Celestion 66 needs mid-range

[WayneSwann]

66's Crossover

Hi Wayne,


Alan has contributed extensively to not only this thread, but also two others on the same theme of these excellent speakers. Do a search for Ditton 66. He (and others) go into great detail with their experiences of changing caps in the various models of 44 & 66. Alan, in particular, gives great advice on replacing these caps and also compensating for the ESR, present in the old electros but absent from the modern polyprops.

Hi Lucas

Thanks, yes I have been looking at some of these older threads. It's just a matter of getting everything in one place to make it as simple as possible for someone like me!

Regards

Wayne

 

[LucasAdamson]

Dloper did a schematic revealing the extent of his cap changes and ESR resistors. He said it was pretty much there as I remember...it's a good starting point.

Certainly the caps should be changed for as close a value to the original as you can find. Here's the places Alan told me about:. Read post 71 and follow the links. He was helping me with my 44s, but the principles, and indeed most of the design and parts are the same. Awesome cap advice!

Good luck!
Lucas

 

[LucasAdamson]

You could also look at my schematic. It's for the 44s, which have a different midrange speaker and circuit, but it shows a better way to lay out a crossover schematic, in three sections for bass, mid & treble.

I have replaced the tweeter with the Seas 19TFF1, the tweeter caps with a 4uF and a 6uF, no resistors (as old caps were film type, low ESR anyway)

The mid I have replaced with (2R7 4W) at 6uF to earth and (2R2 4W) at 33uF to earth. Both caps are polypropylene Jantszen, so I have ESR resistors in there.

The bass circuits are unchanged, but I replaced the 72uF with Alcaps (modern bipolar electros.)

There, that's all I can tell you....

 

[WayneSwann]

You could also look at my schematic. It's for the 44s, which have a different mid range speaker and circuit, but it shows a better way to lay out a crossover schematic, in three sections for bass, mid & treble.

Hi Lucas,

Thanks great! thanks a lot for your help. It's all becoming a little clearer for me now. I have just received a new Multi-meter through the post so I can take and post the necessary readings Alan requested.

Best Regards

Wayne

 

[WayneSwann]

Wayne,


I can calculate you 2 resistors to make a suitable fixed L-pad for adequate attenuation.

I do need to know the DC resistance of the Realistic tweeters.
A good multimeter would be a more useful purchase than some other things,
and can be used more in diy to come.

Hi Alan,

I have purchased a multimeter so I can now give you the readings, and any future readings you require.

One Realistic tweeter measured 6.5 ohms, the other measured 6.2 ohms.
The mid range speaker in my Celestion 66's are model 'MD500 2" dome, impedance 8 ohm, range 0.5 - 5kHz, Nominal loading 80W, Din 45 500'.

I hope this helps with your calculations. Please see my previous posts regarding the use of these Realistic tweeters, Please ignore trying to make them suitable for interchanging between systems. I would much rather you were able to match them to my 66's, so that I can the best out of them.

Best Regards

Wayne

 

[alan-1-b]

Multimeter brand and Accuracy spec ?

Hi Wayne,

I'm sorry, but I'll have to be brief today.

What is the Brand and Model Number of your multimeter ?

Also, look in its User Manual booklet - what is the Accuracy Specification for the Resistance ranges ?

... and there may be different Accuracy for different Ranges
-{such is common, and mid to high resistance ranges are usually greater accuracy than the lowest range and the very highest range}-
thus look to see what is quoted for Less than 10 ohms, or Less than 5 ohms, or whatever the case may be ..?

I do mean Resistance range - the brief Specification that advertisers state for multimters is usually only the DC Voltage accuracy,
as that is the easiest measurement to make reasonable accuracy for.

The Resistance Specification may state a +/- Percentage {%} AND a +/- number of Counts, or something else ...

I estimated from the Realistic data on the pamphlet you posted, using both the 2.2uF and .27mH and the dB drop around 5kHz wrt above 10kHz that the Impedance in the useable bandwidth is likely a little below 8 ohms, and that usually indicates a DC resistance of approximately a little over 6 ohms, thus so far it seems to be close ... but let me know the Accuracy, and I'll calculate 2 resistors for you to make a simple L-pad so that you can hear the tweeters in close to correct level wrt the 66's mids so that you can decide better their possible applications.
Currently as the tweeters dominate the sound so much, you really will not know what they can be usefully used for with any of your speaker systems.

