Hi All.
Wife decided that my Ditton 66 LSs need a good home - priorities change as you become older !
I could split out drivers, but the cabinets are good too, and I still have the original cardboard boxes they came in. However, I live 7 miles NW of Belfast in Northern Ireland, and think everyone interested here is on a different land mass which means cabinet carriage would be very expensive.
No crossovers though, these were stripped out 30+ years ago, and one bass driver has a loose turn which can rub when hot, but still running and being used to this day in our lounge.
Would make an ideal rebuild project or be good for spares, and no hurry if anyone interested, I can be contacted via my Profile e-mail.
Cheers .......... Graham.
Wife decided that my Ditton 66 LSs need a good home - priorities change as you become older !
I could split out drivers, but the cabinets are good too, and I still have the original cardboard boxes they came in. However, I live 7 miles NW of Belfast in Northern Ireland, and think everyone interested here is on a different land mass which means cabinet carriage would be very expensive.
No crossovers though, these were stripped out 30+ years ago, and one bass driver has a loose turn which can rub when hot, but still running and being used to this day in our lounge.
Would make an ideal rebuild project or be good for spares, and no hurry if anyone interested, I can be contacted via my Profile e-mail.
Cheers .......... Graham.
Attachments
What a tragedy! It would be an even bigger tragedy to break them up, though. Why not just fleabay them if there's no interest here?
I'm curious about the image in the thumbnail. Is it the cardboard boxes? If it's the brochure that came with the speakers, is there any juicy technical info on it?
More curiosity - how were you running the speakers without crossovers, if you don't my asking?
I'm curious about the image in the thumbnail. Is it the cardboard boxes? If it's the brochure that came with the speakers, is there any juicy technical info on it?
More curiosity - how were you running the speakers without crossovers, if you don't my asking?
Yes, a good question indeed if we are to consider buying them !
If the woofers alone are driven without cross-over there will be no problem, but the mids and tweeters - do you have those connected and music through them currently ?
Is there anything else printed on that item you thoughtfully attached {and thankyou - it is historically interesting} - anything printed on the back of it you could Post ?
If the woofers alone are driven without cross-over there will be no problem, but the mids and tweeters - do you have those connected and music through them currently ?
Is there anything else printed on that item you thoughtfully attached {and thankyou - it is historically interesting} - anything printed on the back of it you could Post ?
lost email
harnfield,
I'm sorry, but I lost your email.
It arrived in my Junk-mail folder. I read it then attempted to transfer it to Save folder, but it got erased some-how and I could not recover it.
I will try to have a look at a sample of that multimeter this week to read its booklet for more detailed specification, then decide and Post here.
Resistor comments will be in another Post here today.
harnfield,
I'm sorry, but I lost your email.
It arrived in my Junk-mail folder. I read it then attempted to transfer it to Save folder, but it got erased some-how and I could not recover it.
I will try to have a look at a sample of that multimeter this week to read its booklet for more detailed specification, then decide and Post here.
Resistor comments will be in another Post here today.
lost email
harnfield,
I'm sorry, but I lost your email.
It arrived in my Junk-mail folder. I read it then attempted to transfer it to Save folder, but it got erased some-how and I could not recover it.
I will try to have a look at a sample of that multimeter this week to read its booklet for more detailed specification, then decide and Post here.
I see the Bids finish on 31st January, well ...
these meters are not rare, nor expensive new, and if one buys from a shop one gets a Warranty - test equipment is useless if it has a fault or develops a fault.
2.5% is not great accuracy but is adequate for matching pairs or choosing small discriminations between samples for capacitors, and for finding if inductors in one speaker are close to same value in the other speaker.
Unless you intend to wind you own inductors or re-use unknown value inductotrs from old speakers you will not need an L meter.
For Capacitance, better than 2.5% can be bought, eg: my "Fluke" 87 multimeter has cap accuracy is +1% +2 counts.
Resistance 0.8% is OK for some purposes, but that spec. may be only for the middle resistance ranges in that meter - some meters will not measure very low resistances and/or very high resistances to the same accuracy as their 10ohm to 1Mohm specification, thus I would want to see the below 10ohm spec for that meter.
Fluke 87 is +/- 0.2% +1count for all to 4000kohm, including the ability to use its Relative Mode to zero the residual resistance of the test leads so as to get same accuracy for low down to 0.1ohm.
Fluke meters cost more than Digitch, but there will be a few Asian manufactured meters of almost as good spec. as the Flukes for much lower price.
You seem prepared to spend a lot more money on buying resistors than is necessary for good results with the 66s, thus why not buy a good multimeter ?
What do you plan to do with test equipment, only recondition old speakers and recycle old components ?
or do you want to make other measurements, eg: of audio frequencies ?
because though that Digitech meter measures AC voltage and AC current, the summary spec. does not state the frequency range it is accurate to.
Most cheap meters are only designed to measure the AC of the main harmonics of AC Electricity Supply {50Hz or 60Hz} and will measure harmonics up to about 400Hz within the basic accuracy, but not above.
The full spec. in a meter's Manual should give its AC range ability and accuracies up to at least 10kHz to be much use for audio, and, often the AC volts will measure to higher freq's than the AC amps,{not necessarily a problem if sufficient range for the amps, or if one always uses a series resistor when testing and thus can measure volts}.
A Frequency Counter is useful in a meter, and if one has this and an Audio frequency oscillator, one does not need an L meter, as one can resonate an unknown inductor with a known capacitor and calculate the inductance.
I did mean Oscillator, not Oscilloscope.
Audio frequency oscillator or Audio frequency generator, or Signal Generator, or Sine-wave generator - all names for a generator which is use ful if it includes sine waves between 20Hz and 20kHz in one or more of its ranges.
But if you don't have such generator/oscillator, nor want to buy one, then an LCR meter will be more useful than a multimeter as you won't need AC volts and amps, thus could buy an LCR meter with good accuracy spec.s , including for below 10ohms ability.
___________________________________________________
Huh - wretched Post did not Edit, but has left the original {above} and my added to as a new Post here, and which now seems to be Editing OK ! - really I am happier with loudspeakers than I am with computers ...
harnfield,
I'm sorry, but I lost your email.
It arrived in my Junk-mail folder. I read it then attempted to transfer it to Save folder, but it got erased some-how and I could not recover it.
I will try to have a look at a sample of that multimeter this week to read its booklet for more detailed specification, then decide and Post here.
I see the Bids finish on 31st January, well ...
these meters are not rare, nor expensive new, and if one buys from a shop one gets a Warranty - test equipment is useless if it has a fault or develops a fault.
2.5% is not great accuracy but is adequate for matching pairs or choosing small discriminations between samples for capacitors, and for finding if inductors in one speaker are close to same value in the other speaker.
Unless you intend to wind you own inductors or re-use unknown value inductotrs from old speakers you will not need an L meter.
For Capacitance, better than 2.5% can be bought, eg: my "Fluke" 87 multimeter has cap accuracy is +1% +2 counts.
Resistance 0.8% is OK for some purposes, but that spec. may be only for the middle resistance ranges in that meter - some meters will not measure very low resistances and/or very high resistances to the same accuracy as their 10ohm to 1Mohm specification, thus I would want to see the below 10ohm spec for that meter.
Fluke 87 is +/- 0.2% +1count for all to 4000kohm, including the ability to use its Relative Mode to zero the residual resistance of the test leads so as to get same accuracy for low down to 0.1ohm.
Fluke meters cost more than Digitch, but there will be a few Asian manufactured meters of almost as good spec. as the Flukes for much lower price.
You seem prepared to spend a lot more money on buying resistors than is necessary for good results with the 66s, thus why not buy a good multimeter ?
What do you plan to do with test equipment, only recondition old speakers and recycle old components ?
or do you want to make other measurements, eg: of audio frequencies ?
because though that Digitech meter measures AC voltage and AC current, the summary spec. does not state the frequency range it is accurate to.
Most cheap meters are only designed to measure the AC of the main harmonics of AC Electricity Supply {50Hz or 60Hz} and will measure harmonics up to about 400Hz within the basic accuracy, but not above.
The full spec. in a meter's Manual should give its AC range ability and accuracies up to at least 10kHz to be much use for audio, and, often the AC volts will measure to higher freq's than the AC amps,{not necessarily a problem if sufficient range for the amps, or if one always uses a series resistor when testing and thus can measure volts}.
