So sorry, my mistake - as soon as I saw B&W I just assumed it was a modern B&W speaker and not being familiar with the DM2 I didn't realise it was a classic from way back. So I take back what I said about the B&Ws being less warm and more analytical - it could easily be the other way around! But if you already have a classic speaker of this calibre, I wonder if there's really all that much point in doing up a pair of 44s? I can certainly see the point of doing it out of curiosity or just for fun but I'm not sure it would really be a step up from what you already have. Maybe others will disagree - as I say, I've never owned any DM2s so I'm not qualified to say.
I hope others chip in because there's not much more I can add!
I hope others chip in because there's not much more I can add!
Well the Tweeters are on order (had to delay as Falcon were all on their holidays 
now thinking of capacitors & resistors. Anyone hazard a guess what difference using 68uf instead of 72 in the bass circuit would have? Realise of course given tolerances of even 5% the answer could be absolutely none!
now thinking of capacitors & resistors. Anyone hazard a guess what difference using 68uf instead of 72 in the bass circuit would have? Realise of course given tolerances of even 5% the answer could be absolutely none!Thanks, I'd not actually thought of Wilmslow, the only 72uf's I'd found was Alcap from Falcon & eBay - I'd used them once before and to say I was unimpressed would be an understatement (!) I'd been going to use Nichicon GB/DB or possibly Jantzen electros.
I did think of trimming them up but thought that was now frowned upon
Thanks guys, appreciate the thinking material, it's definitely better if you're not making stuff up on your own
if a low esr cap is paralleled with another cap and that pair are subjected to a step signal then there can be ringing, which is not present when one cap alone is subjected to the same stimulus.
But in a speaker crossover where there should be no step change signals, is this a concern?
Many very expensive and well reviewed speakers use multi-parallel low esr capacitors in their crossovers. Those manufacturers seem to consider that getting the right capacitance is more important than avoiding paralleling. I think thus validates the technique.
But in a speaker crossover where there should be no step change signals, is this a concern?
Many very expensive and well reviewed speakers use multi-parallel low esr capacitors in their crossovers. Those manufacturers seem to consider that getting the right capacitance is more important than avoiding paralleling. I think thus validates the technique.
two peculiarities , and other recommendations
'ullo vinyl dragon ,
I am a bit puzzled as you stated you have read both this thread and the Celestion 66 midrange thread ,
but in your first post here you have:
With SEAS 19TFF , 10 ohm and 1.5 ohm will cause a lot of treble loss , AND quite an amount of reduction of the dynamics possible from this tweeter ,
all resulting in a quite dull or slow sound.
Perhaps that will match the characteristics of the particular midrange cone in your 44s ,
or perhaps you will like it in other ways , but damping down the 19TFF that much is a risk.
I recommend starting with 15 ohm in Parallel and 1 ohm in Series for the L-Pad.
That should give a similar level of treble to the original , but with the audible benefits of the better new tweeter.
As you want minimal changes and simple readily available components you will be keeping the 0.14mH inductor ,
thus 4uF or the equivalent 3.9uF will be close enough to optimum for the Input capacitor , but 6uF for the Output capacitor will cause a peculiar sound.
That 6uF cap was used in the original because the Celestion HF2000 has a hump in the low treble of its useable frequency response which needed to be flattened.
The SEAS has no hump and is already a good flat response there. 6uF there will cause a lot of treble loss at and a little above the crossover point.
If you really want to put 6uF there the use no resistors.
Result will be almost correct treble near the crossover point , but elevated treble response at higher frequencies - the opposite of the original HF2000.
Better for the SEAS is to use 10uF or 12uF there { of the readily available values }.
Which one ?
Start with 10uF . Another 44 owner in this thread has that with the 4uF and reported good result.
If you find there is not sufficient in the low-treble region then swap in 12uF.
But if you find there is not sufficient over all the treble range then leave in the 10uF and take out the 15 ohm and 1 ohm and put in 27 ohm in Parallel and no Series resistor.
You don't like the sound from Alcaps , thus try the other brands you mentioned if you will only use bi-polar electro caps in the woofer circuit ,
but neither will sound like the original Elcaps did when new , and most modern electros will not , because most are much lower ESR than the old Elcaps.
Electros of different capacitances can be connected in Parallel with less audible peculiarity result { if both are same model } than an electro in Parallel with a plastic film cap , because the electros relative ESRs tend to compensate each for the other.
I would use Elna RBD bi-polar 22uF//47uF = 69uF , which as electros Tolerance are more often on the + side more so than the - side should by chance give you 70uF or greater , and that will be close enough.