I'll address the other matters when I have time available,
BUT do NOT buy 4uF and 6uF for new tweeters,
because about 3.6uF and larger than 10uF is likely what such will require,
if you want to keep the .14mH inductor, and that will be quite OK.

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

Hi Lucas,

thanks for helping here !

Hey, if you can stand listening to the SEAS tweeter not well matched whilst you are making other components that is your business, but I will be posting about a crossover which will suit it and other similar tweeters as soon as I have time, but do Post in the 44's relevant Thread if you do decide to continue with where you left off there.

An Active crossover can work well for between bass and Cone mid-drivers, but such can be a problem to tweeters and dome mid-drivers because
DC Offset voltages from the direct coupled power amps hold their voice-coils out of the optimum locations in their magnets' space,
thus earlier distortion and compression.
Active for the lower section and passive for the higher section can be got to work well more simply - Post in the other Thread if you want to pursue that.

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

Hi sba,

sorry, all this stuff and I haven't got back to your very interesting material yet !
I will ... eventually ...

 

[WayneSwann]

Accuracy of new multimeter

Hi Wayne,

I'm sorry, but I'll have to be brief today.

What is the Brand and Model Number of your multimeter ?

Also, look in its User Manual booklet - what is the Accuracy Specification for the Resistance ranges ?

... and there may be different Accuracy for different Ranges
-{such is common, and mid to high resistance ranges are usually greater accuracy than the lowest range and the very highest range}-
thus look to see what is quoted for Less than 10 ohms, or Less than 5 ohms, or whatever the case may be ..?

I do mean Resistance range - the brief Specification that advertisers state for multimeters is usually only the DC Voltage accuracy,
as that is the easiest measurement to make reasonable accuracy for.

The Resistance Specification may state a +/- Percentage {%} AND a +/- number of Counts, or something else ...

Hi Alan,

My new multimeter was purchased from Maplins. The model is 'UT50A Manual ranging Digital Multimeter'. Regarding it's accuracy it states the following:-

'The test leads can add 0.1 ohm to 0.3 ohm error to the resistance measurement. To obtain precision readings in low resistance, that is the range of 200 ohm, short-circuit the input terminals beforehand and record the reading obtained (called this reading (x). (x is the additional resistance from the test lead. Then use the equation:
Measured resistance - (x) = precision readings of resistance'.
I measured (x) to be 0.2. So I deducted 0.2 from the actual readings I obtained and gave you the results, (6.5 & 6.2 ohms).

The accuracy specification reads as follows:
Range Resolution Accuracy
200 ohm 0.1 ohm +/- (0.8% + 3) <--- I used this range
2k ohm 1.0 ohm +/- (0.8% + 1)

I hope this helps, I must admit I'm not exactly clear on this last part!.

As far as everything else is concerned I will not order any items until it becomes clear exactly what I need. Thanks very much for all of your help, I know you have gone through all of this in previous Threads but I really appreciate your efforts in this case, bringing it all together. I know this probably sounds simple but why wont an 'off the self' L-pad be OK? The explanation will probably help me to more understand the basic principles involved. It's all good stuff!!

Best Regards

Wayne

 

[WayneSwann]

Speaker Clamps for Celestion 66's

Hi to everyone reading this Thread!

Does anyone know where I can obtain speaker clamps the same or very similar to the small triangular ones used on the 66's?. I have ordered several different types but they have all been too big for the job! I am missing just one.

Thanks

Wayne

 

[alan-1-b]

more about accuracy, and where can you buy from ?

My new multimeter was purchased from Maplins. The model is 'UT50A Manual ranging Digital Multimeter'.

The accuracy specification reads as follows:
Range Resolution Accuracy
200 ohm 0.1 ohm +/- (0.8% + 3) <--- I used this range
2k ohm 1.0 ohm +/- (0.8% + 1)

I hope this helps, I must admit I'm not exactly clear on this last part!.

I know this probably sounds simple but why wont an 'off the self' L-pad be OK? The explanation will probably help me to more understand the basic principles involved. It's all good stuff!!

Wayne

Hi Wayne,

in 200 ohm range, {that is Maximum 199.9 ohms measureable} ,
with +/-(0.8% +3) accuracy
a 6.5 ohm resistance could display anywhere between 6.1 and 6.9 on one of those meters, depending on the particular sample.
Worse is that a 1.5 ohm resistor, {which you will need 2 of at not less than 1.5 ohms each},
could display anywhere between 1.2 and 1.8 ohms,
thus you will need a Calibration resistor to allow you to use your meter for more accurate low ohms measurements.