A Frequency Counter is useful in a meter, and if one has this and an Audio frequency oscillator, one does not need an L meter, as one can resonate an unknown inductor with a known capacitor and calculate the inductance.
I did mean Oscillator, not Oscilloscope.
Audio frequency oscillator or Audio frequency generator, or Signal Generator, or Sine-wave generator - all names for a generator which is use ful if it includes sine waves between 20Hz and 20kHz in one or more of its ranges.
But if you don't have such generator/oscillator, nor want to buy one, then an LCR meter will be more useful than a multimeter as you won't need AC volts and amps, thus could buy an LCR meter with good accuracy spec.s , including for below 10ohms ability.
___________________________________________________
Huh - wretched Post did not Edit, but has left the original {above} and my added to as a new Post here, and which now seems to be Editing OK ! - really I am happier with loudspeakers than I am with computers ...
Dueland versus Mills - where ?
I didn't see any comparison to Mills resistors in that long report about replacing other brand components with Dueland in the x-overs - did I miss it ?
The top line Link accesses but not the other two, fro your Parts Connexion Link.
Mills use the Aryton-Perry winding type. It is very low inductive - almost completely non-inductive. It is also low capacitive, which is useful.
Duelund and large Aluminium foil resistors are very expensive, and quite out of context in this 66 restoration process - more than one would likely hear much benefit from, at least not to the degree of benefit one would likely hear if use ClarityCaps instead of Solen for the mids and treble, and did not use 3.9 + 2.4 or 3.9 + 1.8, {but do not worry, 3.9 + 2.4 can be workable, but let's keep the resistors in context with the quality of the caps you have bought}.
Look at the 5 watt series of the Kiwame resistors available fro the UK seller in that Hi-Fi Collective site you Posted the Link to.
1" long - is that OK to fit on your boards, or too long ?
harnfield said:
I didn't see any comparison to Mills resistors in that long report about replacing other brand components with Dueland in the x-overs - did I miss it ?
The top line Link accesses but not the other two, fro your Parts Connexion Link.
Mills use the Aryton-Perry winding type. It is very low inductive - almost completely non-inductive. It is also low capacitive, which is useful.
Duelund and large Aluminium foil resistors are very expensive, and quite out of context in this 66 restoration process - more than one would likely hear much benefit from, at least not to the degree of benefit one would likely hear if use ClarityCaps instead of Solen for the mids and treble, and did not use 3.9 + 2.4 or 3.9 + 1.8, {but do not worry, 3.9 + 2.4 can be workable, but let's keep the resistors in context with the quality of the caps you have bought}.
Look at the 5 watt series of the Kiwame resistors available fro the UK seller in that Hi-Fi Collective site you Posted the Link to.
1" long - is that OK to fit on your boards, or too long ?
Resistors
Kiwame resistors are not the same as Duelund, but they are similar in part, having a carbon compound resistive substance in cylindrical shape - a solid rod for the Duelund, and for the Kiwame a cylinder coated with carbon as a thin film.
This resistive element is thus a hollow cylinder, thus should in theory present the exact same resistance to high frequencies as it does to low frequencies.
Apparently the carbon film is not helically cut - which is the method of setting the resistance of most Carbon Film and Metal Film resistors - but is simply a complete continuous cylinder, thus should be very low inductance, unlike the inductance caused by a helix cut.
I do not know how Kiwame manage to manufacture to such precise values of resistance as the specification states, however I intend to buy some and file the insulating coating off one and examine its resitive element, etc ...
____________________________________________________
There does not seem to be a Retail Seller of the Mills MRA-5 -{MRA05}- non-inductive 5 watt wirewounds in the UK -{only of the 12 watt}- , thus for UK readers who do not want to buy from USA suppliers some other good quality Brands/Manufacturers of wirewounds are :-
Neohm ; Ohmite ; VTM .
I haven't searched for UK sellers, but I suggest you look in the "RS Components" and "Farnell" on-line catalogs for whose they sell in Non-Inductive resistors.
There is apparently an Ohmite "AudioGold" 5 watt range.
"Caddock" manufacture very good quality resistors, using a mixture of several Oxides to form a compound for use as the resistive element.
Caddock MP-930 are 2.25 watt non-inductive and are available in some very low ohm values one could use in some applications in this 66 ESR simulating, but they cost 3 times the price of the Mills non-inductive 5 watters !
Caddock and Mills power ratings are specified at 25*C, thus will have to be de-rated for power handling when they warm up.
The 5watt Mills are not likely to get more than very slightly warm in the Celestion 66 x-over applications, thus why I specified them, as well as their compact size.
The Caddocks may get warm to hot in the C8 and C7 applications if only a single resistor is being used in each location.
They won't get hot in the C5 application, and probably not in the C6 application if 1.5 ohm is used, but may warm up if a single 3.5 ohm resistor is connected to a single 75uF cap there to increase the lower mids output a little.
Caddock MP-930 de-rate to about 1.5watts at 70*C , unless used with a Heat-sink, but using their type of clip-on heatsink is not practical inside a speaker cabinet where there is strong vibration from the woofers moving !
The Mills web-site is at :- www.millsresistor.com
Kiwame resistors are not the same as Duelund, but they are similar in part, having a carbon compound resistive substance in cylindrical shape - a solid rod for the Duelund, and for the Kiwame a cylinder coated with carbon as a thin film.
This resistive element is thus a hollow cylinder, thus should in theory present the exact same resistance to high frequencies as it does to low frequencies.
Apparently the carbon film is not helically cut - which is the method of setting the resistance of most Carbon Film and Metal Film resistors - but is simply a complete continuous cylinder, thus should be very low inductance, unlike the inductance caused by a helix cut.
I do not know how Kiwame manage to manufacture to such precise values of resistance as the specification states, however I intend to buy some and file the insulating coating off one and examine its resitive element, etc ...
____________________________________________________
There does not seem to be a Retail Seller of the Mills MRA-5 -{MRA05}- non-inductive 5 watt wirewounds in the UK -{only of the 12 watt}- , thus for UK readers who do not want to buy from USA suppliers some other good quality Brands/Manufacturers of wirewounds are :-
Neohm ; Ohmite ; VTM .
I haven't searched for UK sellers, but I suggest you look in the "RS Components" and "Farnell" on-line catalogs for whose they sell in Non-Inductive resistors.
There is apparently an Ohmite "AudioGold" 5 watt range.
"Caddock" manufacture very good quality resistors, using a mixture of several Oxides to form a compound for use as the resistive element.
Caddock MP-930 are 2.25 watt non-inductive and are available in some very low ohm values one could use in some applications in this 66 ESR simulating, but they cost 3 times the price of the Mills non-inductive 5 watters !
Caddock and Mills power ratings are specified at 25*C, thus will have to be de-rated for power handling when they warm up.
The 5watt Mills are not likely to get more than very slightly warm in the Celestion 66 x-over applications, thus why I specified them, as well as their compact size.
The Caddocks may get warm to hot in the C8 and C7 applications if only a single resistor is being used in each location.
They won't get hot in the C5 application, and probably not in the C6 application if 1.5 ohm is used, but may warm up if a single 3.5 ohm resistor is connected to a single 75uF cap there to increase the lower mids output a little.
Caddock MP-930 de-rate to about 1.5watts at 70*C , unless used with a Heat-sink, but using their type of clip-on heatsink is not practical inside a speaker cabinet where there is strong vibration from the woofers moving !
The Mills web-site is at :- www.millsresistor.com
Kwame 5W resistors
Have just purchased resistors from HiFi Collective, a UK supplier that Alan mentioned. Possibly better value at £1.82 each than the Mills 12W at £4.70 (incl.VAT).
Nick at HiFi Collective says that Mills 12W sound better than Kwame and better than the Mills 5W. (Anyway, there appears to be no UK stockist of the Mills 5W.)
www.rswww.com have wirewound resistors but not of Mills quality. Ohmite from www.farnell.com appear too pricey.
Have just purchased resistors from HiFi Collective, a UK supplier that Alan mentioned. Possibly better value at £1.82 each than the Mills 12W at £4.70 (incl.VAT).
