Here's a scan (sorry about the quality) of the MC4 test results from Hi Fi Choice.Definitely looks like the tweeter is voiced a couple of dB too high.Increasing the value of the series resistor in the tweeter network could be worth a try?
Attachments
the Scales' figures ?
Hi eanee,
just what we need - thankyou very much indeed !
The frequency Response plots confirm what was in my memory about the balance of treble to midrange for those l'speakers.
Adding a series connected resistor to the tweeter circuit will alter the shape of the crossover's treble filter.
What I need to know is what is in that filter in the crossover, and there may be 1 or 2 resistors there currently,
thus my questions to Kris in #17.
Changes to the crossover will have to be done carefully so as to not affect the integration of the treble with the midrange.
The Plots themselves are clear enough in your Scan, but not the numerals on the Left Vertical Scale of each plot.
Please can you look at the original and Post here:
(1) - for the top plot, what is the number for the dB spacing between the two horizontal lines ?
(2) - for the 2nd plot, what are the three numbers at the left edge of the three horizontal lines ?
(3) - for the 4th plot {Impedance}, what are the four numbers at the left edge of the four horizontal lines ?
In particular I need to know the Impedance values so as to be able to decide how to not detrimentally change the impedance whilst usefully lowering the treble output.
Hi eanee,
just what we need - thankyou very much indeed !
The frequency Response plots confirm what was in my memory about the balance of treble to midrange for those l'speakers.
Adding a series connected resistor to the tweeter circuit will alter the shape of the crossover's treble filter.
What I need to know is what is in that filter in the crossover, and there may be 1 or 2 resistors there currently,
thus my questions to Kris in #17.
Changes to the crossover will have to be done carefully so as to not affect the integration of the treble with the midrange.
The Plots themselves are clear enough in your Scan, but not the numerals on the Left Vertical Scale of each plot.
Please can you look at the original and Post here:
(1) - for the top plot, what is the number for the dB spacing between the two horizontal lines ?
(2) - for the 2nd plot, what are the three numbers at the left edge of the three horizontal lines ?
(3) - for the 4th plot {Impedance}, what are the four numbers at the left edge of the four horizontal lines ?
In particular I need to know the Impedance values so as to be able to decide how to not detrimentally change the impedance whilst usefully lowering the treble output.
Hi Alan,
Hopefully these scans should be easier to decipher.
Impedance curve looks pretty benign.
IIRC the MC4 crossover circuit had a 12dB low pass section with approx 3mH series inductor,6uF rolloff cap with 1 ohm series damping resistor.
High pass was 18db,3uF/7uF series caps plus .25mh inductor roll off.
Series input resistor to high pass of approx 3 ohm.
Cheers,Ian
Hopefully these scans should be easier to decipher.
Impedance curve looks pretty benign.
IIRC the MC4 crossover circuit had a 12dB low pass section with approx 3mH series inductor,6uF rolloff cap with 1 ohm series damping resistor.
High pass was 18db,3uF/7uF series caps plus .25mh inductor roll off.
Series input resistor to high pass of approx 3 ohm.
Cheers,Ian
Attachments
Wow! I go away for a few days and you guys get busy on my behalf. 😎
I am game to give these ideas a go. Just one thing bothers me a little. Martin Colloms is a highly respected 'speaker designer. Would he really have made a "mistake" in selecting components for the crossover?
Apart from that. Alan, I can open up the MC4's and take a 'photo of the crossover if that will help?
Changing components in the crossover sounds like the first thing to try as that will be inexpensive.
If no joy then Toasters reccomendations for tweeters sound very informed.
I would be inclined to try a modestly priced soft dome tweet that matched the spec of the existing.
This all sounds far more sensible than throwing money at the project.
Thanks again for the help guys.
I am game to give these ideas a go. Just one thing bothers me a little. Martin Colloms is a highly respected 'speaker designer. Would he really have made a "mistake" in selecting components for the crossover?
Apart from that. Alan, I can open up the MC4's and take a 'photo of the crossover if that will help?
Changing components in the crossover sounds like the first thing to try as that will be inexpensive.
If no joy then Toasters reccomendations for tweeters sound very informed.
I would be inclined to try a modestly priced soft dome tweet that matched the spec of the existing.
This all sounds far more sensible than throwing money at the project.
Thanks again for the help guys.
With little information I have (if the tweeters are in good condition/keepers) I would buy new 3/7 caps and adjust resistor then (+/- 3R3).