The 1.5 ohm resistors themselves are specified +/- 5%, thus greater uncertainty via that meter to between 1.1 and 1.9 ohms.

Do not worry, because a 10 ohm/ 0.1% accurate resistor can be bought for 32p. , albeit via mailorder from Farnell,
and that can be used for Calibration comparison very easily.
I will explain how, if you choose this option.

Can you buy from Farnell's UK branch ?
It is accessable via:- Farnell / Electronic Component Distributors / Suppliers / Electronics, Electrical Parts, Electrical Components and Wholesale Electronics.

The resistors you will need for the L-pads can be bought from there at the same time, all for 50p. , or less, each.

Maplin do not stock suitable resistors in the required values.

Alternately, is there a local parts shop you can visit which stocks wirewound resistors in 1.5 ohm/5watt and 3.3 ohm/10 watt ?

+/- (0.8% +3 counts) really only becomes accurate to better than +/- 1% when the resistance is above half of the full scale,
and that will be from 100 ohms with that meter in its 200 ohms range.


A variable L-pad will only match Resistance for the single resistance it is made for, usually 8 ohms,
but to use your Realistic tweeter with the .14mH inductor in the crossover the resistance will have to be reduced to about 4.4 ohms.
I don't see any point in spending money on different inductance values for this experimental circuit when you do not know whether you will even like the sound from a heavily attenuated tweeter, thus I am advising to do it a cheaper way, for you to hear before you commit larger money.

A Calibration resistor will be useful for other purposes with that multimeter,
thus money not wasted.

 

[WayneSwann]

Do not worry, because a 10 ohm/ 0.1% accurate resistor can be bought for 32p. , albeit via mailorder from Farnell,
and that can be used for Calibration comparison very easily.
I will explain how, if you choose this option.

The resistors you will need for the L-pads can be bought from there at the same time, all for 50p. , or less, each.

Hi Alan,

I see! I don't know why these multi-meters cannot be more accurate? This amount of deviation is rather a lot. Anyway that being the case I will gladly purchase a 'Calibration Resistor' from Farnell Electronic Components!
I have had a quick look at their web site and I see one for £0.32p '(Ref 1353330) Thin Film Chip Resistor', is that the one you are referring to? Also which others will I need to order for the L-pad. Are their any other items I need to order from them now in order to save time in the future?

Best Regards

Wayne

 

[kayelem]

Hi,

I see! I don't know why these multi-meters cannot be more accurate?

They can, they just cost a lot more. I have several meters, one of which is a 6-digit bench DMM (big, mains-powered) which uses four test leads (rather than two) to eliminate errors caused by test-lead resistance. It's capable of accurate low resistance measurements. It did cost more s/h than I've just paid for two new tweeters, but I use it a lot for design/repair/restoration/etc. anyway.

It's possible to get meaningful measurements without going to great expense, using a cheap DMM then correcting the measurements having used the same meter and leads to measure a known resistance. Alan's assuming you don't want to spend more money on a DMM than your speakers, just to post a couple of measurements on a forum...

Regards, Kat

 

[WayneSwann]

It's possible to get meaningful measurements without going to great expense, using a cheap DMM then correcting the measurements having used the same meter and leads to measure a known resistance. Alan's assuming you don't want to spend more money on a DMM than your speakers, just to post a couple of measurements on a forum...

Regards, Kat

Hi Kat,

You're right! I certainly do not want to spend a great deal of money on test equipment. Even though my instinct is to always get the best I can, it has it's limits!. I just don't like hanging around waiting for things to be delivered. Unfortunately I cant think of anywhere here in Milton Keynes I can go to to obtain this particularly accurate resistor. I guess there must be somewhere though. I need to know from Alan what else I need to order for the L-pad. That is if he can specify without calibrating my multimeter.
Please keep in touch if you can help in my quest to get my Celestion 66's sounding their best!

Best Regards

Wayne

 

[WayneSwann]

Do not worry, because a 10 ohm/ 0.1% accurate resistor can be bought for 32p. , albeit via mail order from Farnell,
and that can be used for Calibration comparison very easily.
I will explain how, if you choose this option.

