You will have seen that I'm pleased with my ER18RNX/27TDFC MTMs, so why a possible project?
It started when Mark (Islandpink) told me of his very successful speaker, Purify 6.5 inch with Satori beryllium in a TM reflex box; I know Mark and trust his judgement highly, so I know this is an exceptionally good speaker.
I need more sensitivity so I was wondering how good a MTM might be with these drivers.
Fun project?
I like what I have. The usual questions about an upgrade:
What don't you like about the present speakers?
What do you think could be improved?
What kind of change are you looking for?
My answers are pretty much nothing to all. So the project is a bit of a risk, in that I might find no substantial difference, or I might not like the differences. So I'd do it as a fun project,, for the sake of trying something (possibly) different; also I need to keep active for health reasons and a nice bit of DIY work is a good thing.
Research into the Purifi also suggested that Seas W18NX-003 nextel driver is of a similarly high quality, and this is often used with Seas Crescendo tweeter to good effect; yes, I know that tweeter is said to be overpriced, but if it's kind of similar to 27TDFC but usefully better it could suit me very well.
I know there are lots of good drivers out there, but the ER18RNX is excellent so I think only the above woofers are likely to be better.
The WAF factor remains the same, so size is limited to about 40ins high, 8 inches wide (perhaps 8.5), 12 inches deep (perhaps 13); MTM to get the cone area I need for SET amp.
2 choices so far.
Seas W18nx003, Crescendo tweeter; would give me very much what I have, but hopefully somewhat improved, little risk, not a major gain. Fairly easy to design and develop a crossover. Safe option, but less gain?
Purifi PTT6.5W08-NFA-01 (the higher sensitivity one), Satori beryllium; looks to give exceptional transparency, very good sound generally, likely more to gain but higher risk. Crossover could be awkward. Some people find the need for corrections to be applied; Dennis Murphy reckons it's harder to work with than most, may need 4th order crossover and I'm way out of my depth here!
However, these drivers are used in his Salk Bepure 2 in a 2 way TMM, and reports about this speaker are amazingly good, highly tempting. This, and Mark's info, are tempting me to have a go with these drivers.
If so, I'll need to do a lot of research into crossovers and will definitely need help here.
I think both would be excellent speakers.
Yes, I know the SEAS drivers would fit the existing cabinet. The problem is, I'd find it hard to evaluate the results, I really need both at the same time. But it's something to consider for the future.
If I go ahead, it won't be just yet, possibly about May time.
Any comments welcomed.
It started when Mark (Islandpink) told me of his very successful speaker, Purify 6.5 inch with Satori beryllium in a TM reflex box; I know Mark and trust his judgement highly, so I know this is an exceptionally good speaker.
I need more sensitivity so I was wondering how good a MTM might be with these drivers.
Fun project?
I like what I have. The usual questions about an upgrade:
What don't you like about the present speakers?
What do you think could be improved?
What kind of change are you looking for?
My answers are pretty much nothing to all. So the project is a bit of a risk, in that I might find no substantial difference, or I might not like the differences. So I'd do it as a fun project,, for the sake of trying something (possibly) different; also I need to keep active for health reasons and a nice bit of DIY work is a good thing.
Research into the Purifi also suggested that Seas W18NX-003 nextel driver is of a similarly high quality, and this is often used with Seas Crescendo tweeter to good effect; yes, I know that tweeter is said to be overpriced, but if it's kind of similar to 27TDFC but usefully better it could suit me very well.
I know there are lots of good drivers out there, but the ER18RNX is excellent so I think only the above woofers are likely to be better.
The WAF factor remains the same, so size is limited to about 40ins high, 8 inches wide (perhaps 8.5), 12 inches deep (perhaps 13); MTM to get the cone area I need for SET amp.
2 choices so far.
Seas W18nx003, Crescendo tweeter; would give me very much what I have, but hopefully somewhat improved, little risk, not a major gain. Fairly easy to design and develop a crossover. Safe option, but less gain?