I found the classic Celestion HF1300 was one of the best that I tried in combination with the Audax TPX.It's a sweet sounding tweeter that mates very well with the Audax and the existing MC4 crossover.They're easy enough to find on Ebay.
The Celestion is also one that Martin Colloms recommends highly in his book "High Performance Loudspeakers"
The Celestion is also one that Martin Colloms recommends highly in his book "High Performance Loudspeakers"
There seems to be a few interesting options. What I am trying to achieve is a sound that engages me as much as my car stereo does. That has been the result of a lot of work and no small expense plus the help of a very knowledgeable friend. With the home set up I have always had a sense of " Am I having fun yet". Since hearing what a really good set up can do my answer is "NO!" Despite having some very highly regarded kit. The room is a massive limitation as the dimensions create standing waves at unfortunate frequencies. I have worked to ameliorate this with limited success. it is the upper treble which really gets to me though. Altering the crossover is the cheapest and in some ways easiest option but will that change the character of the tweeter for the better?
Try it and see what you think is the best advice. The HF1300 is a good tweeter. I listened to some of these in an old Celestion speaker for years, back in the day. They don't take much power though, and do sound a bit dull. If the crossover mod doesn't do it for you, another serious contender in the replacement tweeter stakes is the BG Neo 3 planar tweeter. I'm just about to hook up a pair of them into my current speaker system/ pile of drive units, so I'll be able to talk sensibly about these soon.
Nice one Toaster. I am waiting for Alan to come back to me on here as he seems to be the man to sort out the crossover mod. If I then go the tweeter route I was hoping to enlist your help. Ribbon tweets eh? I always fancied the idea of that.
I'm also very interested in what the other posters have said. Alan really got into it for me!
I'm also very interested in what the other posters have said. Alan really got into it for me!
I've had a listen to the BG today, along with a Scan-Speak 10F midrange unit, mostly to get an idea of the capabilities of these units as I get into a NaO Note build. First impressions in the current kludge are very good. The tweeter seems particularly good at separating out different strands of music and does 'space' very well. It seems (subjectively) low in colouration, but still allows music to sound lively and interesting- I'm impressed so far. Couldn't hear a lot wrong with the 10F either, although that's not directly relevant here. Good luck with the tweeter level mod.
Your experience with the MC4 ?
Hi eanee,
Thankyou indeed for posting the clearer plots in which I can easily read the numerals on the Left vertical axis,
and thankyou for posting the drawn schematic !
Did you once own these l'speakers, or did a friend ?
I am wondering how you came to know so much about them,
and that you noticed this Thread.
***************
Celestion HF1300 was a very good tweeter in its day, and may still be ... I will have to hear one again to re-evaluate.
As was common with tweeters of its era, it was a bit low in both Power rating and in Sensitivity,
and in its particular case its frequency response extened to only about 13kHz and then sharply dropped off ...
{thus why some manufacturers then used the original Coles supertweeter with it to extend treble towards 20kHz}.
Coles now make a tweeter that looks very like a Celestion HF 1300 !
Perhaps they bought the original Celestion tooling - have a look in:
Magnetman
I see the MC4's treble drops off very sharply from about 11.5kHz in that "Hi-Fi Choice" plot, but perhaps what is printed includes a high frequency limitation error in the particular measurement technique used.
Does the Treble extend to any higher frequencies on any of the other plots for any of the other brands of l'speakers in that Issue of "Hi-Fi Choice" ?
Hi eanee,
Thankyou indeed for posting the clearer plots in which I can easily read the numerals on the Left vertical axis,
and thankyou for posting the drawn schematic !
Did you once own these l'speakers, or did a friend ?
I am wondering how you came to know so much about them,
and that you noticed this Thread.
***************
Celestion HF1300 was a very good tweeter in its day, and may still be ... I will have to hear one again to re-evaluate.
As was common with tweeters of its era, it was a bit low in both Power rating and in Sensitivity,
and in its particular case its frequency response extened to only about 13kHz and then sharply dropped off ...
{thus why some manufacturers then used the original Coles supertweeter with it to extend treble towards 20kHz}.
Coles now make a tweeter that looks very like a Celestion HF 1300 !
Perhaps they bought the original Celestion tooling - have a look in:
Magnetman
I see the MC4's treble drops off very sharply from about 11.5kHz in that "Hi-Fi Choice" plot, but perhaps what is printed includes a high frequency limitation error in the particular measurement technique used.
Does the Treble extend to any higher frequencies on any of the other plots for any of the other brands of l'speakers in that Issue of "Hi-Fi Choice" ?