Can you buy from Farnell's UK branch ?
It is accessible via:- Farnell / Electronic Component Distributors / Suppliers / Electronics, Electrical Parts, Electrical Components and Wholesale Electronics.

The resistors you will need for the L-pads can be bought from there at the same time, all for 50p. , or less, each.

Hi Alan,

I need to now what else I can purchase from 'Farnell's' as there seems to be a minimum limit of £20.00 per mail order. I don't mind but I don't know what else to order?

Best Regards

Wayne

 

[LucasAdamson]

So Alan...

I'd like to wind myself some inductors at some point. It would be nice to be able to reliably inform Ditton 44 users of a proven correct way to upgrade their inductors to Ditton 66 levels.

I will need 4x 2.1mH for the midrange shunt inductors and for the bass series, and 2x 3.2mH for the bass series (Thanks to amazing Katherine for this).

The only question that remains is the thickness of the copper to be used and the dimensions of the spool to wind upon. For this, I think the best thing is to copy that of the 66s, which have similar but not identical values.

Could you, or indeed any Ditton 66 Mk1 owner let me know the copper thickness and spool size of the mid shunt and both bass inductors when you next delve into the crossover please...

Many thanks (and on behalf of all upgrading 44 owners in perpetuity!)
Lucas

 

[alan-1-b]

Tweeters and Resistors

Thankyou Kat for helping out !

***********

Yes Wayne,

minimum Price per Order, and there is an Order Multiple {minimum number of parts} for some of the resistors,
so lets decide which Tweeter first, and then you can but resistors for:
Tweeter L-pads; Mids ESR; Bass ESR; all at the same time.

I have studied the Scanspeak data sheets -{two, the 2008 and similar 2010 model}.

Yes, either can be got to work, but ...
both are higher Sensitivity than the SEAS, thus will need greater attenuation,
which in my experience reduces sound quality unnless one is driving them hard ... as I explained in an earlier Post.

Both have a slightly greater rise in treble above 10kHz than the SEAS, thus if you want that, then buy either,
but the SEAS has extended treble above 10kHz to better degree than the old HF2000, and I think that will be sufficient ... unless you want the extra tilt up.

What you are paying more than twice the price for with the Scanspeaks are:
the damped rear chamber and resultant the low Fs, which enables them to be crossed over lower in frequency for 2-way systems, which you do not need for this 3-way.
the higher Sensitivity, which you do not need, because the SEAS is more than sufficiently sensitive.

Perhaps there are other audible advantages with the Scans, but as they are both re-issues of older Scan models there may be audible advantages to the newer technology dome material and termination of the SEAS - I do not know.

For the extra money not spent on the Scans you can buy better audio quality capacitors, which the SEAS are capable of revealing the advantges of.

I think the SEAS are more than good enough to match the old Celestion mid and bass drivers, however you decide ...
then Post here, and I'll list the resistors for whichever tweeter you want.

***********

Lucas,

if you are going to wind inductors, better can be done than copying the old Celestion ones, and such can improve the bass output and quality,
thus ?

 

[WayneSwann]

I think the SEAS are more than good enough to match the old Celestion mid and bass drivers, however you decide ...
then Post here, and I'll list the resistors for whichever tweeter you want.

***********

Hi,

Once again, thanks for your time and effort. I will go ahead with the SEAS. As you say, they are probably more of a match for the 66's. Plus the money can be better spent on the nuts and bolts!

I have ordered the 3 resistors you specified previously via Wilmslow Audio. They are also supplying me with other stuff as well so there isn't a problem there!
They will source the 10 ohm (+/-0.1%) and the 1.5 ohm/5 watt (+/-1%) from Farnell Electronics. The other 3.3 ohm/10 watt (+/-5%)they can supply themselves.

Regards

Wayne

 

[alan-1-b]

risk of problems

Hi Wayne,

again you rush into ordering !

I'm sorry to say, but this resistor application is NOT simple.
It will take time to explain, which I would have done when I had the time available - now lost on this reply.

If you have irrevocably ordered, then I regret this as you will find out the problems when you receive the resistors and try to use them,
and I don't want to use even more of my limited time solving additional problems whilst there are several other things in this Thread that I have not got to yet, including optimally using the SEAS tweeters.

(1) - have you seen how tiny the particular 10 ohm resistor you selected is ?
It will be difficult to work with, a larger physical size will be better ...
and you should buy 2, not 1.

(2) - you will need more than 2 of each of the others, because some samples will be too low in resistance.

