassis - there were at least 3 versions of that tweeter, and all look much the same from the front.
EXPERIMENT:-
As for the Dynaudios' "open character" - try reversing the Polarity of the tweeter connections to the crossover in the Celestion 66s.
Reverse it at either the board output OR the tweeter terminals - whichever is easier, though at the board output is safer because there one is not applying a hot soldering iron to tweeter leads which if prolonged can conduct sufficient heat into the voice-coil to melt a part of it.
The filter to the Celestion tweeter is 3rd Order.
3rd Order crossovers can be used in Same Polarity or in Reverse Polarity with respect to the two drivers.
Celestion have used Reverse Polarity of tweeter versus mid-dome.
Some listeners prefer the audio character of such, but others, including myself, prefer Same Polarity.
There are technical arguments why either one is supposedly more correct than the other, however I think it is better to decide by listening to the sound.
{Do NOT reverse the mids to bass Polarities, because that is basically a 4th Order crossover there and as such it must be connected as it is in Same Polarity of each driver.}
You may find you want to change the L-pad setting again after Polarity change, but if you do not like the sound after the Polarity change then no amount of L-pad resetting will improve it comprehensively, and if such is the case then change the Polarity back again.
Polarity change should not require any difference to lateral room positioning unless you have a very low ceiling or a very reflective floor and are sitting greater than about 2 metres away from the speakers.
What is often very noticeable in one polarity versus the other is the change of sound in the Vertical axis.
With the 66s the sound should be at its best either on mid-dome axis, or on an axis somewhere in between the mid-dome and the tweeter - regardless of which polarity connection.
Listening from both above tweeter and below mid-dome, sound will be quite different depending on polarity connection of the tweeter.
Stay with your current lateral positions of the speakers in the room when you listen after the polarity change.
Lateral position should only cause audible effects in the bass and low mids,
unless speakers are very close to sidewalls.
I very much doubt the recent crossover changes, even to the bass <--> mids crossover, will cause a problem with lateral positioning,
unless that position was not the optimum in your room prior and you subsequently have heard such better.
*****************************************************
I doubt the "warmer sounding speakers ..... provoke it" {the "thickening"}- I think the "thickening" is mostly in those two loudspeakers, as it definitely in KEF 104s I've heard, and it is to some degree in the 12" cone of the 66 bass driver, even if no longer noticeable in the cabinet walls after your tensioning modifications.
{Don't sell your KEFs yet - I'll Post some ideas to improve those when I have time to.}
Last edited:
Dielectric Absorption -{DA}-, and subst. cap. sizes
I have posted about the dielectric absorption in electrolytic capacitors previously in this Thread.
Basically what happens is a small part of the signal is held in the capacitor for longer than the remainder, and is released sufficiently later to be separately audible.
If every bit of the music was exactly the same volume and intensity as every other bit then the later released signal would mostly cause a loudness increase.
However ... most music contains variations in emphasis and intensity, and when the absorbed parts of the louder or more intense parts are released late - during quieter sections or pauses between transients, etc ... - the audible effect is heard as a smearing of the signal.
When this smearing is only in part of the frequency range it is heard as a colouration - similar to a resonance.
In dloper's 66s there is now low dielectric absorption caps in the upper mids and treble circuits, and high dielectric absorption caps in the lower mids and bass circuit, thus there will be audible colouration from low mids down through the bass.
Some of this is not audibly unpleasant because it is still wide-band and as all across the lower end it is basically heard as warmth of tone,
BUT, there is at least one low mids' fairly narrow band of resonance in the bass driver's cone, thus when the late released, dielectrically absorbed, signal is added to that resonant area in the 12" driver's cone the colouration there is further prolonged and heard as a thickening of the tone.
If the electrolytic caps are replaced with low DA polypropylene caps there will be no late signal release, thus less audible thickening ... though still a small amount as result of the resonance in the cone.
So, Stuart, you can reduce that thickening a little, and I hope sufficiently for your satisfaction, but it will not entirely disappear.
Suitable poly caps:-
For smallest physical size these will have to be bought from Canada or France,
if you want Solen brand - which is what your Wilmslow "Supersound" caps are,
because only the larger size 400volt Solens are available from other UK sellers.