Hi Alan. I have to put this project on hold as I am under orders to decorate my daughters bedroom. While that is going on she is moving into my listening room. As soon as these traumatic events are behind me I will get back on here and post up the info you need. Thank you very much for your interest.
Hi Alan,
"Did you once own these l'speakers, or did a friend ?"
I first heard the MC4s about 20years ago and was suitably impressed. Shortly after I came across a DIY speaker article by Martin Colloms in Stereophile magazine which I decided to build.This featured the Audax TPX in a 2way design which appeared to be a clone of the MC4.
Research at that time confirmed the crossover details as being virtually identical.The MC4 was biwired using ICW (related to Claritycap)polypropylenes.
I ended up tweaking the crossover by reducing the 3mH inductor to 2.5mH to bring up the mid level slightly for better integration.
Also introduced a 4dB constant impedance L pad for the tweeter using Mills WW resistors instead of the 3.3 ohm amp side attenuation.This was an audible improvement to my ear at least.
"Celestion HF1300 was a very good tweeter in its day, and may still be ... I will have to hear one again to re-evaluate.
As was common with tweeters of its era, it was a bit low in both Power rating and in Sensitivity"
I managed to procure a pair of HF1300 tweeters ex Spendor who specially selected them to meet tight performance limits for their BC1 monitors.
Despite having evaluated many possible replacements over the years these Celestions have always sounded more natural and less agressive than any would-be contestants,despite their supposed limitations.The HF1300 was certainly popular with the BBC engineers who used them in classic BBC designs such as the LS5/1,LS5/5,LS 5/6. They do roll off but good examples show level response to 15k before doing so.Sensitivity for the stronger magnet version was fairly normal at around 89dB and still required 3-4dB attenuation to match a typical equalised midrange level of around 85-86dB.
Using top quality crossover components is mandatory for the best results of course.
"I see the MC4's treble drops off very sharply from about 11.5kHz in that "Hi-Fi Choice" plot, but perhaps what is printed includes a high frequency limitation error in the particular measurement technique used.
Does the Treble extend to any higher frequencies on any of the other plots for any of the other brands of l'speakers in that Issue of "Hi-Fi Choice" ?
Rechecking the Hi Fi Choice tests it appears that the premature 12k treble rolloff was unique to MC4 and MC2 both using the 30mm Elac dome.Other designs showed good extension to 20k.
Cheers,Ian
"Did you once own these l'speakers, or did a friend ?"
I first heard the MC4s about 20years ago and was suitably impressed. Shortly after I came across a DIY speaker article by Martin Colloms in Stereophile magazine which I decided to build.This featured the Audax TPX in a 2way design which appeared to be a clone of the MC4.
Research at that time confirmed the crossover details as being virtually identical.The MC4 was biwired using ICW (related to Claritycap)polypropylenes.
I ended up tweaking the crossover by reducing the 3mH inductor to 2.5mH to bring up the mid level slightly for better integration.
Also introduced a 4dB constant impedance L pad for the tweeter using Mills WW resistors instead of the 3.3 ohm amp side attenuation.This was an audible improvement to my ear at least.
"Celestion HF1300 was a very good tweeter in its day, and may still be ... I will have to hear one again to re-evaluate.
As was common with tweeters of its era, it was a bit low in both Power rating and in Sensitivity"
I managed to procure a pair of HF1300 tweeters ex Spendor who specially selected them to meet tight performance limits for their BC1 monitors.
Despite having evaluated many possible replacements over the years these Celestions have always sounded more natural and less agressive than any would-be contestants,despite their supposed limitations.The HF1300 was certainly popular with the BBC engineers who used them in classic BBC designs such as the LS5/1,LS5/5,LS 5/6. They do roll off but good examples show level response to 15k before doing so.Sensitivity for the stronger magnet version was fairly normal at around 89dB and still required 3-4dB attenuation to match a typical equalised midrange level of around 85-86dB.
Using top quality crossover components is mandatory for the best results of course.
"I see the MC4's treble drops off very sharply from about 11.5kHz in that "Hi-Fi Choice" plot, but perhaps what is printed includes a high frequency limitation error in the particular measurement technique used.
Does the Treble extend to any higher frequencies on any of the other plots for any of the other brands of l'speakers in that Issue of "Hi-Fi Choice" ?
Rechecking the Hi Fi Choice tests it appears that the premature 12k treble rolloff was unique to MC4 and MC2 both using the 30mm Elac dome.Other designs showed good extension to 20k.