(3) - now that you have decided on a tweeter for the Celestions, do you still want to try the Realistic there as sole tweeter
- to hear what it can and can't do ? ,
or, do you want it to be set up to use as a true Super-Tweeter - highest treble frequencies only - in such way that you can try it with any/all your loudspeaker systems ?
If the latter, then 1.5 ohm and 3.3 ohm are not necessary, but different components - a specific variable L-pad plus different capacitors than specified.

(4) - where in the Farnell catalog are there 1.5 ohm 5 watt resistors in +/- 1% tolerance ?
It may be good if there are, depending on what they are made of ...

I'll return to this when I have time available, but you may have wasted money.

 

[WayneSwann]

Hi Wayne,

again you rush into ordering !

I'm sorry to say, but this resistor application is NOT simple.
It will take time to explain, which I would have done when I had the time available - now lost on this reply.

If you have irrevocably ordered, then I regret this as you will find out the problems when you receive the resistors and try to use them,

Hi Alan,

Sorry, don't worry I have cancelled the order. All is OK

Regards

Wayne

 

[WayneSwann]

(1) - have you seen how tiny the particular 10 ohm resistor you selected is ?
It will be difficult to work with, a larger physical size will be better ...
and you should buy 2, not 1.

(3) - now that you have decided on a tweeter for the Celestions, do you still want to try the Realistic there as sole tweeter
- to hear what it can and can't do ? ,
or, do you want it to be set up to use as a true Super-Tweeter - highest treble frequencies only - in such way that you can try it with any/all your loudspeaker systems ?

(4) - where in the Farnell catalog are there 1.5 ohm 5 watt resistors in +/- 1% tolerance ?

Hi Alan,

As I said The order for the resistors has been cancelled. Sorry but I mistook your previous comments as a request to order them. In future I will not order anything until you say 'GO AHEAD AND ORDER!'. So as to avoid any more confusion.

Just for your information -

There are 19 resistors rated 10 ohm (+/- 0.1%).

The one I ordered was not the one costing £0.32, but the large and more expensive:-

Manufacturer: TYCO ELECTRONICS / NEOHM Order Code: 1754843 Manufacturer Part No: UPW25B10RV RESISTOR, 0.1%, 5PPM, 10R

• Resistance:10ohm
• Resistance Tolerance:± 0.1%
• Power Rating:0.25W
• Voltage Rating:300V
• Series:UPW
• Temperature Coefficient:± 5ppm/°C
• Resistor Element Type:Wire Wound
• Resistor Case Style:Axial Leaded
• No. of Pins:2
• Body Diameter:4.75mm
• Case Style:Axial
• Body Length:9.53mm
• Lead Diameter:0.64mm
• Lead Length:25mm
• Operating Temperature Range:-55°C to +145°C

But I could order any of them, dependant on the power rating.

I think I would like the challenge of building a pair of real Super-Tweeters so as to interchange them between systems, but only if you feel this is a feasible and reasonable proposition.
I would also like to configure the 66's crossovers to best use the SEAS tweeter. I would like to give the crossovers a general overhaul, replacing the old capacitors with the newer polypropylene ones.

The more accurate 1.5 ohm/5 watt resistors is:-

Order Code: 1614415 RCD (RESISTORS COILS DELAYLINES) 160-1R50-FBW RESISTOR, WIREWOUND, 1.5OHM, 5W, 1%
Resistance:1.5ohm; Resistance Tolerance:± 1%; Power Rating:5W; Voltage Rating:400V; Series:100; Temperature Coefficient:± 50ppm/°C; Resistor Element Material:Ceram 46M1925

I look forward to hearing from you when you get the time.

Best Regards

Wayne

 

[LucasAdamson]

Wayne,

If I may be so bold as to tell you what you should do, here it is:

Forget about super-tweeters at the very least until you have an excellent fully restored and improved pair of Ditton 66s.

Buy the Seas 19TFF1. It's awesome! (Best get 2 of them )

Replace the caps as per Alan's ordering instructions.

Fit ESR compensating resistors as per Alan's ordering instructions.

Replace the binding posts. Remember that there is an aluminium plate on the current ones that will short out your speakers unless new binding posts are isolated from it.

Sit back, relax, enjoy! You've got one of the most pleasing pairs of loudspeakers ever made, improved beyond compare...

My twopence. Sorry if I'm out of line...
All the best,