There is not much point in buying significantly more expensive caps than Solens for the bass circuit, because IF they are an audibly superior brand it would have been better they were used in the mids and treble circuits than Solens, thus stay with the lower priced caps unless you plan to replace the mids and treble caps again.
Sizes:-
250 volt Solens are 44mm {1 3/4"} diameter for the values needed in the bass circuit, and 60mm long {almost 2 1/2"}.
400 volt Solens are between 49mm and 52mm diameter -{I am not sure which for the UK sold 70uF as that is not a standard cap value}- and 65mm long.
There is a 26mm {1"} diameter, low price, 75uF cap of likely similar audio quality to Solen available from Denmark via Mail-order, but it is 150mm long {6"} !
One of these could be used for the C1 position if only one 6" long cap will fit somewhere suitable on the board, and a 68uF, 39mm dia. x 56mm long, Danish cap can be used for C2 in dloper's circuit because he has slightly raised the crossover point as result of 27uF instead of 30uF to the MF500 version mid-dome in his 66s.
{I am presuming your mid-dome is the MF version Stuart ... ?}
If 44mm diameter will not fit for C1 and C2, and 150mm/6" is too long, then a Parallel connected pair of 36uF Solens - 34mm x 48mm - could be used, but from Canada, {and perhaps France}.
So, what diameters and what lengths can you fit suitably on your boards Stuart ?
... and where do you prefer to buy from ?
I have posted about the dielectric absorption in electrolytic capacitors previously in this Thread.
Basically what happens is a small part of the signal is held in the capacitor for longer than the remainder, and is released sufficiently later to be separately audible.
If every bit of the music was exactly the same volume and intensity as every other bit then the later released signal would mostly cause a loudness increase.
However ... most music contains variations in emphasis and intensity, and when the absorbed parts of the louder or more intense parts are released late - during quieter sections or pauses between transients, etc ... - the audible effect is heard as a smearing of the signal.
When this smearing is only in part of the frequency range it is heard as a colouration - similar to a resonance.
In dloper's 66s there is now low dielectric absorption caps in the upper mids and treble circuits, and high dielectric absorption caps in the lower mids and bass circuit, thus there will be audible colouration from low mids down through the bass.
Some of this is not audibly unpleasant because it is still wide-band and as all across the lower end it is basically heard as warmth of tone,
BUT, there is at least one low mids' fairly narrow band of resonance in the bass driver's cone, thus when the late released, dielectrically absorbed, signal is added to that resonant area in the 12" driver's cone the colouration there is further prolonged and heard as a thickening of the tone.
If the electrolytic caps are replaced with low DA polypropylene caps there will be no late signal release, thus less audible thickening ... though still a small amount as result of the resonance in the cone.
So, Stuart, you can reduce that thickening a little, and I hope sufficiently for your satisfaction, but it will not entirely disappear.
Suitable poly caps:-
For smallest physical size these will have to be bought from Canada or France,
if you want Solen brand - which is what your Wilmslow "Supersound" caps are,
because only the larger size 400volt Solens are available from other UK sellers.
There is not much point in buying significantly more expensive caps than Solens for the bass circuit, because IF they are an audibly superior brand it would have been better they were used in the mids and treble circuits than Solens, thus stay with the lower priced caps unless you plan to replace the mids and treble caps again.
Sizes:-
250 volt Solens are 44mm {1 3/4"} diameter for the values needed in the bass circuit, and 60mm long {almost 2 1/2"}.
400 volt Solens are between 49mm and 52mm diameter -{I am not sure which for the UK sold 70uF as that is not a standard cap value}- and 65mm long.
There is a 26mm {1"} diameter, low price, 75uF cap of likely similar audio quality to Solen available from Denmark via Mail-order, but it is 150mm long {6"} !
One of these could be used for the C1 position if only one 6" long cap will fit somewhere suitable on the board, and a 68uF, 39mm dia. x 56mm long, Danish cap can be used for C2 in dloper's circuit because he has slightly raised the crossover point as result of 27uF instead of 30uF to the MF500 version mid-dome in his 66s.