Purifi PTT6.5W08-NFA-01 (the higher sensitivity one), Satori beryllium; looks to give exceptional transparency, very good sound generally, likely more to gain but higher risk. Crossover could be awkward. Some people find the need for corrections to be applied; Dennis Murphy reckons it's harder to work with than most, may need 4th order crossover and I'm way out of my depth here!
However, these drivers are used in his Salk Bepure 2 in a 2 way TMM, and reports about this speaker are amazingly good, highly tempting. This, and Mark's info, are tempting me to have a go with these drivers.
If so, I'll need to do a lot of research into crossovers and will definitely need help here.
I think both would be excellent speakers.
Yes, I know the SEAS drivers would fit the existing cabinet. The problem is, I'd find it hard to evaluate the results, I really need both at the same time. But it's something to consider for the future.
If I go ahead, it won't be just yet, possibly about May time.
Any comments welcomed.
Woofer choice, BSC, sensitivity, bass alignment.
I still think the SEAS nextel and Crescendo MTM would be very good, but I've been looking further into Purifi/ beryllium.
The sensitivity issue of Purify has gone away, because of BSC and correction.
I know the ER18RNX and W18NX003 bass drops away significantly and needs a correction rather like BSC, by a large inductor. Typically about 2.5 to 3 mH with one woofer, 1.4mH with parallel woofers. Looking at FR plots and other data suggests that these drivers need about 3 to 4 dB correction, with about 3dB of BSC in addition.
In my case with the ER18s, the speakers are a touch closer to the rear wall than ideal so my inductor is 1.2mH and I guess the BSC is 2 to 3 dB.
Now the Purifi seems to need no such correction. Typical systems seem to have about 2dB of BSC, such as Mark's with 0.56mH on his 4 ohm woofer. I've also seen 0.7mH.
It all seems to add up. PTT6.5X08 is 87dB roughly; allow 2dB BSC, allow 1dB loss in crossover, add 6 for twin drivers and result is 90dB sensitivity, which agrees with the Bepure 2.
Simulation of my existing speakers gives 90dB; a similar speaker with W18NX003 would be about the same, a touch less.
Modelling crossovers with the Purifi in parallel and about 0.75mH looks to give what I think is the right BSC (no correction needed) of about 2 dB.
So I'm confident the Purifi would give me adequate sensitivity, and I'm planning to use this make..
Now, which? X or W version?
I had been thinking of the W for the extra sensitivity, but I tend to favour the X because it's more of a known quantity; used by Mark, Salk and others.
The bass of the X seems to be better - a major plus factor of the Bepure 2. Also, perhaps the W might benefit from a touch more BSC, if so this will lose some of its advantage of sensitivity.
Bass alignment looks OK; in about a 45 litre box, port tuned to 25 Hz gives a slight peak about 0.8dB at 60-70 Hz; -3dB at 30Hz, -6dB at about 24 Hz; group delay 6 at 50Hz, 9 at 40Hz, 13 at 30Hz. The peak is similar to the ER18 MTM, I do not think it would be a problem.
Any comments welcomed
I still think the SEAS nextel and Crescendo MTM would be very good, but I've been looking further into Purifi/ beryllium.
The sensitivity issue of Purify has gone away, because of BSC and correction.
I know the ER18RNX and W18NX003 bass drops away significantly and needs a correction rather like BSC, by a large inductor. Typically about 2.5 to 3 mH with one woofer, 1.4mH with parallel woofers. Looking at FR plots and other data suggests that these drivers need about 3 to 4 dB correction, with about 3dB of BSC in addition.
In my case with the ER18s, the speakers are a touch closer to the rear wall than ideal so my inductor is 1.2mH and I guess the BSC is 2 to 3 dB.
Now the Purifi seems to need no such correction. Typical systems seem to have about 2dB of BSC, such as Mark's with 0.56mH on his 4 ohm woofer. I've also seen 0.7mH.
It all seems to add up. PTT6.5X08 is 87dB roughly; allow 2dB BSC, allow 1dB loss in crossover, add 6 for twin drivers and result is 90dB sensitivity, which agrees with the Bepure 2.
Simulation of my existing speakers gives 90dB; a similar speaker with W18NX003 would be about the same, a touch less.