Elna RBD will have longer Service Life than any other commonly available electro cap. They are fairly low price , and small size to fit.
As you are intending to use the physically large 400 volt Solens from Falcon ,
much better is to buy from Wilmslow Audio their Supersound caps which cost very little more but do sound better , and the SEAS tweeter deserves such.
That SEAS tweeter will reveal the inadequacies of Solen caps.
Supersound include: 4uf , 5.6uF , 10uF , 30uF.
Use the 30uF Supersound as Input to the midrange circuit for consistency of sound with the treble ,
and their 5.6uF will suffice instead of 6uF in the mids' output Parallel position as that cap's value is not absolutely critical ,
however if you ask Wilmslow they might measure some samples and select two that measure on the + side of the Tolerance thus close to 6uF.
Alternately have them measure some samples of 6.8uF and select two that are in the - side of the Tolerance , thus little more than 6uF.
If your mid-cones are still in good audible condition they are worth at least the Supersound caps.
Hi-Fi Collective:
Amplifier valve kits, HIFI pre-amplifiers, speaker kits, AMP Parts, upgrade components - Hifi Collective
sell the Elna RBD bi-polars , and also the Nichicon Muse bi-polars which would be my second choice for electros ,
but I would use polypropylene film caps , thus if you change your mind post here and I'll advise a simple choice.
Hi-Fi Collective also sell better audio quality resistors than Falcon and Wilmslow , and some for little higher price.
I recommend either Jantzen Superes in 5 watt or Mundorf M-resist 5 watt MR5 MOX if you don't want to pay high prices.
The Superes may be the better option over the long term of life , but I do not know for sure.
B&W DM2 was available in at least two different versions.
Neither sounded anything like a Celestion 44 , AND both versions of the DM2 drove low frequency room acoustics quite differently to 44s ,
thus your hearing even with the aged loudspeaker samples is correct.
Post here if there is anything above that I have not explained well enough.
--- --- --- --- --- --- --- --- --- --- ---
'ullo AndrewT ,
Some high-end manufacturers do Parallel polypropylene film caps ,
however those poly caps are usually chosen by listening , AND in relation to the audible resonances of the drivers , but don't expect those manufacturers to tell us their secrets of how they decide such !
Some high-end manufacturers use only a single cap , eg: B&W used a single expensive Mundorf cap in their not long ago 800 series models , and they may still or may have changed to one of the expensive ClarityCap types.
Some manufacturers do not listen to such degree and make good sounding speakers by co-incidence , but often as not those can be improved by careful choice capacitor arrangement changes.
Different listeners have often have quite different ear-brain audio priorities ,
thus why so many differences of opinion about which hi-end brand is best , etc ... for seemingly very similar technically capable loudspeakers.
I think you will have experience of some of the above , and all your comments are welcome by me !
'ullo vinyl dragon ,
I am a bit puzzled as you stated you have read both this thread and the Celestion 66 midrange thread ,
but in your first post here you have:
With SEAS 19TFF , 10 ohm and 1.5 ohm will cause a lot of treble loss , AND quite an amount of reduction of the dynamics possible from this tweeter ,
all resulting in a quite dull or slow sound.
Perhaps that will match the characteristics of the particular midrange cone in your 44s ,
or perhaps you will like it in other ways , but damping down the 19TFF that much is a risk.
I recommend starting with 15 ohm in Parallel and 1 ohm in Series for the L-Pad.
That should give a similar level of treble to the original , but with the audible benefits of the better new tweeter.
As you want minimal changes and simple readily available components you will be keeping the 0.14mH inductor ,
thus 4uF or the equivalent 3.9uF will be close enough to optimum for the Input capacitor , but 6uF for the Output capacitor will cause a peculiar sound.
That 6uF cap was used in the original because the Celestion HF2000 has a hump in the low treble of its useable frequency response which needed to be flattened.
The SEAS has no hump and is already a good flat response there. 6uF there will cause a lot of treble loss at and a little above the crossover point.
If you really want to put 6uF there the use no resistors.
Result will be almost correct treble near the crossover point , but elevated treble response at higher frequencies - the opposite of the original HF2000.
Better for the SEAS is to use 10uF or 12uF there { of the readily available values }.
Which one ?
Start with 10uF . Another 44 owner in this thread has that with the 4uF and reported good result.
If you find there is not sufficient in the low-treble region then swap in 12uF.