{I am presuming your mid-dome is the MF version Stuart ... ?}
If 44mm diameter will not fit for C1 and C2, and 150mm/6" is too long, then a Parallel connected pair of 36uF Solens - 34mm x 48mm - could be used, but from Canada, {and perhaps France}.
So, what diameters and what lengths can you fit suitably on your boards Stuart ?
... and where do you prefer to buy from ?
Hi Alan - sorry to take so long to get back to you. Haven't had much chance to listen lately, or do much else for that matter (looking after small grandchildren) so I'll get back with the Dynaudio specs when I can. I prefer to listen later in the evenings, as we live next to a main road which is quite noisy through the day. I am also of the opinion that my hi-fi sounds better late at night for some reason. Less background noise or a cleaner mains supply perhaps?
I did make time to swap the polarity on the treble units to the same polarity as the other drivers. I swapped back and forth several times, and I do prefer it with the polarity swapped. The treble seems better integrated with the rest of the sound and there does seem to be a more 'open' quality. I did find that I prefer the sound with more attenuation on the L-pads after changing the polarity on the tweeters. After a little more experiment I'll post the L-pad resistance figures. I also tried moving the speakers back and forwards, but ended up back at my original position, with the centre of the bass dome 750mm from the rear wall and 950mm from the sides.
I am quite happy to fit physically large caps in the bass circuit and I don't mind reworking the boards to suit. It doesn't matter where they are sourced. By the way, the mid-range drivers are MD 500s.
I'm away for a couple of days now, but I'll look at the Dynaudios when I get back.
Stuart
I did make time to swap the polarity on the treble units to the same polarity as the other drivers. I swapped back and forth several times, and I do prefer it with the polarity swapped. The treble seems better integrated with the rest of the sound and there does seem to be a more 'open' quality. I did find that I prefer the sound with more attenuation on the L-pads after changing the polarity on the tweeters. After a little more experiment I'll post the L-pad resistance figures. I also tried moving the speakers back and forwards, but ended up back at my original position, with the centre of the bass dome 750mm from the rear wall and 950mm from the sides.
I am quite happy to fit physically large caps in the bass circuit and I don't mind reworking the boards to suit. It doesn't matter where they are sourced. By the way, the mid-range drivers are MD 500s.
I'm away for a couple of days now, but I'll look at the Dynaudios when I get back.
Stuart
Correcting a mistake, and more about the capacitors
Hi Stuart,
"small grandchildren" ... the likely attraction to the vibratings ABRs in 66s may be too much for them to resist ... and then all the active drivers to feel !
Hi-Fi sounding better late in the evening is a common phenomenon, and the reason is often given that the mains are cleaner then and I think that is likely correct as busier times' voltage fluctuations and electrical noises superimposed on the basic AC are not completely controlled by most audio equipments' power supplies.
There is also the psychological factor resulting from less other noises, and likely also as less other sensory distraction as less bright light then.
*****************************************************
Correction:-
I have several times typed 27uF in recent Posts for the series cap at input to the midrange filter.
That has probably stuck in my mind as I recommended in early Posts in this Thread to use 27uF instead of a parallel composite for the 28uF in the intermediate versions of the 66.
Indeed if one wants to keep the crossover as it was with the 28uF versions, then do use a modern 27uF poly cap plus an ESR simulating resistor,
BUT in recent Posts about dloper's 66s I should have been typing 25uF, because that is what he had installed.
Now I read that he has the MD version of the mid-dome, thus a single 25uF cap is the correct modern equivalent, though if one wants 24uF as in final MkII 66, then use a parallel pair or 12uFs, but I doubt this will be result in better sound unless a 25uF sample is substantially on the +ive side of its Tolerance.
Also, if one 12uF cap is largely +ive toleranced and the other is largely -ive toleranced the audible result may not be as clear as a single 25uF cap - some critical listeners have reported this phenomena when using nominally same value parallelled caps versus a single cap.
This would be more audible in the mids and treble than it would be with caps in the shunt circuits of bass filters.