Modelling crossovers with the Purifi in parallel and about 0.75mH looks to give what I think is the right BSC (no correction needed) of about 2 dB.
So I'm confident the Purifi would give me adequate sensitivity, and I'm planning to use this make..
Now, which? X or W version?
I had been thinking of the W for the extra sensitivity, but I tend to favour the X because it's more of a known quantity; used by Mark, Salk and others.
The bass of the X seems to be better - a major plus factor of the Bepure 2. Also, perhaps the W might benefit from a touch more BSC, if so this will lose some of its advantage of sensitivity.
Bass alignment looks OK; in about a 45 litre box, port tuned to 25 Hz gives a slight peak about 0.8dB at 60-70 Hz; -3dB at 30Hz, -6dB at about 24 Hz; group delay 6 at 50Hz, 9 at 40Hz, 13 at 30Hz. The peak is similar to the ER18 MTM, I do not think it would be a problem.
Any comments welcomed
W18NX003 needs only a 7dB notch at 4.3k to flatten and thats pretty much it. If the cost wasn't so astronomical, I'd build MTMs with it using a very good HF ribbon crossed around 2.8k 3rd order.
I used this driver in a smaller 2 way with the Audax TW025A28. Midrange sounded like an electrostat. Likely the highest resolving, lowest distortion 6" cone driver I've ever heard, better than Purifi or ScanSpeak. Bass is very resolving and analytical, but not in any way lean. Compared to my HD650s running balanced through the RME ADI2 (about 0.2% THD) just as resolving and very little difference throughout the entire midrange. Only thing better I've heard is a large 3 way with Bliesma M74A.
This Seas Reminds me oddly of the older Peerless revelators without the typical big 1k wiggle that spoils the mids. This would be a great driver for no-compromise 2 way nearfields.
I used this driver in a smaller 2 way with the Audax TW025A28. Midrange sounded like an electrostat. Likely the highest resolving, lowest distortion 6" cone driver I've ever heard, better than Purifi or ScanSpeak. Bass is very resolving and analytical, but not in any way lean. Compared to my HD650s running balanced through the RME ADI2 (about 0.2% THD) just as resolving and very little difference throughout the entire midrange. Only thing better I've heard is a large 3 way with Bliesma M74A.
This Seas Reminds me oddly of the older Peerless revelators without the typical big 1k wiggle that spoils the mids. This would be a great driver for no-compromise 2 way nearfields.
Whatever I build, I will use an external crossover to allow easy tuniing.
My initial thoughts are to make and try:
2nd order on both, as that's what I use now
2nd order on woofer, 3rd order on tweeter, as it's a common method and may have better phase
Here is an attempt in Xsim; Purifi 6.5 X 8 ohm; the tweeter is 4ohm, I have yet to model with 8ohm and decide which.
Looks good to me but what do I know.
My initial thoughts are to make and try:
2nd order on both, as that's what I use now
2nd order on woofer, 3rd order on tweeter, as it's a common method and may have better phase
Here is an attempt in Xsim; Purifi 6.5 X 8 ohm; the tweeter is 4ohm, I have yet to model with 8ohm and decide which.
Looks good to me but what do I know.
You clearly enjoy your existing build and with no clear goals other than to surpass it, there’s a few ways to look at things…
One….objectively ’better’ drivers……..if the implementation is well thought out, and adding in suggestive bias that you worked hard on the design and spent more, your goal might be realized.
Two……re evaluate your goal and use DIY to the full spectrum advantage it offers……design with a very specific use case where it can perform best in your unique environment and to your performance preferences……the latter being much more difficult to define
the ER drivers are excellent and given their affordability, IMO, there’s no proportional improvement out there……modest improvements for a LOT more $$$ and effort……same goes for the SB26 tweets.
One….objectively ’better’ drivers……..if the implementation is well thought out, and adding in suggestive bias that you worked hard on the design and spent more, your goal might be realized.
Two……re evaluate your goal and use DIY to the full spectrum advantage it offers……design with a very specific use case where it can perform best in your unique environment and to your performance preferences……the latter being much more difficult to define
the ER drivers are excellent and given their affordability, IMO, there’s no proportional improvement out there……modest improvements for a LOT more $$$ and effort……same goes for the SB26 tweets.