But if you find there is not sufficient over all the treble range then leave in the 10uF and take out the 15 ohm and 1 ohm and put in 27 ohm in Parallel and no Series resistor.
You don't like the sound from Alcaps , thus try the other brands you mentioned if you will only use bi-polar electro caps in the woofer circuit ,
but neither will sound like the original Elcaps did when new , and most modern electros will not , because most are much lower ESR than the old Elcaps.
Electros of different capacitances can be connected in Parallel with less audible peculiarity result { if both are same model } than an electro in Parallel with a plastic film cap , because the electros relative ESRs tend to compensate each for the other.
I would use Elna RBD bi-polar 22uF//47uF = 69uF , which as electros Tolerance are more often on the + side more so than the - side should by chance give you 70uF or greater , and that will be close enough.
Elna RBD will have longer Service Life than any other commonly available electro cap. They are fairly low price , and small size to fit.
As you are intending to use the physically large 400 volt Solens from Falcon ,
much better is to buy from Wilmslow Audio their Supersound caps which cost very little more but do sound better , and the SEAS tweeter deserves such.
That SEAS tweeter will reveal the inadequacies of Solen caps.
Supersound include: 4uf , 5.6uF , 10uF , 30uF.
Use the 30uF Supersound as Input to the midrange circuit for consistency of sound with the treble ,
and their 5.6uF will suffice instead of 6uF in the mids' output Parallel position as that cap's value is not absolutely critical ,
however if you ask Wilmslow they might measure some samples and select two that measure on the + side of the Tolerance thus close to 6uF.
Alternately have them measure some samples of 6.8uF and select two that are in the - side of the Tolerance , thus little more than 6uF.
If your mid-cones are still in good audible condition they are worth at least the Supersound caps.
Hi-Fi Collective:
Amplifier valve kits, HIFI pre-amplifiers, speaker kits, AMP Parts, upgrade components - Hifi Collective
sell the Elna RBD bi-polars , and also the Nichicon Muse bi-polars which would be my second choice for electros ,
but I would use polypropylene film caps , thus if you change your mind post here and I'll advise a simple choice.
Hi-Fi Collective also sell better audio quality resistors than Falcon and Wilmslow , and some for little higher price.
I recommend either Jantzen Superes in 5 watt or Mundorf M-resist 5 watt MR5 MOX if you don't want to pay high prices.
The Superes may be the better option over the long term of life , but I do not know for sure.
B&W DM2 was available in at least two different versions.
Neither sounded anything like a Celestion 44 , AND both versions of the DM2 drove low frequency room acoustics quite differently to 44s ,
thus your hearing even with the aged loudspeaker samples is correct.
Post here if there is anything above that I have not explained well enough.
--- --- --- --- --- --- --- --- --- --- ---
'ullo AndrewT ,
Some high-end manufacturers do Parallel polypropylene film caps ,
however those poly caps are usually chosen by listening , AND in relation to the audible resonances of the drivers , but don't expect those manufacturers to tell us their secrets of how they decide such !
Some high-end manufacturers use only a single cap , eg: B&W used a single expensive Mundorf cap in their not long ago 800 series models , and they may still or may have changed to one of the expensive ClarityCap types.
Some manufacturers do not listen to such degree and make good sounding speakers by co-incidence , but often as not those can be improved by careful choice capacitor arrangement changes.
Different listeners have often have quite different ear-brain audio priorities ,
thus why so many differences of opinion about which hi-end brand is best , etc ... for seemingly very similar technically capable loudspeakers.
I think you will have experience of some of the above , and all your comments are welcome by me !
Last edited:
Hi Alan
To be honest I got myself so confused reading both the threads & lots of advice, much of it conflicting (you should see my notes!) in the end I didn't quite know what to think.
Basically after a lot of thought what I really want to achieve is what Kat Manton on Vintage Radio was trying to do, simply return them to full working order (I took my resistor values from her). As I said I have one dead and one dodgy tweeter so leaving them as they are is not an option but even in that state, they're absolute *magic*. Up till now I've always used speakers by Spendor, Quad & B&W, these Dittons are a revelation, and the more I listen to them the more I love them.
I did think of trying to source a pair of HF2000's but that just not an option for all sorts of reasons..
Anyway I've bought a pair of Seas 19TFF's and would just like them to integrate with the rest of the speaker. I did read on AOS (jk Wynn's Ditton 66 thread) that this can be done with a replacement L5 of 0.17MH and 3.3 & 9.0uf caps. Though I always understood the Seas has higher efficency then the HF2000 so surely would need at least one resistor to match levels? Basically it's the sound I want to keep original..