Nick at HiFi Collective says that Mills 12W sound better than Kwame and better than the Mills 5W. (Anyway, there appears to be no UK stockist of the Mills 5W.)
www.rswww.com have wirewound resistors but not of Mills quality. Ohmite from www.farnell.com appear too pricey.
Hey, you found the seller ! - I merely noticed what else he had to sell ...
Well, he would say the most expensive of the three sounds better, but what sounds "better" is often relevant to the application.
I doubt a physically large 12 watt Mills would sound any better than their physically smaller 5 watt , {which almost certainly uses the same wire}, in this particular application where there will be low dissipation, and the larger size may not sound as good when passing upper mids and treble as the smaller, as long as the smaller does not heat up much - and it shouldn't in this application.
There are always high frequency losses when signal is sent through physically larger components than necessary, unless those larger components have some other feature that is of benefit to the signal.
I do expect the 5watt Mills to sound better than the Kiwame in at least the Series to Mid Dome circuit, but it may be only by a very small amount.
To hear these differences, and the drivers and new cross-over components at optimum the Diffraction from the cabinet edges and grille frame will have to be attended to, at least from the protruding lip at the top of the cabinet as first priority even if one listens with the grilles off.
___________________________________________________
Ohmite "Audio Gold" - I have read the specifications but not found a seller .. yet .. but they seem likely very similar to the Mills if one can obtain the non-inductive option of the Ohmite from anywhere.
It is designed, but may not be in production for sale until someone orders in Quantity.
Well, he would say the most expensive of the three sounds better, but what sounds "better" is often relevant to the application.
I doubt a physically large 12 watt Mills would sound any better than their physically smaller 5 watt , {which almost certainly uses the same wire}, in this particular application where there will be low dissipation, and the larger size may not sound as good when passing upper mids and treble as the smaller, as long as the smaller does not heat up much - and it shouldn't in this application.
There are always high frequency losses when signal is sent through physically larger components than necessary, unless those larger components have some other feature that is of benefit to the signal.
I do expect the 5watt Mills to sound better than the Kiwame in at least the Series to Mid Dome circuit, but it may be only by a very small amount.
To hear these differences, and the drivers and new cross-over components at optimum the Diffraction from the cabinet edges and grille frame will have to be attended to, at least from the protruding lip at the top of the cabinet as first priority even if one listens with the grilles off.
___________________________________________________
Ohmite "Audio Gold" - I have read the specifications but not found a seller .. yet .. but they seem likely very similar to the Mills if one can obtain the non-inductive option of the Ohmite from anywhere.
It is designed, but may not be in production for sale until someone orders in Quantity.
Capacitors in Parallel, and Pulse Rise times
Before I Post any more about specific caps for options for the x-over I am going to address the subject of connecting Capacitors in Parallel as it is not well understood in critical Audio applications.
In Basic Electronics one learns that any two or more caps can be connected in parallel to increase the over-all capacitance to the sum of those connected, but that is simplistic by itself.
One has to consider the Application.
Yes, any caps can be paralleled and will the store a larger quantity of electrical charge, to the sum of their total capacitance, but that does not indicate how the charging occurs, nor the implication of the charging/discharging process in any particular application.
It is worthwhile to study texts where caps are applied critically to learn more about their behaviour.
Audio is NOT regarded as CRITICAL to anything except the transmission and reception of speech information, beacause the general public will accept mediocrity, and thus are sold such.
Study R.F. applications of caps. Anything where there is application to Military or Aerospace is regarded as Critical, and from there one can learn things to apply to Audio.
Pulse Rise Time is basically the response time of the cap to the signal.
It is dependant on the characteristics of the materials used in the caps, including the type of Dielectric and its thickness in application, and on the surface area the electrical charge has to spread on the metal film or foil.
Caps with "extended foil" design have electrical contact along the entire outer edges of their metal films or foils, thus only the width of the film or foil is then the determining dimension.
In Axial caps, such as all the Cross-over caps' brands we have mentioned in this Thread, the width of film or foil is basically almost the full length of the capacitior.
Other relevant determinants of Pulse Rise Time can be included by simply using the same model caps - that is same dielectric type and internal construction and of the same voltage rating and from the same Manufacturer.
Thus for equal Pulse Rise times in all caps in a parallel bundle use caps from the above that are of the same axial length.
If significantly different Pulse Rise times through a bundle of parallel caps the highest frequency components of the signal lead the lower frequency components and the sound will be as if spitting or sibilant or if an undamped resonance in the tweeter .
Yes, some listeners have blamed tweeters for audible fault when the cause has been parallel caps to it of different pulse rise times.
Some DIYers keep changing tweeters ,or caps presumed "microphonic", without ever realizing the true cause of the audible fault.
Have to go now - more on this later.
Before I Post any more about specific caps for options for the x-over I am going to address the subject of connecting Capacitors in Parallel as it is not well understood in critical Audio applications.
In Basic Electronics one learns that any two or more caps can be connected in parallel to increase the over-all capacitance to the sum of those connected, but that is simplistic by itself.
One has to consider the Application.
Yes, any caps can be paralleled and will the store a larger quantity of electrical charge, to the sum of their total capacitance, but that does not indicate how the charging occurs, nor the implication of the charging/discharging process in any particular application.
It is worthwhile to study texts where caps are applied critically to learn more about their behaviour.
Audio is NOT regarded as CRITICAL to anything except the transmission and reception of speech information, beacause the general public will accept mediocrity, and thus are sold such.
Study R.F. applications of caps. Anything where there is application to Military or Aerospace is regarded as Critical, and from there one can learn things to apply to Audio.
Pulse Rise Time is basically the response time of the cap to the signal.
It is dependant on the characteristics of the materials used in the caps, including the type of Dielectric and its thickness in application, and on the surface area the electrical charge has to spread on the metal film or foil.
Caps with "extended foil" design have electrical contact along the entire outer edges of their metal films or foils, thus only the width of the film or foil is then the determining dimension.
In Axial caps, such as all the Cross-over caps' brands we have mentioned in this Thread, the width of film or foil is basically almost the full length of the capacitior.
Other relevant determinants of Pulse Rise Time can be included by simply using the same model caps - that is same dielectric type and internal construction and of the same voltage rating and from the same Manufacturer.
Thus for equal Pulse Rise times in all caps in a parallel bundle use caps from the above that are of the same axial length.
If significantly different Pulse Rise times through a bundle of parallel caps the highest frequency components of the signal lead the lower frequency components and the sound will be as if spitting or sibilant or if an undamped resonance in the tweeter .
Yes, some listeners have blamed tweeters for audible fault when the cause has been parallel caps to it of different pulse rise times.
Some DIYers keep changing tweeters ,or caps presumed "microphonic", without ever realizing the true cause of the audible fault.
Have to go now - more on this later.
CORRECTION and ADDITIONS to last Post
... "and will the store" ... should be :-
and will THEN store a larger quantity of electrical charge ...
____________________________________________________
PULSE RISE TIMES , continued :-
For "parallel" connection I will use the symbol // between the cap values - this is a convention in Electronics.
We see 4.7uF // 1.5uF and 3.3uF//.68uF in the later 66's x-over in the tweeter filter - two pairs of different physical size Radial caps - different pulse rise times.
Why are they there if audibly inferior to the equal length caps in the parallel bundles in the older 66s ?
I suspect the reason is simply Price of caps - cheaper is to use what is readily available, and fewer pieces -
Four of 1.5uF cost more than one 4.7 + one 1.5.
After Plastic Film caps were introduced to the market, the available capacitance values were Standardized, and the E-Series values were the Preferred Values for Capacitor, and Resistor, manufacturers to make, thus no more 2uF , etc ... in the new plastic film caps unless one was prepared to pay for a very large batch to be manufactured of a particular cap value.
For large cap values this remained the case for about 20 years, until the sufficient interest in DIY and Hi-End Audio Maunufacturers' requests for a greater range of cap values provided market opportunity for some of the cap manufacturers to issue specialist Audio Capacitor series, such as the Solens made in France by SCR.
{"Solen" brand is Canadian owned - the same SCR manufactured caps were sold in France with the "Chateauroux" brand on them.}
Also - those tweeter filter caps in the later 66s are Polyester dielectric -{probably "Mylar" a Trade-name owned by DuPont}.