I would suggest an alternative, a more difficult route of talking to the spouse in such a way that
waf might be altered to allow a wider cabinet employing drive units for a "simple" 4 way, midrange
dome based solution.
waf might be altered to allow a wider cabinet employing drive units for a "simple" 4 way, midrange
dome based solution.
You can make narrow front baffle for 4way, with woofer on the side. NHT had few models like that.
Good thoughts, thanks for your interest. Just a comment or two.
Baffle width is pretty much fixed as is, no chance of a 8inch driver on the front. Woofer on the side is possible.
But I don't think I need a larger woofer or 4 way if I go for Purifi; the bass performance of dual Purifi is excellent (see comments abour Bepure 2).
My limited experience with 6.5in drivers (2 different speakers) suggests that they can give a very good midrange, so the benefit of a dedicated midrange might not be that much in my case, particularly as I'd have the complexity of the complex crossover - and if, as is likely, I don't get it quite right this may lose any benefit.
So I intend to stay with the relatively simple MTM 6.5 inch.
Best regards!
Baffle width is pretty much fixed as is, no chance of a 8inch driver on the front. Woofer on the side is possible.
But I don't think I need a larger woofer or 4 way if I go for Purifi; the bass performance of dual Purifi is excellent (see comments abour Bepure 2).
My limited experience with 6.5in drivers (2 different speakers) suggests that they can give a very good midrange, so the benefit of a dedicated midrange might not be that much in my case, particularly as I'd have the complexity of the complex crossover - and if, as is likely, I don't get it quite right this may lose any benefit.
So I intend to stay with the relatively simple MTM 6.5 inch.
Best regards!
After a long time, here is an update and questions.
I'm going ahead with the Purifi/beryllium MTM, I have the bass drivers and have started the woodwork. I am enjoying the latter, which was half the reason for doing this project.
My wife thinks I'm utterly mad as she really loves the existing speakers (MTM, ER18RNX, 27TDFC); and so do I !
I think the idea and basic design of the new one is good. I had an awkward moment when I saw that the Purifi PTT6.5-X08-NFA-01 works well into a relatively small enclosure, discussed here. I modelled the box allowing for my amp's high Zout which has a major effect, ended up at 44 litres ported to 24 to 25 Hz, very different to what was quoted. However, an expert concluded that a larger volume could work well if Q was increased - by adding resistance in series. This is my situation, so I think I'm OK, I think my model and design is OK. I can try changing the port tuning later if necessary.
My issue now is the choice of version of the Satori beryllium tweeter; 4 or 8 ohms nominal?
4 would seem somewhat compatible with the 4 ohms of the bass drivers in parallel. But 8 might put less of a load on the SET amp which is limited in power and current (but has good drive and musicality, far better than the watts would suggest).
I attach a possible crossover and response for the 4 ohm tweeter. Note the 2dB scale. The notch has been discussed here very recently, I think it is OK but will try without it.
Now similar for the 8 ohm tweeter.
The FR of the 4 ohm one appears better. I've tried a fair amount of playing around and these are about the best I can come up with but of course I am no expert! Others may well be able to improve on this.
But based on the above, I tend to go for the 4 ohm tweeter and I think this xover is worth building - then trying and measuring.
I do not know what crossover is used in the Salk Bepure 2 but I suspect it may be 2nd bass, 3rd tweeter as Dennis tends to use this e.g. in the 'New DIY MTM Towers on Audioholics.
Any comments on this decision, and on the crossovers in general, welcomed.
Both give very good nulls when the tweeter is reversed; 4 ohm at the top at about 1850Hz, 8 ohm (the lower one) at about 1950Hz.
I'm going ahead with the Purifi/beryllium MTM, I have the bass drivers and have started the woodwork. I am enjoying the latter, which was half the reason for doing this project.
My wife thinks I'm utterly mad as she really loves the existing speakers (MTM, ER18RNX, 27TDFC); and so do I !