I think for the older MF500 use 30uF or 15//15uF, if you want to stay with the original sound, but 27uF there will give slightly increased volume level capacity from the older driver, and very little audible tonal difference to 30uF - when the optimum value of ESR simulating resistor is found.
For the MD500 I think it better to use 24 or 25uF and not 28uF -{or 27uF}-, because Celestion only used 28uF in the interim version and settled on 24uF for the final MkII, but if you are a critical listener then try both options and decide by sound.
*****************************************************
Stuart, you may have "Blackie" in external finish on your 66s, but as MD500 equipped you have MkII 66 - unless the bass drivers have the older centre dust cap, and that is not likely a significant audible concern for 500Hz region and lower.
I'm out of available time now - I'll list caps and sellers next time.
Hi Stuart,
"small grandchildren" ... the likely attraction to the vibratings ABRs in 66s may be too much for them to resist ... and then all the active drivers to feel !
Hi-Fi sounding better late in the evening is a common phenomenon, and the reason is often given that the mains are cleaner then and I think that is likely correct as busier times' voltage fluctuations and electrical noises superimposed on the basic AC are not completely controlled by most audio equipments' power supplies.
There is also the psychological factor resulting from less other noises, and likely also as less other sensory distraction as less bright light then.
*****************************************************
Correction:-
I have several times typed 27uF in recent Posts for the series cap at input to the midrange filter.
That has probably stuck in my mind as I recommended in early Posts in this Thread to use 27uF instead of a parallel composite for the 28uF in the intermediate versions of the 66.
Indeed if one wants to keep the crossover as it was with the 28uF versions, then do use a modern 27uF poly cap plus an ESR simulating resistor,
BUT in recent Posts about dloper's 66s I should have been typing 25uF, because that is what he had installed.
Now I read that he has the MD version of the mid-dome, thus a single 25uF cap is the correct modern equivalent, though if one wants 24uF as in final MkII 66, then use a parallel pair or 12uFs, but I doubt this will be result in better sound unless a 25uF sample is substantially on the +ive side of its Tolerance.
Also, if one 12uF cap is largely +ive toleranced and the other is largely -ive toleranced the audible result may not be as clear as a single 25uF cap - some critical listeners have reported this phenomena when using nominally same value parallelled caps versus a single cap.
This would be more audible in the mids and treble than it would be with caps in the shunt circuits of bass filters.
I think for the older MF500 use 30uF or 15//15uF, if you want to stay with the original sound, but 27uF there will give slightly increased volume level capacity from the older driver, and very little audible tonal difference to 30uF - when the optimum value of ESR simulating resistor is found.
For the MD500 I think it better to use 24 or 25uF and not 28uF -{or 27uF}-, because Celestion only used 28uF in the interim version and settled on 24uF for the final MkII, but if you are a critical listener then try both options and decide by sound.
*****************************************************
Stuart, you may have "Blackie" in external finish on your 66s, but as MD500 equipped you have MkII 66 - unless the bass drivers have the older centre dust cap, and that is not likely a significant audible concern for 500Hz region and lower.
I'm out of available time now - I'll list caps and sellers next time.
Alan, sorry for the delay…
IIRC, the mic was at the centre of the mid-dome
The fronts on all four of my 66s have been painted and now appear similar, so I don‘t recall which mid-dome was being measured. However, I wonder about there being any actual difference between the MD and the MF mid-domes. The label change appeared around halfway through the 66’s run… and perhaps during a label reorder someone decided that 30w was too conservative a rating.
I was just looking over my mid-domes and, among my six units (I have two backup units), the only apparent difference is in the color of the surrounds. Half of the surrounds are black ( 3 MFs ) and half are orange ( 2 MDs & 1 MF ). I don’t know what is typical.
I suppose it’s possible that the orange ones might indicate newly designed surrounds for higher power handling (and perhaps one of my MFs received one of those newer diaphragms along the way). But who knows! Until I see more evidence, I’m sticking with the notion that the MDs and the MFs are one and the same… they look similar, they measure similar, and, to my ear, they sound similar.