Pulse Rise Time for Polyester dielectric is about ten time slower than for Polypropylene dielectric - if using the same type of internal construction for both caps,
AND,
Polyester caps have significantly higher dielectric absorption than Polypropylene caps, and quite audibly so, and this muddies the sound sufficiently to not hear different parallel pulse rise times as distinctly.
I doubt the specific audio effect of significantly different pulse rise times of parallel connected caps would be as obvious with Polyester caps as with Polypropylene caps, distinct from the overall clearer sound of the Polypropylenes.
Paper-in-Oil caps went out of use in those cap sizes in the older 66s because they are much less long term reliable than Polyester, thus no longer used in Industrial applications, etc ... , hence not manufactured much for many years until a few audio enthusiasts noticed the better sound they could get using oil filled Motor-run caps instead of Electrotytics, and then they tried the oil-filled caps in other audio applications, and perhaps reconditioned a few old paper-in-oil caps, and liking what they heard they requested manufacture, but the new manufacture paper-in-oil caps - referred to as PIOs - are very expensive, as they are not used in Industrial applications thus only small quantities are made, and they are likely still not long-life.
So, did the older 66s sound better than the later 66s ?
Well, one would have to ask the enthusiasts who heard both at the time of the change.
With the later tweeter of probably better quality, and the new MD500 possibly better quality, and certainly capable of being played a bit louder before distortion than the older MF500, the two versions of the 66 may have simply sounded different to each other, with neither necessarily being better, except for personal preference perhaps based on the type of music played and how loudly.
sba's old series "blackie" 66s had significantly deteriorated tweeter caps, and both his 66 versions had deteriorating electrotytic caps, thus neither were a sample of what either speaker sounded like when new.
My intention is to enable both versions of 66 to be restored to suitable condition to hear the quality of the drivers when crossed over at the frequencies Celestion decided were optimum for those drivers, and this includes the Slopes of the filters and the small differences in Sensitivity between the drivers, and the overall audio character of the complete loudspeaker - this final aspect is set by the series resistances of the two large inductors in the bass filter.
This is why I've written at length about parallel connected caps and ESR simulating resistors.
I hope I'll get to the specific caps and resistors in my next Post !
alan-1-b said:
... "and will the store" ... should be :-
and will THEN store a larger quantity of electrical charge ...
____________________________________________________
PULSE RISE TIMES , continued :-
For "parallel" connection I will use the symbol // between the cap values - this is a convention in Electronics.
We see 4.7uF // 1.5uF and 3.3uF//.68uF in the later 66's x-over in the tweeter filter - two pairs of different physical size Radial caps - different pulse rise times.
Why are they there if audibly inferior to the equal length caps in the parallel bundles in the older 66s ?
I suspect the reason is simply Price of caps - cheaper is to use what is readily available, and fewer pieces -
Four of 1.5uF cost more than one 4.7 + one 1.5.
After Plastic Film caps were introduced to the market, the available capacitance values were Standardized, and the E-Series values were the Preferred Values for Capacitor, and Resistor, manufacturers to make, thus no more 2uF , etc ... in the new plastic film caps unless one was prepared to pay for a very large batch to be manufactured of a particular cap value.
For large cap values this remained the case for about 20 years, until the sufficient interest in DIY and Hi-End Audio Maunufacturers' requests for a greater range of cap values provided market opportunity for some of the cap manufacturers to issue specialist Audio Capacitor series, such as the Solens made in France by SCR.
{"Solen" brand is Canadian owned - the same SCR manufactured caps were sold in France with the "Chateauroux" brand on them.}
Also - those tweeter filter caps in the later 66s are Polyester dielectric -{probably "Mylar" a Trade-name owned by DuPont}.
Pulse Rise Time for Polyester dielectric is about ten time slower than for Polypropylene dielectric - if using the same type of internal construction for both caps,
AND,
Polyester caps have significantly higher dielectric absorption than Polypropylene caps, and quite audibly so, and this muddies the sound sufficiently to not hear different parallel pulse rise times as distinctly.
I doubt the specific audio effect of significantly different pulse rise times of parallel connected caps would be as obvious with Polyester caps as with Polypropylene caps, distinct from the overall clearer sound of the Polypropylenes.
Paper-in-Oil caps went out of use in those cap sizes in the older 66s because they are much less long term reliable than Polyester, thus no longer used in Industrial applications, etc ... , hence not manufactured much for many years until a few audio enthusiasts noticed the better sound they could get using oil filled Motor-run caps instead of Electrotytics, and then they tried the oil-filled caps in other audio applications, and perhaps reconditioned a few old paper-in-oil caps, and liking what they heard they requested manufacture, but the new manufacture paper-in-oil caps - referred to as PIOs - are very expensive, as they are not used in Industrial applications thus only small quantities are made, and they are likely still not long-life.
So, did the older 66s sound better than the later 66s ?
Well, one would have to ask the enthusiasts who heard both at the time of the change.
With the later tweeter of probably better quality, and the new MD500 possibly better quality, and certainly capable of being played a bit louder before distortion than the older MF500, the two versions of the 66 may have simply sounded different to each other, with neither necessarily being better, except for personal preference perhaps based on the type of music played and how loudly.
sba's old series "blackie" 66s had significantly deteriorated tweeter caps, and both his 66 versions had deteriorating electrotytic caps, thus neither were a sample of what either speaker sounded like when new.
My intention is to enable both versions of 66 to be restored to suitable condition to hear the quality of the drivers when crossed over at the frequencies Celestion decided were optimum for those drivers, and this includes the Slopes of the filters and the small differences in Sensitivity between the drivers, and the overall audio character of the complete loudspeaker - this final aspect is set by the series resistances of the two large inductors in the bass filter.
This is why I've written at length about parallel connected caps and ESR simulating resistors.
I hope I'll get to the specific caps and resistors in my next Post !
Hi All.
Thanks for the e-mails Adrian. I don't dip in here very often.
That Page I posted was a Celection advert on the back of an early '70s publication. Being the back and very well thumbed it is dirty, scratched and crumpled. I will see if I can re-copy it and upload to a hosting website in better detail for others to download, but I don't have any other info on the loudspeakers. I can't even remember any leaflet being in the box with them.
I believe Hi-Fi News did a review, but I don't think it was very complementary.
They were brilliant for Pop Parties. Again in the early '70s I even saw a pair of 66's mounted vertically above reaching height in the well appointed Officer's bar at RAF Aldergrove (NW of Belfast in N. Ireland) - must have made formidable sound.
Re post#196.
That crossover layout could lead to an audible inductive coupling between the windings, so maybe the 0.3mH coil should be rotated by 90 degrees with respect to those either side of it.
Someone asked about my Ditton 66s.
I did not like the sound of the mid-range integration with the original crossover, so both were removed almost immediately. I fitted a 6.5FR in place of the mid (just drops in) and built my own crossover.
My 66s were early versions, and when I went to re-use the parts to build my own crossover I was disappointed with their specifications.
For me, the coils had too high resistance, and the capacitors were too lossy for the frequencies/impedances they were supposed to be coping with.
Hence I am so glad to see these points being addressed via the on-going efforts of subscribers here so that the drivers can be properly driven and crossed over.
When a bass/ABR arrangement is used the coils in series with the main driver need to have as low R as possible so that the amplifier does the driving and damping, and not the cabinet air motion. My choice here was for thick wire and ferrite cored.
Cheers for now ............ Graham.
Thanks for the e-mails Adrian. I don't dip in here very often.
That Page I posted was a Celection advert on the back of an early '70s publication. Being the back and very well thumbed it is dirty, scratched and crumpled. I will see if I can re-copy it and upload to a hosting website in better detail for others to download, but I don't have any other info on the loudspeakers. I can't even remember any leaflet being in the box with them.
I believe Hi-Fi News did a review, but I don't think it was very complementary.
They were brilliant for Pop Parties. Again in the early '70s I even saw a pair of 66's mounted vertically above reaching height in the well appointed Officer's bar at RAF Aldergrove (NW of Belfast in N. Ireland) - must have made formidable sound.
Re post#196.
That crossover layout could lead to an audible inductive coupling between the windings, so maybe the 0.3mH coil should be rotated by 90 degrees with respect to those either side of it.
Someone asked about my Ditton 66s.
I did not like the sound of the mid-range integration with the original crossover, so both were removed almost immediately. I fitted a 6.5FR in place of the mid (just drops in) and built my own crossover.