I think the idea and basic design of the new one is good. I had an awkward moment when I saw that the Purifi PTT6.5-X08-NFA-01 works well into a relatively small enclosure, discussed here. I modelled the box allowing for my amp's high Zout which has a major effect, ended up at 44 litres ported to 24 to 25 Hz, very different to what was quoted. However, an expert concluded that a larger volume could work well if Q was increased - by adding resistance in series. This is my situation, so I think I'm OK, I think my model and design is OK. I can try changing the port tuning later if necessary.
My issue now is the choice of version of the Satori beryllium tweeter; 4 or 8 ohms nominal?
4 would seem somewhat compatible with the 4 ohms of the bass drivers in parallel. But 8 might put less of a load on the SET amp which is limited in power and current (but has good drive and musicality, far better than the watts would suggest).
I attach a possible crossover and response for the 4 ohm tweeter. Note the 2dB scale. The notch has been discussed here very recently, I think it is OK but will try without it.
Now similar for the 8 ohm tweeter.
The FR of the 4 ohm one appears better. I've tried a fair amount of playing around and these are about the best I can come up with but of course I am no expert! Others may well be able to improve on this.
But based on the above, I tend to go for the 4 ohm tweeter and I think this xover is worth building - then trying and measuring.
I do not know what crossover is used in the Salk Bepure 2 but I suspect it may be 2nd bass, 3rd tweeter as Dennis tends to use this e.g. in the 'New DIY MTM Towers on Audioholics.
Any comments on this decision, and on the crossovers in general, welcomed.
Both give very good nulls when the tweeter is reversed; 4 ohm at the top at about 1850Hz, 8 ohm (the lower one) at about 1950Hz.
This 2 ohms represents the output impedance of the SET amp. OK, it may not be a pure resistance, but I think it will be mainly resistive as it is the reflection through the output transformer of the internal resistance of the output valve, and this is mainly resistive. So it will not be 100% accurate, but close enough to be useful I think. It does make a substantial difference to the crossover.
Sounds like a fun project and in the spirit of YOLO, why not also look at the Scan 6640, 7140, or Bliesma T34B?
(I love notionally spending other people's money)
(I love notionally spending other people's money)
Yes, it is a fun project as I may not be quite the first to do this but I'm not following someone else's design, I'm enjoying working things out for myself.
And I do enjoy and appreciate the advice and support here.
As I enjoy the engineering, development, doing something new (I'm an engineer), I'm getting a lot out of this; whatever happens it is not a waste of money, so don't worry about spending mine for me! I may look up these other units but I'm almost certain to use the Satori as it's used by Mark (who I know and trust) and in the Bepure 2, so it's somewhat a known quantity.
Today I listened on the existing speakers to this:
Exquisite. Emotional, compelling, delicate yet powerful, the tonality was superb, sound quality of everything was excellent. Top notch enjoyment.
I think my wife might be right, how on earth can I even match this, never mind improve on it. I will enjoy trying and seeing where I get to.
And I do enjoy and appreciate the advice and support here.
As I enjoy the engineering, development, doing something new (I'm an engineer), I'm getting a lot out of this; whatever happens it is not a waste of money, so don't worry about spending mine for me! I may look up these other units but I'm almost certain to use the Satori as it's used by Mark (who I know and trust) and in the Bepure 2, so it's somewhat a known quantity.
Today I listened on the existing speakers to this:
Exquisite. Emotional, compelling, delicate yet powerful, the tonality was superb, sound quality of everything was excellent. Top notch enjoyment.
I think my wife might be right, how on earth can I even match this, never mind improve on it. I will enjoy trying and seeing where I get to.
Progress is slow but positive; I have to work at low speed and effort but it's good therapy.
The laminated sides are made and chamfered. These are 3 layers of 4mm birch ply, 2 layers 4mm MDF (alternated) giving 20mm thickness..
Rear panels are ready, front panels almost done, internal braces are here, so I will start assembly soon.
Now I have questions about damping the panels. I have read up a lot here and elsewhere (Troels seems to talk sense on the subject) so I have ideas but I'm not sure how good..
I read that bitumen does not absorb but reflects, useful to add mass; but with my 30mm front MDF panels and 20mm sides, I guess I don't need this. I might add some to the 18mm rear panel.