*****************************
NEW MEASUREMENTS--
Here are some T/S measurements of my mid-domes. It doesn’t look good! Half of my domes have Fs that are too high to be used safely with a 500hz X-O. Perhaps the Fs were lower when the units were new, and then migrated upwards with age. I think the x-o point should be moved up to at least 800hz, and maybe even higher. IIRC, the Ditton 25 is like a 66 without a midrange, so the woofers should be capable of going there.
Alan, what do you think about moving the x-o point ? Shouldn’t it be at least an octave above the Fs ?
Harnfield, by any chance were you able to test the mid-domes for distortion?
Have a nice holiday !
Last edited:
Great stuff !!!
No apology is necessary sba, as we all are busy with other matters at times,
but thankyou for your consideration.
That is a clever measuring device you have sba !
- what is its brand and model number ?
Does yours do all these things harnfield ?
I'll have to be brief today sba as I'm almost out of available time,
but don't worry too much as you have sufficient to work with to have two very good 66s and two OK enough ...
There are some tricks you can do with resistors to enable your Blue and Purple tweeters to work more safely in their x-overs,
AND with all the capacitors you bought - if you can measure the capacitance of each sample for the treble filters' caps you can select those best suited for a filter for each tweeter - I mean the +/- tolerance differences - some caps will be larger and some will be smaller.
Label them all after you measure them.
{Groan ... , yes a lot of work as you have them all soldered into circuits, but only one end of each cap has to be disconnected from the circuit to measure it.}
I'll describe caps for tweeter samples and about the resistors next time ... and in preparation for that can you measure the relative SPLs of each tweeter ?
or, simply list their apparent relative to your ear ?
The woofers are fine - I'll comment further next time if I think anything relevant to pair selection for.
NOTICE - and dloper also - the small short plateau on the woofers' impedance plots between 700 and 800 Hz.
This is the lowest frequency break-up resonance, and will be audible from the cone even though a little above the nominal 500Hz crossover point.
This may be part of what dloper is hearing, as described in his Posts about the low mids.
The next two - at about 1.6kHz and 2.1kHz will be less audible with low crossover frequency.
Highest reasonable crossover would seem at about 720Hz ,
but I would have to see frequency response plots to be sure.
Can your measure and Post such sba, with the woofers disconnected from their crossovers ?
Can you Post the ones you had done harnfield ?
Celestion 25 was not particually good sounding, so far as I can remember -{though not really poor}- and Celestion 44 was better,
thus I'd be wary of thinking those woofers could work to significantly higher frequency for Hi-Fi.
I'm not up on modern abbreviations - please inform me, what does IIRC mean ?
Mid-domes:-
the best two are the two Greens.
Purple will need a parallel resistor to use safely at high volume in the current crossover - note its high Qms - it is not well damped.
Remember "tonedef2's" description early in this Thread about the clever mechanical damping of the MD500 - well the Purple does not have it - perhaps a faulty or age deteriorated sample.
Yellow has poor electrical damping - Qes - perhaps from an out of specification magnet or voice-coil or age deteriorated change of position of its voice-coil.
Blue and Orange can be got to work as a pair, and those are the two I'd try with slightly higher crossover frequency.
Which MD and MF and surround colour apply to which measured sample domes ?
If you can measure all the capacitances of your mids' filters caps sba, some may be able to be applied for a sufficient raising of crossover,
and otherwise applied better to the sample differences between the mid-domes.
AND, also all your samples of bass filters' caps, though simply using some 62uF alone there may work ,
BUT I'll have to think about that for a while ...
Are the L(e) measurements at any particular frequencies for the 3 different driver types - woofers, mid-domes, tweeters - and do Post the frequencies -
or are these some other type of L(e) measurement ?
Depending on frequencies, thse may assist if you want to attempt change of crossover to a pair of 66s.
I too hope everyone has a nice holiday !
No apology is necessary sba, as we all are busy with other matters at times,
but thankyou for your consideration.
That is a clever measuring device you have sba !
- what is its brand and model number ?
Does yours do all these things harnfield ?
I'll have to be brief today sba as I'm almost out of available time,
but don't worry too much as you have sufficient to work with to have two very good 66s and two OK enough ...