As for the rest of the crossover I shall simply measure the caps (I have access to an accurate capacitance/ESR meter now) and replace any that are out of spec. I realise at least in the bass it would have to be Alcaps as I have some Jantzen electrolytics and they have an ESR of only R0.11
I really hadn't realised about the controlled HF2000 'hump' I must somehow have missed that bit..
I realise what I'm trying to do is very in-glamorous but I'd rather just get them working properly & spend my cash on new music
So what would be the *best* option to simply integrate the Seas do you think?
Thanks for the advice
All the best
Graham
To be honest I got myself so confused reading both the threads & lots of advice, much of it conflicting (you should see my notes!) in the end I didn't quite know what to think.
Basically after a lot of thought what I really want to achieve is what Kat Manton on Vintage Radio was trying to do, simply return them to full working order (I took my resistor values from her). As I said I have one dead and one dodgy tweeter so leaving them as they are is not an option but even in that state, they're absolute *magic*. Up till now I've always used speakers by Spendor, Quad & B&W, these Dittons are a revelation, and the more I listen to them the more I love them.
I did think of trying to source a pair of HF2000's but that just not an option for all sorts of reasons..
Anyway I've bought a pair of Seas 19TFF's and would just like them to integrate with the rest of the speaker. I did read on AOS (jk Wynn's Ditton 66 thread) that this can be done with a replacement L5 of 0.17MH and 3.3 & 9.0uf caps. Though I always understood the Seas has higher efficency then the HF2000 so surely would need at least one resistor to match levels? Basically it's the sound I want to keep original..
As for the rest of the crossover I shall simply measure the caps (I have access to an accurate capacitance/ESR meter now) and replace any that are out of spec. I realise at least in the bass it would have to be Alcaps as I have some Jantzen electrolytics and they have an ESR of only R0.11
I really hadn't realised about the controlled HF2000 'hump' I must somehow have missed that bit..
I realise what I'm trying to do is very in-glamorous but I'd rather just get them working properly & spend my cash on new music
So what would be the *best* option to simply integrate the Seas do you think?
Thanks for the advice
All the best
Graham
A Correction and Addition to #308 , and a brief reply to #309
*
To whoever may be following my Post #308 recommendations:
I left the final letter off a word in that Post ,
thus I will post the correction in the Quote Box below ,
and I will add another option for the Elna RBD Bi-polars which I should have thought of then ,
but which did not come into my mind till later.
'ullo vinyl dragon ,
I have another matter I have to attend to now , however I will be back as soon as I have time available ,
{ which I hope will be later today , but it might not be till 3 days time } ,
as I have another electrolytic caps' option that might get you closer to the sound of the originals ,
and I have two options for the 6uF cap for the midrange filter.
*
To whoever may be following my Post #308 recommendations:
I left the final letter off a word in that Post ,
thus I will post the correction in the Quote Box below ,
and I will add another option for the Elna RBD Bi-polars which I should have thought of then ,
but which did not come into my mind till later.
*
'ullo vinyl dragon ,
I have another matter I have to attend to now , however I will be back as soon as I have time available ,
{ which I hope will be later today , but it might not be till 3 days time } ,
as I have another electrolytic caps' option that might get you closer to the sound of the originals ,
and I have two options for the 6uF cap for the midrange filter.
Last edited:
Continuing Reply to #309
'ullo Graham ,
and to whoever else may be following this subject.
I apologize it has taken me longer than I had anticipated till I got time to come back to this.
I'll type a few Replies inside the Quote box to specific points there ,
and follow with more after the Quote box.
Within , following each --->
Measuring Capacitance of Electrolytic capacitors is not simple , not even with new electros , regardless of whether Polar or Bipolar.
I have stated various of the reasons in at least one of the Celestion threads.
Here I will re-state that an electro that is leaking charge , which all those old Elcaps will be ,
will not measure its actual Capacitance on various types of Capacitance meters
when the type of Meter is one that measures the result of a Charge/Discharge cycle ,
because a leaky Capacitor { regardless of Dielectric type } does not have a definite Capacitance in so far as storing charge because it is always leaking it ,
nor does it have a definite Impedance with Frequency { Capacitive Reactance } , because it is leaking charge that is applied at any Frequency.