My 66s were early versions, and when I went to re-use the parts to build my own crossover I was disappointed with their specifications.
For me, the coils had too high resistance, and the capacitors were too lossy for the frequencies/impedances they were supposed to be coping with.
Hence I am so glad to see these points being addressed via the on-going efforts of subscribers here so that the drivers can be properly driven and crossed over.
When a bass/ABR arrangement is used the coils in series with the main driver need to have as low R as possible so that the amplifier does the driving and damping, and not the cabinet air motion. My choice here was for thick wire and ferrite cored.
Cheers for now ............ Graham.
Inductive Coupling
Hi Graham,
indeed I expect some degree of inductive coupling in that B&O x-over - I would never have used that layout, and I'm surprised that B&O did, but perhaps their design engineer didn't get any say about how the production department placed components ...
Did you see the Link sba posted in #204 ?
{great stuff sba ! - excellent that you posted a Link to that research and results on Inductor layout}
However, there will be some degree of inductive coupling in the Celestion 66 x-over also.
I wondered about it from the first, when its photos were Linked to this thread.
Its probably not excessive in this case, AND, the Celestion design engineer will almost certainly have known about it.
Perhaps he included the coupling effect into filter turnover frequencies and slopes.
Such could be got to work to some degree for the 500 Hz filters, but I wouldn't want the low mids from those coupled into the 5kHz filters !
I suspect it would have been for cost cutting, to rely on coupling instead of using larger inductors.
As we do not know the designer's intention I recommend the inductors' layout is not changed if one makes new boards, but if you want to space the inductors further apart, do so but ENSURE that you do not reverse the coupling co-efficients .
Simplest way to not reverse those is to re-lay the inductors exactly the same direction - tops facing up as now, bottoms on the board as now, AND, connect the end leads exactly the same as were to the capacitors and input links - do not reverse the connections to the inductors.
It would be of benefit to reduce any coupling between the 500Hz filters and the 5kHz filters, thus place the 2 smallest inductors further away from the 3 largest inductors, and don't put any closer to each other than they are on the boards you have now.
____________________________________________________
Graham,
the 6.5FR - do you mean you put in a 6.5" Full Range driver, or is 6.5FR a particular Manufacturer's Model Number ?
Whose driver did you use ?
This may be of some assistance to later readers who have non-working mid-domes.
As you say yours were early versions I think you will have the MF-500/T.1806 domes ... ?
Notice that B&O cross that MF version over at a higher frequency than Celestion did - pehaps it sounded better with a little more low mids removed, and such would have improved its power handling -
{perhaps why those sba found in the overdriven 5700s survived}.
sba can try his "blackies" with the lower x-over and report on the sound in comparison to how he currently has them.
Put in 30uF temporarily, with a parallel pair of your 15uF caps there sba, and with a 1.5 ohm Series connected resistor-
but if you haven't bought any resistors yet, wait awhile till I get to Posting more about those.
I think the older series 66s' MF-500s may benefit from crossing over a little higher using 24uF as for the later MD-500, and adding a Series connected resistor to the closest 72uF cap to the woofer to raise the midrange output of it a little to compensate.
I'll Post more about that when I post about the caps and resistors.
____________________________________________________
Graham,
I also think the coils in the 500Hz filters have too much resistance for accurate reproduction of the input signal, but, to reduce resistance one has to use physically larger coils, or use ferrite or metal cored coils, and those will saturate and produce compression and distortion effects with signal peaks and general high volume levels.
rwthomkins reported the sound of cored coils in his comments about his Celestion 44s - Post #177 on Page 8.
Yes, thick wire for alternate woofer filter coils, but if Ferrite cored they will still need to be large size so as to not too soon saturate when signal levels increase.
OK Graham, tell me, did you get satisfactory sound with what physically sized coils compared to the air-cores in the original ?
I completely agree with your final point about the bass drivers and the ABRs and the amplifier - and I should have included that in may last Post's final paragraph about the audio character of the complete loudspeaker - I was too brief there.
{me, brief ? - rarely !}
Hi Graham,
indeed I expect some degree of inductive coupling in that B&O x-over - I would never have used that layout, and I'm surprised that B&O did, but perhaps their design engineer didn't get any say about how the production department placed components ...
Did you see the Link sba posted in #204 ?
{great stuff sba ! - excellent that you posted a Link to that research and results on Inductor layout}
However, there will be some degree of inductive coupling in the Celestion 66 x-over also.
I wondered about it from the first, when its photos were Linked to this thread.
Its probably not excessive in this case, AND, the Celestion design engineer will almost certainly have known about it.
Perhaps he included the coupling effect into filter turnover frequencies and slopes.
Such could be got to work to some degree for the 500 Hz filters, but I wouldn't want the low mids from those coupled into the 5kHz filters !
I suspect it would have been for cost cutting, to rely on coupling instead of using larger inductors.
As we do not know the designer's intention I recommend the inductors' layout is not changed if one makes new boards, but if you want to space the inductors further apart, do so but ENSURE that you do not reverse the coupling co-efficients .
Simplest way to not reverse those is to re-lay the inductors exactly the same direction - tops facing up as now, bottoms on the board as now, AND, connect the end leads exactly the same as were to the capacitors and input links - do not reverse the connections to the inductors.
It would be of benefit to reduce any coupling between the 500Hz filters and the 5kHz filters, thus place the 2 smallest inductors further away from the 3 largest inductors, and don't put any closer to each other than they are on the boards you have now.
____________________________________________________
Graham,
the 6.5FR - do you mean you put in a 6.5" Full Range driver, or is 6.5FR a particular Manufacturer's Model Number ?
Whose driver did you use ?
This may be of some assistance to later readers who have non-working mid-domes.
As you say yours were early versions I think you will have the MF-500/T.1806 domes ... ?
Notice that B&O cross that MF version over at a higher frequency than Celestion did - pehaps it sounded better with a little more low mids removed, and such would have improved its power handling -
{perhaps why those sba found in the overdriven 5700s survived}.
sba can try his "blackies" with the lower x-over and report on the sound in comparison to how he currently has them.
Put in 30uF temporarily, with a parallel pair of your 15uF caps there sba, and with a 1.5 ohm Series connected resistor-
but if you haven't bought any resistors yet, wait awhile till I get to Posting more about those.
I think the older series 66s' MF-500s may benefit from crossing over a little higher using 24uF as for the later MD-500, and adding a Series connected resistor to the closest 72uF cap to the woofer to raise the midrange output of it a little to compensate.
I'll Post more about that when I post about the caps and resistors.
____________________________________________________
Graham,
I also think the coils in the 500Hz filters have too much resistance for accurate reproduction of the input signal, but, to reduce resistance one has to use physically larger coils, or use ferrite or metal cored coils, and those will saturate and produce compression and distortion effects with signal peaks and general high volume levels.
rwthomkins reported the sound of cored coils in his comments about his Celestion 44s - Post #177 on Page 8.
Yes, thick wire for alternate woofer filter coils, but if Ferrite cored they will still need to be large size so as to not too soon saturate when signal levels increase.
OK Graham, tell me, did you get satisfactory sound with what physically sized coils compared to the air-cores in the original ?
I completely agree with your final point about the bass drivers and the ABRs and the amplifier - and I should have included that in may last Post's final paragraph about the audio character of the complete loudspeaker - I was too brief there.
{me, brief ? - rarely !}
ADDITION to Post #231
I'll separate the passage in the Quote Box that I should have been more comprehensive about, and type a more comprehensive description below the Box.
-----> and the overall audio character of the complete loudspeaker -
this final aspect is set by the series resistances of the two large inductors in the bass filter and specifically their amount of resistance in relation to the fairly low DC resistance of the woofers, and the subsequent fairly low Impedance of the woofers, along with the audio effect of the ABRs themselves as well as how they are controlled by the woofers - less so in this case than what is considered optimum,
{but some listeners like the sound that produces !}.