There are some tricks you can do with resistors to enable your Blue and Purple tweeters to work more safely in their x-overs,
AND with all the capacitors you bought - if you can measure the capacitance of each sample for the treble filters' caps you can select those best suited for a filter for each tweeter - I mean the +/- tolerance differences - some caps will be larger and some will be smaller.
Label them all after you measure them.
{Groan ... , yes a lot of work as you have them all soldered into circuits, but only one end of each cap has to be disconnected from the circuit to measure it.}
I'll describe caps for tweeter samples and about the resistors next time ... and in preparation for that can you measure the relative SPLs of each tweeter ?
or, simply list their apparent relative to your ear ?
The woofers are fine - I'll comment further next time if I think anything relevant to pair selection for.
NOTICE - and dloper also - the small short plateau on the woofers' impedance plots between 700 and 800 Hz.
This is the lowest frequency break-up resonance, and will be audible from the cone even though a little above the nominal 500Hz crossover point.
This may be part of what dloper is hearing, as described in his Posts about the low mids.
The next two - at about 1.6kHz and 2.1kHz will be less audible with low crossover frequency.
Highest reasonable crossover would seem at about 720Hz ,
but I would have to see frequency response plots to be sure.
Can your measure and Post such sba, with the woofers disconnected from their crossovers ?
Can you Post the ones you had done harnfield ?
Celestion 25 was not particually good sounding, so far as I can remember -{though not really poor}- and Celestion 44 was better,
thus I'd be wary of thinking those woofers could work to significantly higher frequency for Hi-Fi.
I'm not up on modern abbreviations - please inform me, what does IIRC mean ?
Mid-domes:-
the best two are the two Greens.
Purple will need a parallel resistor to use safely at high volume in the current crossover - note its high Qms - it is not well damped.
Remember "tonedef2's" description early in this Thread about the clever mechanical damping of the MD500 - well the Purple does not have it - perhaps a faulty or age deteriorated sample.
Yellow has poor electrical damping - Qes - perhaps from an out of specification magnet or voice-coil or age deteriorated change of position of its voice-coil.
Blue and Orange can be got to work as a pair, and those are the two I'd try with slightly higher crossover frequency.
Which MD and MF and surround colour apply to which measured sample domes ?
If you can measure all the capacitances of your mids' filters caps sba, some may be able to be applied for a sufficient raising of crossover,
and otherwise applied better to the sample differences between the mid-domes.
AND, also all your samples of bass filters' caps, though simply using some 62uF alone there may work ,
BUT I'll have to think about that for a while ...
Are the L(e) measurements at any particular frequencies for the 3 different driver types - woofers, mid-domes, tweeters - and do Post the frequencies -
or are these some other type of L(e) measurement ?
Depending on frequencies, thse may assist if you want to attempt change of crossover to a pair of 66s.
I too hope everyone has a nice holiday !
Last edited:
Measuring woofers' mid- frequency resoponse.
Edit time expired, so here instead ...
Drivers disconnected from their crossovers, but still in their cabinets,
and if you measure each in its own cabinet, then do place each cabinet in exactly the same position in the room, and distant from walls,
and pointing towards an open doorway may assist greater validity of measurement result, or at least towards a soft sofa !
Place the mic about 18" <---> 2ft in front of the centre of the woofer cone.
Not closer, as that will cause annomolies in the measured result in the midrange.
Don't worry about annomolies in the bass - we only need to see the midrange response.
Place a thick, soft, sound absorbing item on the floor halfway between the cabinet and the mic.
A loosely bundled wool blanket, or a cotton or similar filled pillow, or very soft cushion would likely suffice.
No filters or tone controls active on in the amplifier the mic feeds.
If you want to measure the mid-domes in a similar way, that may be useful for fine-tuning crossovers to your differnt tweeter samples.
Edit time expired, so here instead ...
Drivers disconnected from their crossovers, but still in their cabinets,
and if you measure each in its own cabinet, then do place each cabinet in exactly the same position in the room, and distant from walls,
and pointing towards an open doorway may assist greater validity of measurement result, or at least towards a soft sofa !
Place the mic about 18" <---> 2ft in front of the centre of the woofer cone.
Not closer, as that will cause annomolies in the measured result in the midrange.