The Meter is continually trying to Charge Up the cap to measure its Time Constant in the meter's circuit , from which the Capacitance is calculated ,
but as the cap leaks it is never fully charged thus the meter is continually charging it , though when the leakage is below some critical amount the meter will eventually display a Result ,
however that result is usually one indicating significantly higher capacitance than is actually the case ,
and as I stated there is no actual fixed capacitance for a leaking capacitor.
Before you decide whether to use the old Elcaps , test them for Leakage , as follows:
Use a new or close to fully charged 9 volt battery , or two of 9 volt connected in electrical Series { with wire and alligator clips , do not solder onto battery terminal },
or use a DC output Power Supply at any voltage to NOT higher than HALF the Voltage printed on the body of the Elcap ,
because those are AC voltage rated Caps , not DC rated like Polar electros ,
and being old samples their voltage stress tolerances will have deteriorated.
The caps will charge quickly , but leave them for longer than a few minutes till whatever of the initial Leakage stabilizes ,
then BEFORE you disconnect the Battery or Power Supply ,
measure the DC Voltage and write it some where.
Next , quickly disconnect one terminal from the Power Supply and take care to not short circuit that lead of the cap onto any metal.
Then disconnect the other end lead , taking care to not short circuit anything because a short circuit will almost instantly discharge a lot of the voltage.
Pick up the cap by holding its insulated body - do not touch either lead because the cap will discharge , albeit slowly , through your body if both its leads can complete a circuit anywhere.
Put each charged cap somewhere it will not move to the other caps nor to anything it can complete a circuit through , nor let it get wet/damp.
Charge each Elcap by itself , because the more leaky samples will prevent the less leaky sample from charging when all are connected together in Parallel.
Label each cap and write its charged voltage.
All should show the same charged voltage , unless the battery is running flat { almost exhausted of its charge reserve }.
Any Elcaps that will not charge as high as the others to the same Voltage will be too leaky to use.
For the samples that all do charge to the same voltage , leave them for at least an hour , but not too much longer because all will discharge somewhat over time ... example try leaving them overnight or longer and measure and many will have less than half the initial voltage - for electrolytic caps.
{ Plastic Film caps should hold close to full voltage for days at least and the best one hold charge for weeks or longer , depending on atmospheric humidity.}
Any Elcap that is close to half or lower of its initial charged voltage after about an hour is fairly useless in a crossover , and also is not safe to use in Series with a midrange driver nor tweeter.
Thus after about an hour or whatever period longer connect one lead to a DC Voltmeter or DC Voltage terminal of a Multimeter ,
then quickly connect the other lead and watch for the displayed result.
The meter may flash a few digits before it stabilizes , but remember the first almost stable reading it displays then quickly disconnect one lead ,
because the cap is actually discharging through the circuit of the meter.
If you leave one connected you will see the voltage reading reducing.
Write down all the then measure Voltages , then compare each to the initial charged voltage for each cap.
If the voltage has reduced by more than about 10% I would not use the cap ,
but you can compromise a bit more if you want to to about a 20% voltage reduction for the particular caps that were connected in Parallel with the speaker drivers in the crossover circuit ,
those being the 72uF in the woofer circuit and the 6uF output cap in the midrange circuit.
The 24uf and its Parallel 6uF , or single 28uF or 30uF in the midrange circuit ,
and both caps in the Tweeter circuit are all Series connected components ,
thus significantly leaky samples should not be used , because they will not sufficiently filter low frequencies from reaching the drivers.
Post the Initial Charge Voltage , and the 1 hour or whatever time later measured voltage for each sample here , and I will comment about each if you wish.
--- --- ---
The Jantzen bipolar electros are not manufactured by Jantzen.
They are bought from perhaps a major European caps manufacturer , or more likely from an Asian manufacturer.
Look at Mundorf's , ECap AC Bipolar , and you will see two types:
"raw" and "plain".
"Raw" are etched foil capacitors - that is a later technology developed the enable smaller physical sized components.
It works by exposing larger surface are of the foil as result of multiple scratches cut into the foil.
The trade-off is that this type usually leaks charge to greater degree than non-etched types { even when new after initial Forming }.
"Plain" are smooth foil caps , and that was the original type of rolled electrolytic cap.
That is what the LL marked Elcaps are , and perhaps the other Elcaps are also plain foil -{ I have forgotten when etched foil caps were introduced }.
Plain foil electros usually leak to significantly lesser degree than etched foil types , that is when both types are subject to the same Quality Control in manufacture.
Only part of the Sound of an electrolytic cap is related to its Dissipation Factor/ESR.
Another significant contributor to the Sound is the Dielectric Absorption.