Cheers,Ian
Subjectively, and according to admittedly crude 1/3 8ve measurements I took at the time, the Elac certainly had more hf extension on axis than the Celestion. It may be that the magazine in-room responses are composites/averages of various microphone positions, as Stereophile's in-room measurements are now. With a large dome like the Elac, the power response will be obviously be 'duller' than smaller diameter tweeters, all else being equal. Given the likely need for a supertweeter and powerhandling limitations, the HF1300 or the Coles equivalent wouldn't be my first choice for the money these days. They are good though, and even though the speaker it was in wasn't that great, the HF1300 made it surprisingly hard to move on. I didn't think it was exceptonally smooth, but the basic clarity and liveliness was really good.
Technicalities aside I found the Elac/TDL 30mm dome very explicit sounding but also a bit hard and relentless,paralleling the experience of krismusic in this regard.
Unfortunately this appears to have been an all too common complaint made about the early generation of metal dome tweeters.
Unfortunately this appears to have been an all too common complaint made about the early generation of metal dome tweeters.
follow-ups
Hi Kris,
as I'm not sure of your technical experience, I'll post a few suggestions to include with what I asked in # above:
Most multimeters are not specifically accurate with their low ohms readings,
thus to assist:
(1) - before you measure, touch the measuring probes together and adjust whatever may be possible on your meter to get as close as possible to 0 ohms reading.
(2) - measure a low ohms' resistor also, if you have one, so that we have a Reference.
(3) - the resistors in the MC4 crossover will still be in-circuit after the tweeter is disconnected, thus you will not get a correct reading across those resistors, but do not desolder them from the board, because that will reduce the electrical integrity of the existing joints.
(4) - measure BOTH tweeters, because there may be sample variations.
(5) -and I forgot to include in my last post, if there is any Part Number printed on the back of the tweeters, please write it in your post also.
***************
Hi eanee and Toaster,
Thankyou for the various additional comments.
I have little time today, but I'll be back with a bit more when time is available.
For now I will conclude with a comment about eanee's:
"Technicalities aside I found the Elac/TDL 30mm dome very explicit sounding but also a bit hard and relentless"
There are two possible matters relevant to that.
One is the type of recordings or source material used to evaluate the treble frequencies ... the Elac tweeter may have been revealing the qualitive states of those.
The other possible relevant matter I will post about later, when we get further into modifying this crossover.
Hi Kris,
as I'm not sure of your technical experience, I'll post a few suggestions to include with what I asked in # above:
Most multimeters are not specifically accurate with their low ohms readings,
thus to assist:
(1) - before you measure, touch the measuring probes together and adjust whatever may be possible on your meter to get as close as possible to 0 ohms reading.
(2) - measure a low ohms' resistor also, if you have one, so that we have a Reference.
(3) - the resistors in the MC4 crossover will still be in-circuit after the tweeter is disconnected, thus you will not get a correct reading across those resistors, but do not desolder them from the board, because that will reduce the electrical integrity of the existing joints.
(4) - measure BOTH tweeters, because there may be sample variations.
(5) -and I forgot to include in my last post, if there is any Part Number printed on the back of the tweeters, please write it in your post also.
***************
Hi eanee and Toaster,
Thankyou for the various additional comments.
I have little time today, but I'll be back with a bit more when time is available.
For now I will conclude with a comment about eanee's:
"Technicalities aside I found the Elac/TDL 30mm dome very explicit sounding but also a bit hard and relentless"
There are two possible matters relevant to that.
One is the type of recordings or source material used to evaluate the treble frequencies ... the Elac tweeter may have been revealing the qualitive states of those.
The other possible relevant matter I will post about later, when we get further into modifying this crossover.
Hi Alan. I trust that you saw my previous post saying that I have to put this on hold for a short time. I will take the MC4's apart and get the info you need posted on here as soon as I can but it will inevitably take longer than I expect to sort out my dahghters bedroom. 🙁
interim
Hi Kris,
yes, I saw and read your post.
Those things do take time, but no hurry ...
Post here with the follow-up information when you have the time available.
***************
... and I myself have very little time available today, thus to eanee and Toaster,
I regret I will have to get back to the other points you raised later.
Hi Kris,
yes, I saw and read your post.
Those things do take time, but no hurry ...
Post here with the follow-up information when you have the time available.
***************
... and I myself have very little time available today, thus to eanee and Toaster,
I regret I will have to get back to the other points you raised later.
Last edited:
- Status
- Not open for further replies.