I'll separate the passage in the Quote Box that I should have been more comprehensive about, and type a more comprehensive description below the Box.
alan-1-b said:
-----> and the overall audio character of the complete loudspeaker -
this final aspect is set by the series resistances of the two large inductors in the bass filter and specifically their amount of resistance in relation to the fairly low DC resistance of the woofers, and the subsequent fairly low Impedance of the woofers, along with the audio effect of the ABRs themselves as well as how they are controlled by the woofers - less so in this case than what is considered optimum,
{but some listeners like the sound that produces !}.
initial audition
Alan
Have incorporated 5W Kwame resistors to substitute for ESR as you recommended for my "blackie" Ditton 66s with MF500 mid driver:
100R across the tweeter
1R5 in series with the 24 microF
1R2 in series with the first 72 microF
1R5 in series with the second 72 microF
As Alan recommended, the 24 microF was made up of two parallel arms, each containing a 12 microF and 3R in series rather than "commoning" both capacitors together in series with both resistors together
Have replaced the original capacitors by Solen metallised polypropylene capacitors, all internal wiring, rearranged components within the existing tagboard, new terminals for biwiring, potted the inductors, applied Dynamat Xtreme to all internal surfaces, and pinned a fat sock around the top lip of the cabinet to curb reflections with the tweeter nearby.
Initial reactions: the sound is cleaner and has more delicacy, less of the cuppy muffled upper range yet perhaps a tad less of the enjoyable warmth! Maybe because the MF500 is receiving more drive at the lower end plus the upper range of the bass driver is also increased. However I will try them with other kit which is likely will stiffen up the bottom end.
One speaker sounds too trebly and suspect that I damaged the 3.9 microF capacitor by a tiny pit while soldering - fortunate to have a spare. (Put the crossover together twice - once with Welwyn resistors then with Kwame resistors - and the board is rather cramped. Rusty at soldering.)
---
BTW The Ditton 33 crossover has a 2R2 Welwyn resistor before the signal goes to the mid section of the crossover. (The crossovers for Dittons 33, 44 and 66 are otherwise similar in topology.)
Alan
Have incorporated 5W Kwame resistors to substitute for ESR as you recommended for my "blackie" Ditton 66s with MF500 mid driver:
100R across the tweeter
1R5 in series with the 24 microF
1R2 in series with the first 72 microF
1R5 in series with the second 72 microF
As Alan recommended, the 24 microF was made up of two parallel arms, each containing a 12 microF and 3R in series rather than "commoning" both capacitors together in series with both resistors together
Have replaced the original capacitors by Solen metallised polypropylene capacitors, all internal wiring, rearranged components within the existing tagboard, new terminals for biwiring, potted the inductors, applied Dynamat Xtreme to all internal surfaces, and pinned a fat sock around the top lip of the cabinet to curb reflections with the tweeter nearby.
Initial reactions: the sound is cleaner and has more delicacy, less of the cuppy muffled upper range yet perhaps a tad less of the enjoyable warmth! Maybe because the MF500 is receiving more drive at the lower end plus the upper range of the bass driver is also increased. However I will try them with other kit which is likely will stiffen up the bottom end.
One speaker sounds too trebly and suspect that I damaged the 3.9 microF capacitor by a tiny pit while soldering - fortunate to have a spare. (Put the crossover together twice - once with Welwyn resistors then with Kwame resistors - and the board is rather cramped. Rusty at soldering.)
---
BTW The Ditton 33 crossover has a 2R2 Welwyn resistor before the signal goes to the mid section of the crossover. (The crossovers for Dittons 33, 44 and 66 are otherwise similar in topology.)
As expected, the 3.9 microF needed replacing.
---
Sound is neutral and pleasant. No objection if would sweeten or warm slightly. Maybe this will come as the capacitors break in (if that process indeed occurs and isn't just the listener adapting to the sound presentation
).
Maybe bass unit rolls off a touch too early.
---
Some absence of "air". Cymbals are dry and truncated. May try bipassing the inline capacitors with Vishays and remove the 100R. Sorry Alan.
The 1R5 in series with the MF500 is higher than original ESR at 5kHz of ~ 0R7. This reduces slightly the upper response of the MF500 relative to the HF2000. Considering replacing the 2.4 microF (in parallel with 3.9 microF) with 1.8 microF, to roll in the tweeter further up. (Earlier crossover, for the MF500, had 6 microF while the later had 6.2 microF.)
---
Sound is neutral and pleasant. No objection if would sweeten or warm slightly. Maybe this will come as the capacitors break in (if that process indeed occurs and isn't just the listener adapting to the sound presentation
). Maybe bass unit rolls off a touch too early.
---
Some absence of "air". Cymbals are dry and truncated. May try bipassing the inline capacitors with Vishays and remove the 100R. Sorry Alan.
The 1R5 in series with the MF500 is higher than original ESR at 5kHz of ~ 0R7. This reduces slightly the upper response of the MF500 relative to the HF2000. Considering replacing the 2.4 microF (in parallel with 3.9 microF) with 1.8 microF, to roll in the tweeter further up. (Earlier crossover, for the MF500, had 6 microF while the later had 6.2 microF.)
Good result in part, and a tweek for more .
Hi harnfield,
I'm very pleased you have heard some improvements to the sound, as one would hope to be worth all the work you've done !
The improvements you describe to the upper range are as I expected.
But, regarding the - "tad less of the enjoyable warmth" - no, the MF500 will not be receiving more drive at the lower end, it will be receiving less because 24uF will be rolling off low-mids at a higher frequency than 30uF did in the originals, or than a charge leaking old electro-cap of whatever was in yours,
{24uF or 28uF or 30uF .. ? .. - I apologise, I have forgotten if you posted what was in there}.
Increasing from 24uF will reduce power handling, and result may damage the now old MF500s more readily than when they were new, thus unless you have a small room, or only intend to push to moderate volume level in a large room, I recommend you keep the 24uF at mids, but add some more output from the woofers to fill in the gap.
Upper end from the woofers will not have increased, because those series resistors are only doing what the ESR in the old 72uF caps did.
To increase the midrange output from the woofers larger series resistance needs to be added to C6 - the cap electrically closest to woofer, {but NOT to the other 72uF - C8 , because that one is filtering upper mids that one does not want to have in the woofer}.
In your case of parallel 36uF// 36uF for 72uF :-
(1) I thought you put parallel 2.2ohm Kiwames to C8, for effective 1.1ohm .. ? .. , but 1.2 will not be a major problem there.
(2) If I understand correctly you have parallel 3ohm pair for total 1.5ohm at C6 ...
if so, remove one of the 3R Kiwames, and have simply one 3ohm Kiwame there with the parallel pair of caps both connected to the resistor - 5 watts is sufficient as there will not be much current through it in that location.
This small real increase in resistance there will lift the midrange output a little in the 500Hz region and above and restore a little warmth.
You can increase resistance there a little more if still not sufficient improvement.
Maximum effective resistance for this type of tweek is equal to the voice-coil resistance of the woofer, which given the measurements of it posted by others in this thread is about 4.3ohms.
Using more than the voice-coil DCR may cause some higher frequency output from the woofer that is currently reproduced adequately by the MF500 and may simply over-accentuate that portion of the frequency spectrum.
This type of tweeking for fine tuning is best done by ear, thus listen with 3 ohms, and if not sufficient then experiment by using your Welwyns till you get the sound you want, before you buy more Kiwames.
There is a 4.3R Kiwame manufactured, but I've forgotten if on that UK sellers list.
The less number of Series solder joints, the better the sound, thus a single 4.3ohm Kiwame will probably give better result than 3ohm plus a spare 1R2 in series with it, but series is of course OK when one is experimenting to get close to preferred result.
____________________________________________________
Hah, I see you have Posted another whilst I am still typing this one , well I'll include comment about it also, but first I am please the initial teble fault was simply the 3.9uF cap
- burning a hole in a film cap will reduce its capacitance, thus cause phase change and resulting interference around the intended cross-over frequency.
{I am guessing it is the 3.9uF in parallel with the MF500 that was damaged ? ... but even if it was a 3.9uF in the tweeter filter it will still have caused some lower treble loss thus upper treble stands out a bit}.
But, NO, reducing from the parallel 2.4 to 1.8uF will not help, as that will cause more loss of out put to tweeter just above intended cross-over frequency.
Removing the 100R will give a little more treble output in part of the spectrum, so try that before you change from 2.4uF to 1.8
{if you want to hear what such change will cause, and any experiment that increases one's knowledge can be worthwhile}.
Try all the above before you add any Vishays - those create a sound effect, not reproduce the actual signal sent by the amp.
Post your results, and if still not happy I'll give some thought to the resistor in series with the MF500s.