Various things affect dielectric absorption , however Plain foil and Etched foil electros have somewhat different dielectric absorption amounts even when the same type of Dielectric is used in both.
Of the various things that cause different electro caps to sound the way they do foil type is one factor ,
thus if you use Plain foil electros you will get closer to the original sound at least for the LL type Elcaps , and perhaps for all the Elcaps ,
BUT both the Mundorf ECap AC Bipolar types have much lower ESR than the old Elcaps , thus you will likely still need the compensating resistors connected in Series with whatever capacitances you sum to within the 68uf , 72uf , 75uf range for the woofer circuit , and for the 6uF output cap in the midrange circuit.
You may not need a resistor to the 24uF or 30uF Input cap of the midrange circuit if you do not attenuate the SEAS tweeter with L-pad nor Series resistor ,
though the mids will be a little louder than previously and noticeable as thus more prominent over the Bass output ,
thus decide whether to keep the relative Bass versus Midrange outputs as they are or allow change by not using a Series resistor at the mids' Input cap.
Mundorf have their ECap AC Bipolars manufactured for them by a long in business major German manufacturer of many types of capacitors.
They will likely be very well manufactured , and should have long Service Life albeit within their specification for use.
I've described enough for you to think about for now , so I'll post this and await you reply before proceeding any further.
'ullo Graham ,
and to whoever else may be following this subject.
I apologize it has taken me longer than I had anticipated till I got time to come back to this.
I'll type a few Replies inside the Quote box to specific points there ,
and follow with more after the Quote box.
Within , following each --->
---To be honest I got myself so confused reading both the threads & lots of advice, much of it conflicting (you should see my notes!) in the end I didn't quite know what to think.
---> That happens to a lot of people who are fairly new to DIY where several opinions are offered.
Some of it derives from learning general technical matter ,
and other derives from how differently different people hear , by which I mean the specific Audio related priorities one's brain allocates which differ from how another listener's brain allocates.
We all differ to some degree , however there are distinct groupings of people who have similar brain audio priority allocations , and those groups include as many listeners who have different tastes in Music as have similar - that is: tastes in Music with regard to Styles of Music or performance attributes within a Style seem to have no correlation with brain priority allocation for some distinct Audio attributes ... and thus the differences of opinion , and some arguments , in posts in this Forum.
---
Basically after a lot of thought what I really want to achieve is what Kat Manton on Vintage Radio was trying to do, simply return them to full working order (I took my resistor values from her).
---> I had also been following Kat Manton's experiments.
Note however that she had some Russian Military Teflon dielectric capacitors.
The sound through those is nothing like the sound through any electrolytic cap ,
however she may have only use those Teflon caps to the SEAS tweeter.
I have lost the Link I had to her thread in whatever Forum it was ... do you have it ?
---
As I said I have one dead and one dodgy tweeter so leaving them as they are is not an option but even in that state, they're absolute *magic*. Up till now I've always used speakers by Spendor, Quad & B&W, these Dittons are a revelation, and the more I listen to them the more I love them.
I did think of trying to source a pair of HF2000's but that just not an option for all sorts of reasons..
---> ridiculously high prices are asked for them , given the limited Service Life possibility of any sample with its unknown use history to the buyer.
Price may be worth it only if a listener must have that exact sound ,
but better would be to leave them to be bought by Owners of the vintage Rogers loudspeakers which used HF2000 for those Owners who want to continue with Rogers' particular crossovers.
---
Anyway I've bought a pair of Seas 19TFF's and would just like them to integrate with the rest of the speaker. I did read on AOS (jk Wynn's Ditton 66 thread) that this can be done with a replacement L5 of 0.17MH and 3.3 & 9.0uf caps. Though I always understood the Seas has higher efficency then the HF2000 so surely would need at least one resistor to match levels?
---> Not 3.3uF but 3.0uF , with 9uF or 10uF and 0.17mH.
---
Basically it's the sound I want to keep original..
---> You will not be able to keep the original sound , however you can try to get close to it.
---
As for the rest of the crossover I shall simply measure the caps (I have access to an accurate capacitance/ESR meter now) and replace any that are out of spec.
---> If you have already done that , then Post the measured results here ,
however they may not be accurate ... I will describe why later in this post.
---
I realise at least in the bass it would have to be Alcaps as I have some Jantzen electrolytics and they have an ESR of only R0.11
---> I do not know of any current manufacture Bipolar electros which have Dissipation Factor of close to 0.2 at 1kHz and the higher Dissipation Factors at higher frequencies similar to the old Elcaps , and I doubt that Alcaps do.