Don't worry about annomolies in the bass - we only need to see the midrange response.
Place a thick, soft, sound absorbing item on the floor halfway between the cabinet and the mic.
A loosely bundled wool blanket, or a cotton or similar filled pillow, or very soft cushion would likely suffice.
No filters or tone controls active on in the amplifier the mic feeds.
If you want to measure the mid-domes in a similar way, that may be useful for fine-tuning crossovers to your differnt tweeter samples.
question on bass circuit
first - I moved back here from the 44 *nightmare*. scoping all the brands for the 72 uf's, think I will use electrolytics ( I'll be gone before they wear out) and the space consideration !
Closest is 68 uf in most brands. will 68 suffice for the 72's?? what would be the end result? Alan, think I am going to not practise with varying resistors, going to pick middle values and put em in.
starting to look like get all from either connexion or express.
before ordering just wanted thought on using 68 in place of 72. thanks.
DM
first - I moved back here from the 44 *nightmare*. scoping all the brands for the 72 uf's, think I will use electrolytics ( I'll be gone before they wear out) and the space consideration !
Closest is 68 uf in most brands. will 68 suffice for the 72's?? what would be the end result? Alan, think I am going to not practise with varying resistors, going to pick middle values and put em in.
starting to look like get all from either connexion or express.
before ordering just wanted thought on using 68 in place of 72. thanks.
DM
capacitors for bass filter's circuit, and mods to come ...
Hello DM - and everyone else interested in the bass filter circuit -
it is good that you have joined us here.
(1) - how much of a hurry are you in ?
Given the measurements that sba has posted I may be able to calculate for fine-tuning the bass filter a little tighter now than was done economically years ago.
Hopefully sba - and harnfield - will be able to Post frequency response plots for the woofers, from which it will be easier to decide about 68uF versus 72uF versus 75uF versus what I will describe below in this Post.
Given that you have 24uF -{or 25uF ?}- to your MD500 mid-domes, it may be better to use 68uF for at least the cap that is electrically closest to the woofer in the circuit.
(2) - if you want to buy now and not wait for further development of the crossover, then simplest is to go to:-
Untitled Document -{solen.ca}-
if I have correctly recognised your Flag ... ?
There, look under Solen 250volt PA series polypropylene caps.
These are thinner diameter than the 400volt Solens sold elsewhere, and are lower price !
Easiest to fit, and closest to Celestion's specification is to use parallel connected pairs of 36uF for 36//36 = 72uF.
Lay one cap on the board and lay the second cap on top of the first,
and solder at one point only the wires at each end - preferably not twisting the caps' leads together for a significant distance.
Use a 1.5ohm/5watt wirewound resistor in Series with the cap pair that are closest electrically to the woofer,
and use a 1ohm/5watt in Series with the distant caps' pair.
These do not need to be non-inductive, and you can use a cheap wirewound, but better is to buy in Welwyn WP4S series from Farnell.com , or in "Ohmite" brand -{USA manufactured}- from where-ever you can.
(3) - if you don't want to use parallel pairs, then in the same 250volt PA series there is the thin diameter 75uF cap.
One in each position will be OK, and use the same resistors as above.
(4) - perhaps better may be to use a 250volt PA, 68uF cap for the closest electrically to the woofer, and a 75uF cap for the distant located cap ... and with the same resistors as above.
This staggered filter may be better, particully when using 25uF to the MD500 in the mids filter.
I will be thinking through this more as I study sba's posted plots, and whatever else anyone may post.
(5) - don't think that you may not be able to hear the difference between electrolytic caps and polypropylene caps in the bass filter.
Read everything that dloper has posted about what he hears in the low mids from his electros' fitted bass filter.
I have heard similar in several crossovers, and there is a published diy article in a British magazine by an experienced crossover designer who heard similar to dloper and modified the crossover in his client's loudspeaker who was also hearing the same - from electro caps in shunt location in the bass filter of a 3-way loudspeaker.
(6) - if you will only buy electrolytic caps, then at Parts Express in their Crossover Capacitors section are 68uF and 80uF.
You could try one of each in each filter in the way I described above for the 68uF and 75uF poly caps.