I also read that open cell foam is better than closed cell so I propose to use this as it is open cell with excellent sound absorption:
https://www.ebay.co.uk/itm/302640616227
I'm thinking of lining the sides and rear with 25mm in the upper half of the speakers, i.e. behind the drivers, and perhaps 12mm in the lower half. Also perhaps 2 x 25mm under the top and 3 x 25mm on the base to reduce vertical waves. It's guesswork, aiming at not being over damped.
The existing speakers of the same size and type use bitumen/foam pads from Wilmslow on surfaces in the upper half, nothing on surfaces in the lower half, 3 layers felt carpet underlay on the bottom and results must be OK as I love the speakers. I'm half tempted to do the same again.
Any comments welcomed.
The laminated sides are made and chamfered. These are 3 layers of 4mm birch ply, 2 layers 4mm MDF (alternated) giving 20mm thickness..
Rear panels are ready, front panels almost done, internal braces are here, so I will start assembly soon.
Now I have questions about damping the panels. I have read up a lot here and elsewhere (Troels seems to talk sense on the subject) so I have ideas but I'm not sure how good..
I read that bitumen does not absorb but reflects, useful to add mass; but with my 30mm front MDF panels and 20mm sides, I guess I don't need this. I might add some to the 18mm rear panel.
I also read that open cell foam is better than closed cell so I propose to use this as it is open cell with excellent sound absorption:
https://www.ebay.co.uk/itm/302640616227
I'm thinking of lining the sides and rear with 25mm in the upper half of the speakers, i.e. behind the drivers, and perhaps 12mm in the lower half. Also perhaps 2 x 25mm under the top and 3 x 25mm on the base to reduce vertical waves. It's guesswork, aiming at not being over damped.
The existing speakers of the same size and type use bitumen/foam pads from Wilmslow on surfaces in the upper half, nothing on surfaces in the lower half, 3 layers felt carpet underlay on the bottom and results must be OK as I love the speakers. I'm half tempted to do the same again.
Any comments welcomed.
Thanks, noted, and I will do this click test when the speakers are built.
I plan to leave the top and base removeable at first, screwed in place and sealed with a bead of silicone sealer. I will then be able to adjust the damping. Later those pieces will be fully glued.
I'm hoping my initial plan as above might be reasonable as changing this lining later will be possible, but not that easy.
I plan to leave the top and base removeable at first, screwed in place and sealed with a bead of silicone sealer. I will then be able to adjust the damping. Later those pieces will be fully glued.
I'm hoping my initial plan as above might be reasonable as changing this lining later will be possible, but not that easy.
As nobody has criticised my proposal, and I have a very useful response from GM, I'm going ahead as above with a small change.
First, I noted Norman Tracy's comment about the Purifi drivers:
"every time the enclosure system is made stiffer, more inert, and damped one learns the driver has more to give"
With 20mm curved laminated sides, 30mm front panel, and 5 braces, stiffness should be OK, but I decided to try slightly more damping.
In looking into damping, I noticed that EPDM seems well regarded. Now it seems to me that 2 different materials could work well together so I will use some.
In the top half of the speakers, where the drivers are (MTM, remember), 5mm EPDM and 25mm open foam as above.
In the bottom half:
5mm EPDM and 12mm open foam on the rear panel, 12mm open foam on other surfaces.
2 layers 25mm open foam under the top, 2 or 3 layers open foam on the base, to damp vertical reflections.
First, I noted Norman Tracy's comment about the Purifi drivers:
"every time the enclosure system is made stiffer, more inert, and damped one learns the driver has more to give"
With 20mm curved laminated sides, 30mm front panel, and 5 braces, stiffness should be OK, but I decided to try slightly more damping.
In looking into damping, I noticed that EPDM seems well regarded. Now it seems to me that 2 different materials could work well together so I will use some.
In the top half of the speakers, where the drivers are (MTM, remember), 5mm EPDM and 25mm open foam as above.
In the bottom half:
5mm EPDM and 12mm open foam on the rear panel, 12mm open foam on other surfaces.
2 layers 25mm open foam under the top, 2 or 3 layers open foam on the base, to damp vertical reflections.