There will be some low-cost manufacture bipolars somewhere , but they will be compromised in other ways - they are unlikely to be optimised for passive crossover applications.
The old Elcaps were a coincidence of a Cap design { two designs at least - the ordinary version and the LL version } which sufficient Crossover Designers happened to find worked well enough for their Loudspeakers at a cost effective-price in that era.
Those two designs are no longer manufactured , because there is no large scale Commercial Industrial nor Military application for them in modern technology.
Alcaps were designed for passive crossovers , albeit with specification decided by a few of the then large British loudspeaker manufacturers ,
and as such not all designers , nor all listeners , agree.
---
I really hadn't realised about the controlled HF2000 'hump' I must somehow have missed that bit..
I realise what I'm trying to do is very in-glamorous but I'd rather just get them working properly & spend my cash on new music
---> So would I , and seemingly most music lovers , but it is not always simple to do , thus why this DIY Audio Forum and others similar !
---
So what would be the *best* option to simply integrate the Seas do you think?
---> Hah , "best" indeed ! ... the answer will have to be: What You Think ,
because I have both in this thread and in the 66 midrange thread already stated what I think !!
OK , I'll be a bit merciful and will TRY to recommend following what you have stated about what you are trying to achieve ,
but please remember that we are not together in a room actually listening through your sound system.
Measuring Capacitance of Electrolytic capacitors is not simple , not even with new electros , regardless of whether Polar or Bipolar.
I have stated various of the reasons in at least one of the Celestion threads.
Here I will re-state that an electro that is leaking charge , which all those old Elcaps will be ,
will not measure its actual Capacitance on various types of Capacitance meters
when the type of Meter is one that measures the result of a Charge/Discharge cycle ,
because a leaky Capacitor { regardless of Dielectric type } does not have a definite Capacitance in so far as storing charge because it is always leaking it ,
nor does it have a definite Impedance with Frequency { Capacitive Reactance } , because it is leaking charge that is applied at any Frequency.
The Meter is continually trying to Charge Up the cap to measure its Time Constant in the meter's circuit , from which the Capacitance is calculated ,
but as the cap leaks it is never fully charged thus the meter is continually charging it , though when the leakage is below some critical amount the meter will eventually display a Result ,
however that result is usually one indicating significantly higher capacitance than is actually the case ,
and as I stated there is no actual fixed capacitance for a leaking capacitor.
Before you decide whether to use the old Elcaps , test them for Leakage , as follows:
Use a new or close to fully charged 9 volt battery , or two of 9 volt connected in electrical Series { with wire and alligator clips , do not solder onto battery terminal },
or use a DC output Power Supply at any voltage to NOT higher than HALF the Voltage printed on the body of the Elcap ,
because those are AC voltage rated Caps , not DC rated like Polar electros ,
and being old samples their voltage stress tolerances will have deteriorated.
The caps will charge quickly , but leave them for longer than a few minutes till whatever of the initial Leakage stabilizes ,
then BEFORE you disconnect the Battery or Power Supply ,
measure the DC Voltage and write it some where.
Next , quickly disconnect one terminal from the Power Supply and take care to not short circuit that lead of the cap onto any metal.
Then disconnect the other end lead , taking care to not short circuit anything because a short circuit will almost instantly discharge a lot of the voltage.
Pick up the cap by holding its insulated body - do not touch either lead because the cap will discharge , albeit slowly , through your body if both its leads can complete a circuit anywhere.
Put each charged cap somewhere it will not move to the other caps nor to anything it can complete a circuit through , nor let it get wet/damp.
Charge each Elcap by itself , because the more leaky samples will prevent the less leaky sample from charging when all are connected together in Parallel.
Label each cap and write its charged voltage.
All should show the same charged voltage , unless the battery is running flat { almost exhausted of its charge reserve }.
Any Elcaps that will not charge as high as the others to the same Voltage will be too leaky to use.
For the samples that all do charge to the same voltage , leave them for at least an hour , but not too much longer because all will discharge somewhat over time ... example try leaving them overnight or longer and measure and many will have less than half the initial voltage - for electrolytic caps.
{ Plastic Film caps should hold close to full voltage for days at least and the best one hold charge for weeks or longer , depending on atmospheric humidity.}
Any Elcap that is close to half or lower of its initial charged voltage after about an hour is fairly useless in a crossover , and also is not safe to use in Series with a midrange driver nor tweeter.