___________________________________________________
Do you have a schematic for the Ditton 33 x-over ?
If so, can you Post it or a LInk to it, as there may be something of interest in it that we can use in this 6 project.
Hi harnfield,
I'm very pleased you have heard some improvements to the sound, as one would hope to be worth all the work you've done !
The improvements you describe to the upper range are as I expected.
But, regarding the - "tad less of the enjoyable warmth" - no, the MF500 will not be receiving more drive at the lower end, it will be receiving less because 24uF will be rolling off low-mids at a higher frequency than 30uF did in the originals, or than a charge leaking old electro-cap of whatever was in yours,
{24uF or 28uF or 30uF .. ? .. - I apologise, I have forgotten if you posted what was in there}.
Increasing from 24uF will reduce power handling, and result may damage the now old MF500s more readily than when they were new, thus unless you have a small room, or only intend to push to moderate volume level in a large room, I recommend you keep the 24uF at mids, but add some more output from the woofers to fill in the gap.
Upper end from the woofers will not have increased, because those series resistors are only doing what the ESR in the old 72uF caps did.
To increase the midrange output from the woofers larger series resistance needs to be added to C6 - the cap electrically closest to woofer, {but NOT to the other 72uF - C8 , because that one is filtering upper mids that one does not want to have in the woofer}.
In your case of parallel 36uF// 36uF for 72uF :-
(1) I thought you put parallel 2.2ohm Kiwames to C8, for effective 1.1ohm .. ? .. , but 1.2 will not be a major problem there.
(2) If I understand correctly you have parallel 3ohm pair for total 1.5ohm at C6 ...
if so, remove one of the 3R Kiwames, and have simply one 3ohm Kiwame there with the parallel pair of caps both connected to the resistor - 5 watts is sufficient as there will not be much current through it in that location.
This small real increase in resistance there will lift the midrange output a little in the 500Hz region and above and restore a little warmth.
You can increase resistance there a little more if still not sufficient improvement.
Maximum effective resistance for this type of tweek is equal to the voice-coil resistance of the woofer, which given the measurements of it posted by others in this thread is about 4.3ohms.
Using more than the voice-coil DCR may cause some higher frequency output from the woofer that is currently reproduced adequately by the MF500 and may simply over-accentuate that portion of the frequency spectrum.
This type of tweeking for fine tuning is best done by ear, thus listen with 3 ohms, and if not sufficient then experiment by using your Welwyns till you get the sound you want, before you buy more Kiwames.
There is a 4.3R Kiwame manufactured, but I've forgotten if on that UK sellers list.
The less number of Series solder joints, the better the sound, thus a single 4.3ohm Kiwame will probably give better result than 3ohm plus a spare 1R2 in series with it, but series is of course OK when one is experimenting to get close to preferred result.
____________________________________________________
Hah, I see you have Posted another whilst I am still typing this one , well I'll include comment about it also, but first I am please the initial teble fault was simply the 3.9uF cap
- burning a hole in a film cap will reduce its capacitance, thus cause phase change and resulting interference around the intended cross-over frequency.
{I am guessing it is the 3.9uF in parallel with the MF500 that was damaged ? ... but even if it was a 3.9uF in the tweeter filter it will still have caused some lower treble loss thus upper treble stands out a bit}.
But, NO, reducing from the parallel 2.4 to 1.8uF will not help, as that will cause more loss of out put to tweeter just above intended cross-over frequency.
Removing the 100R will give a little more treble output in part of the spectrum, so try that before you change from 2.4uF to 1.8
{if you want to hear what such change will cause, and any experiment that increases one's knowledge can be worthwhile}.
Try all the above before you add any Vishays - those create a sound effect, not reproduce the actual signal sent by the amp.
Post your results, and if still not happy I'll give some thought to the resistor in series with the MF500s.
___________________________________________________
Do you have a schematic for the Ditton 33 x-over ?
If so, can you Post it or a LInk to it, as there may be something of interest in it that we can use in this 6 project.
My "woodies" the crossover had only 24 microF, not 28 or 30 microF in line with the MF500.
Possibly values changed according to make of capacitors selected - these ELCAPS were black with red "caps", the 24 microF was CCL 25V reversible WE 99RAS JAN73, and the 4 & 6 microF for the tweeter were made up from banks of Metallised 2 microF caps.
---
Am wondering if the sound is not as warm as I like because I need to reduce the level of signal arriving at the mid driver. (I dampened the wall panels that may have been lending warmth.)
---
All voices sound clear, a bit similar and a little more forward than it need be - this could be explained if there is too much overlap just below or around 5kHz, which why am considering reducing the 6.3 microF to 5.7, shifting up the roll in frequency of the tweeter. (I seem to recall that "forwardness" can be due to emphasis/recess in a band around 2kHz to 3kHz.)
---
Have not changed the yellow foam that lined the cavity. Possibly other material would absorb better.
Possibly values changed according to make of capacitors selected - these ELCAPS were black with red "caps", the 24 microF was CCL 25V reversible WE 99RAS JAN73, and the 4 & 6 microF for the tweeter were made up from banks of Metallised 2 microF caps.
---
Am wondering if the sound is not as warm as I like because I need to reduce the level of signal arriving at the mid driver. (I dampened the wall panels that may have been lending warmth.)
---
All voices sound clear, a bit similar and a little more forward than it need be - this could be explained if there is too much overlap just below or around 5kHz, which why am considering reducing the 6.3 microF to 5.7, shifting up the roll in frequency of the tweeter. (I seem to recall that "forwardness" can be due to emphasis/recess in a band around 2kHz to 3kHz.)
---
Have not changed the yellow foam that lined the cavity. Possibly other material would absorb better.
Changes -
Yes, the resonance of the wall panels would have added the impression of warmth to the sound.
Being damped by added mass, the walls will now resonate lower in the frequency spectrum, and probably at lower amplitude, and it seems this is the case from your description of the sound.
As there was 24uF to MF500 in your interim model 66, I think it is worth trying the single 3ohm to C6 36uF//pair to add a little warmth.
This will not sound like the old resonant panels' warmth though.
The forwardness will be from the mid-domes, as result of much clearer sound through the new caps.
Obviously 1.5ohms there is not too much resistance, but listen again after you try the resistor to C6 change.
The purpose of the 100ohm resistor in parallel with the tweeter was to ensure not too much overlap around the 5kHz region.
Removing this resistor, as you wanted to do, and substituting 1.8uF for the 2.4uF will push a little further the effect that is currently heard with the resistor and the 2.4 in,
but, you may as well try it to satisfy your curiosity, and you might hear a small change you like, but I doubt it will reduce "forwardness" of human voices.
Further reduction of 5kHz region overlap may be better achieved by reducing the resistance that is in Series with C5/3.9uF {in parallel with the mid-dome}.
I think you have 3.9ohms there now.
Try 2.2ohm or 2.4ohm if no spare 2.2s currently.
This is to be tried with the 2.4uF and 100ohm in the tweeter circuit.
If you don't like the sound with lower ohms to the 3.9uF, then put the 3.9ohm resistor back in before you try the change to 1.8uF in the tweeter circuit.
Try the resistor change in woofer circuit first, then which-ever of the mid or treble changes you want to try, with the woofer resistor in, if you liked the sound of that change.
I am keeping in mind your sound preference you described in the first paragraph of your Post #178 on Page 8.
Also, I hope you know that you will never obtain the presentation of the midrange and treble of your Apogee "Scintillas" from the 66s no matter how many mods you try, nor with any expensive caps and resistors.
I think you do know that, but I have to include this because some people do think such things can be achieved !
____________________________________________________
If the yellow foam is not crumbling or otherways falling apart, and hasn't gone sticky to touch, then it will be still OK.
Better absorption can be achieved with a mineral fibre wool, and almost as good as that with Fibreglass, but both materials are slightly hazardous to work with - may irritate skin and will irritate eyes and can damage the surface of eyes,
{wear gloves, and goggles if you work close to it in the cabinet}.
Long-haired wool is safe, and better than plastic foam, but wool if in batts and the mineral and glass fibre products need to be optimally aligned in the enclosure to absorb optimally, and to not reduce efficiency of the woofer by mass loading if they are free to flop about inside whilst the woofer is pumping !
Yes, the resonance of the wall panels would have added the impression of warmth to the sound.