You will not need ESR simulating resistors with the electro caps, but if we find a better way to modify this filter later you may have to add resistors to get close to it.
"Parts ConneXion" have Mills MRA-5 non-inductive wirewounds for the mids and treble filters.
I recommend you start with 1.5ohm and 3.9 ohm in the mids filter and 0.5ohm(0.47ohm) in the tweeter filter, connected between the tweeter and the 6uF cap.
I will be posting more about another modification to the mids' filter based on the Impedance plots that sba has posted, and that will include the 1.5ohm and 3.9ohm resistors to try there first.
Post the brands and types of caps are you using in the mids and tweeter filters currently ?
I will have better idea then of the degree it is worth developing your bass filter towards.
Hello DM - and everyone else interested in the bass filter circuit -
it is good that you have joined us here.
(1) - how much of a hurry are you in ?
Given the measurements that sba has posted I may be able to calculate for fine-tuning the bass filter a little tighter now than was done economically years ago.
Hopefully sba - and harnfield - will be able to Post frequency response plots for the woofers, from which it will be easier to decide about 68uF versus 72uF versus 75uF versus what I will describe below in this Post.
Given that you have 24uF -{or 25uF ?}- to your MD500 mid-domes, it may be better to use 68uF for at least the cap that is electrically closest to the woofer in the circuit.
(2) - if you want to buy now and not wait for further development of the crossover, then simplest is to go to:-
Untitled Document -{solen.ca}-
if I have correctly recognised your Flag ... ?
There, look under Solen 250volt PA series polypropylene caps.
These are thinner diameter than the 400volt Solens sold elsewhere, and are lower price !
Easiest to fit, and closest to Celestion's specification is to use parallel connected pairs of 36uF for 36//36 = 72uF.
Lay one cap on the board and lay the second cap on top of the first,
and solder at one point only the wires at each end - preferably not twisting the caps' leads together for a significant distance.
Use a 1.5ohm/5watt wirewound resistor in Series with the cap pair that are closest electrically to the woofer,
and use a 1ohm/5watt in Series with the distant caps' pair.
These do not need to be non-inductive, and you can use a cheap wirewound, but better is to buy in Welwyn WP4S series from Farnell.com , or in "Ohmite" brand -{USA manufactured}- from where-ever you can.
(3) - if you don't want to use parallel pairs, then in the same 250volt PA series there is the thin diameter 75uF cap.
One in each position will be OK, and use the same resistors as above.
(4) - perhaps better may be to use a 250volt PA, 68uF cap for the closest electrically to the woofer, and a 75uF cap for the distant located cap ... and with the same resistors as above.
This staggered filter may be better, particully when using 25uF to the MD500 in the mids filter.
I will be thinking through this more as I study sba's posted plots, and whatever else anyone may post.
(5) - don't think that you may not be able to hear the difference between electrolytic caps and polypropylene caps in the bass filter.
Read everything that dloper has posted about what he hears in the low mids from his electros' fitted bass filter.
I have heard similar in several crossovers, and there is a published diy article in a British magazine by an experienced crossover designer who heard similar to dloper and modified the crossover in his client's loudspeaker who was also hearing the same - from electro caps in shunt location in the bass filter of a 3-way loudspeaker.
(6) - if you will only buy electrolytic caps, then at Parts Express in their Crossover Capacitors section are 68uF and 80uF.
You could try one of each in each filter in the way I described above for the 68uF and 75uF poly caps.
You will not need ESR simulating resistors with the electro caps, but if we find a better way to modify this filter later you may have to add resistors to get close to it.
"Parts ConneXion" have Mills MRA-5 non-inductive wirewounds for the mids and treble filters.
I recommend you start with 1.5ohm and 3.9 ohm in the mids filter and 0.5ohm(0.47ohm) in the tweeter filter, connected between the tweeter and the 6uF cap.
I will be posting more about another modification to the mids' filter based on the Impedance plots that sba has posted, and that will include the 1.5ohm and 3.9ohm resistors to try there first.
Post the brands and types of caps are you using in the mids and tweeter filters currently ?
I will have better idea then of the degree it is worth developing your bass filter towards.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.