Thus after about an hour or whatever period longer connect one lead to a DC Voltmeter or DC Voltage terminal of a Multimeter ,
then quickly connect the other lead and watch for the displayed result.
The meter may flash a few digits before it stabilizes , but remember the first almost stable reading it displays then quickly disconnect one lead ,
because the cap is actually discharging through the circuit of the meter.
If you leave one connected you will see the voltage reading reducing.
Write down all the then measure Voltages , then compare each to the initial charged voltage for each cap.
If the voltage has reduced by more than about 10% I would not use the cap ,
but you can compromise a bit more if you want to to about a 20% voltage reduction for the particular caps that were connected in Parallel with the speaker drivers in the crossover circuit ,
those being the 72uF in the woofer circuit and the 6uF output cap in the midrange circuit.
The 24uf and its Parallel 6uF , or single 28uF or 30uF in the midrange circuit ,
and both caps in the Tweeter circuit are all Series connected components ,
thus significantly leaky samples should not be used , because they will not sufficiently filter low frequencies from reaching the drivers.
Post the Initial Charge Voltage , and the 1 hour or whatever time later measured voltage for each sample here , and I will comment about each if you wish.
--- --- ---
The Jantzen bipolar electros are not manufactured by Jantzen.
They are bought from perhaps a major European caps manufacturer , or more likely from an Asian manufacturer.
Look at Mundorf's , ECap AC Bipolar , and you will see two types:
"raw" and "plain".
"Raw" are etched foil capacitors - that is a later technology developed the enable smaller physical sized components.
It works by exposing larger surface are of the foil as result of multiple scratches cut into the foil.
The trade-off is that this type usually leaks charge to greater degree than non-etched types { even when new after initial Forming }.
"Plain" are smooth foil caps , and that was the original type of rolled electrolytic cap.
That is what the LL marked Elcaps are , and perhaps the other Elcaps are also plain foil -{ I have forgotten when etched foil caps were introduced }.
Plain foil electros usually leak to significantly lesser degree than etched foil types , that is when both types are subject to the same Quality Control in manufacture.
Only part of the Sound of an electrolytic cap is related to its Dissipation Factor/ESR.
Another significant contributor to the Sound is the Dielectric Absorption.
Various things affect dielectric absorption , however Plain foil and Etched foil electros have somewhat different dielectric absorption amounts even when the same type of Dielectric is used in both.
Of the various things that cause different electro caps to sound the way they do foil type is one factor ,
thus if you use Plain foil electros you will get closer to the original sound at least for the LL type Elcaps , and perhaps for all the Elcaps ,
BUT both the Mundorf ECap AC Bipolar types have much lower ESR than the old Elcaps , thus you will likely still need the compensating resistors connected in Series with whatever capacitances you sum to within the 68uf , 72uf , 75uf range for the woofer circuit , and for the 6uF output cap in the midrange circuit.
You may not need a resistor to the 24uF or 30uF Input cap of the midrange circuit if you do not attenuate the SEAS tweeter with L-pad nor Series resistor ,
though the mids will be a little louder than previously and noticeable as thus more prominent over the Bass output ,
thus decide whether to keep the relative Bass versus Midrange outputs as they are or allow change by not using a Series resistor at the mids' Input cap.
Mundorf have their ECap AC Bipolars manufactured for them by a long in business major German manufacturer of many types of capacitors.
They will likely be very well manufactured , and should have long Service Life albeit within their specification for use.
I've described enough for you to think about for now , so I'll post this and await you reply before proceeding any further.
Piezo Tweeter Mishap
Good day guys... i am a newbie here... I just want to ask
I just bought a speaker with a Piezo tweeter... i just test it with my Behringer 1202 Mixer. When i put it in, the Piezo Tweeter keeps on ringing and it is very annoying. So what seems to be the problem? the Piezo? the Capacitor? the Crossover? Thank you so much
Good day guys... i am a newbie here... I just want to ask
I just bought a speaker with a Piezo tweeter... i just test it with my Behringer 1202 Mixer. When i put it in, the Piezo Tweeter keeps on ringing and it is very annoying. So what seems to be the problem? the Piezo? the Capacitor? the Crossover? Thank you so much
Is it a POEWER mixer ? I.e. a mixing console with built-in power amp ? If yes, then maybe the power amp doesn't like piezos (-> capacitive load) and is prone to instability.
In this case a small modification to your speaker would be necessary (series resistor for the piezo).
Regards
Charles
In this case a small modification to your speaker would be necessary (series resistor for the piezo).
Regards
Charles
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.