Being damped by added mass, the walls will now resonate lower in the frequency spectrum, and probably at lower amplitude, and it seems this is the case from your description of the sound.
As there was 24uF to MF500 in your interim model 66, I think it is worth trying the single 3ohm to C6 36uF//pair to add a little warmth.
This will not sound like the old resonant panels' warmth though.
The forwardness will be from the mid-domes, as result of much clearer sound through the new caps.
Obviously 1.5ohms there is not too much resistance, but listen again after you try the resistor to C6 change.
The purpose of the 100ohm resistor in parallel with the tweeter was to ensure not too much overlap around the 5kHz region.
Removing this resistor, as you wanted to do, and substituting 1.8uF for the 2.4uF will push a little further the effect that is currently heard with the resistor and the 2.4 in,
but, you may as well try it to satisfy your curiosity, and you might hear a small change you like, but I doubt it will reduce "forwardness" of human voices.
Further reduction of 5kHz region overlap may be better achieved by reducing the resistance that is in Series with C5/3.9uF {in parallel with the mid-dome}.
I think you have 3.9ohms there now.
Try 2.2ohm or 2.4ohm if no spare 2.2s currently.
This is to be tried with the 2.4uF and 100ohm in the tweeter circuit.
If you don't like the sound with lower ohms to the 3.9uF, then put the 3.9ohm resistor back in before you try the change to 1.8uF in the tweeter circuit.
Try the resistor change in woofer circuit first, then which-ever of the mid or treble changes you want to try, with the woofer resistor in, if you liked the sound of that change.
I am keeping in mind your sound preference you described in the first paragraph of your Post #178 on Page 8.
Also, I hope you know that you will never obtain the presentation of the midrange and treble of your Apogee "Scintillas" from the 66s no matter how many mods you try, nor with any expensive caps and resistors.
I think you do know that, but I have to include this because some people do think such things can be achieved !
____________________________________________________
If the yellow foam is not crumbling or otherways falling apart, and hasn't gone sticky to touch, then it will be still OK.
Better absorption can be achieved with a mineral fibre wool, and almost as good as that with Fibreglass, but both materials are slightly hazardous to work with - may irritate skin and will irritate eyes and can damage the surface of eyes,
{wear gloves, and goggles if you work close to it in the cabinet}.
Long-haired wool is safe, and better than plastic foam, but wool if in batts and the mineral and glass fibre products need to be optimally aligned in the enclosure to absorb optimally, and to not reduce efficiency of the woofer by mass loading if they are free to flop about inside whilst the woofer is pumping !
CAPACITORS for each 66 version
There are some interim versions of the 66 which use some of the capacitor values of the original T.B.C. model and some of the capacitor values from the later P.C.C. model.
I do not know why, but perhaps Celestion were changing their Inventory, and, I think it likely they were considering the change from 30uF in Series to the older MF500 mid-dome to 24uF before they had the later MD-500 mid-dome in production.
If one wants to hear what their particular model of 66 sounds like with new caps, then use the same Total Capacitance in each location as was in the original model with the particular drivers.
Also, there will be some differences in frequency response between the older and newer versions of the drivers, thus where a change of capacitance was made it was likely to have been to adjust the cross-over for optimum response with the new drivers.
****************************************************
TWEETER FILTER
For the original T.1637 version of the HF2000 tweeter there was a Total of 6uF in the parallel connected bunch of caps which is located at the output of the filter directly to the tweeter.
For replacement here use a single 6uF cap,
or a parallel pair of 3uF caps,
or parallel three of 2uF caps,
or parallel four of 1.5uF caps,
AND, use all of the same Brand , type and Voltage rating.
Any of those options will produce a more coherent sound with modern Polypropylene capacitors than would occur if one paralleled different value caps which sum to 6uF, eg: 2uF//4uF - which will mess up the transient response of the signal.
// is the symbol used in Electronics for a Parallel connection.
+ is the smbol used for a Series connection.
If you have 6.2uF cap and the T.1637 tweeter, connect a 100 ohm, 5 watt resistor in Parallel across the connections to the tweeter.
This will reduce the Load Impedance of the tweeter sufficiently to adjust it for a 6.2uF cap, and the rest of the filter.
There will be no loss of output from the tweeter.
The resistor does not need to be non-inductive there, but do use a good quality resistor, and mount it so that its body is not touching any metal or any of the caps.
For the later T.2373 version of the HF2000 there is a Total of 6.2uF for that output cap.
For replacement use a single 6.2uF cap,
or 3uF // 3.3uF ONLY if both caps are of the same length,
or 2uF // 2uF // 2.2uF ONLY if all caps are of the same length.
For the same reasons I wrote above about the 6uF cap, do not use a parallel connected 1.5uF // 4.7uF in this location with modern Polypropylene caps.
If you already have only 6uF for the later T.2373 tweeter, don't worry too much, as there is only small difference, thus put it in and listen.
Connecting a 0.2uF or 0.22uF cap in parallel with it will not produce more accurate sound because it will mess up the transient response.
Both versions of the tweeter filter used 4uF for the input capacitor, thus simply put in a new 4uF cap,
or a 3.9uF - close enough,
or a pair of 2uF // 2uF.
This same situation occurs for the capacitor connected in parallel at the output to the mid-dome in both versions of the 66,
thus simply use same there as you buy for the tweeter 4uF cap.
I will post about the Series connected capacitor at the input to the Midrange Filter in my next Post, as I have run out of ime now.
There are some interim versions of the 66 which use some of the capacitor values of the original T.B.C. model and some of the capacitor values from the later P.C.C. model.
I do not know why, but perhaps Celestion were changing their Inventory, and, I think it likely they were considering the change from 30uF in Series to the older MF500 mid-dome to 24uF before they had the later MD-500 mid-dome in production.
If one wants to hear what their particular model of 66 sounds like with new caps, then use the same Total Capacitance in each location as was in the original model with the particular drivers.
Also, there will be some differences in frequency response between the older and newer versions of the drivers, thus where a change of capacitance was made it was likely to have been to adjust the cross-over for optimum response with the new drivers.
****************************************************
TWEETER FILTER
For the original T.1637 version of the HF2000 tweeter there was a Total of 6uF in the parallel connected bunch of caps which is located at the output of the filter directly to the tweeter.
For replacement here use a single 6uF cap,
or a parallel pair of 3uF caps,
or parallel three of 2uF caps,
or parallel four of 1.5uF caps,
AND, use all of the same Brand , type and Voltage rating.
Any of those options will produce a more coherent sound with modern Polypropylene capacitors than would occur if one paralleled different value caps which sum to 6uF, eg: 2uF//4uF - which will mess up the transient response of the signal.
// is the symbol used in Electronics for a Parallel connection.
+ is the smbol used for a Series connection.
If you have 6.2uF cap and the T.1637 tweeter, connect a 100 ohm, 5 watt resistor in Parallel across the connections to the tweeter.
This will reduce the Load Impedance of the tweeter sufficiently to adjust it for a 6.2uF cap, and the rest of the filter.
There will be no loss of output from the tweeter.
The resistor does not need to be non-inductive there, but do use a good quality resistor, and mount it so that its body is not touching any metal or any of the caps.
For the later T.2373 version of the HF2000 there is a Total of 6.2uF for that output cap.
For replacement use a single 6.2uF cap,
or 3uF // 3.3uF ONLY if both caps are of the same length,
or 2uF // 2uF // 2.2uF ONLY if all caps are of the same length.
For the same reasons I wrote above about the 6uF cap, do not use a parallel connected 1.5uF // 4.7uF in this location with modern Polypropylene caps.
If you already have only 6uF for the later T.2373 tweeter, don't worry too much, as there is only small difference, thus put it in and listen.
Connecting a 0.2uF or 0.22uF cap in parallel with it will not produce more accurate sound because it will mess up the transient response.
Both versions of the tweeter filter used 4uF for the input capacitor, thus simply put in a new 4uF cap,
or a 3.9uF - close enough,
or a pair of 2uF // 2uF.
This same situation occurs for the capacitor connected in parallel at the output to the mid-dome in both versions of the 66,
thus simply use same there as you buy for the tweeter 4uF cap.
I will post about the Series connected capacitor at the input to the Midrange Filter in my next Post, as I have run out of ime now.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.