This thread really seems to be taking on a life of its own.
So since I started this I'll throw my thoughts into the mix.
Now for my part I'm not the least bit interested in breaking out the microphones and setting up tests that way, I don't deny its valide and I have avtually made my own test tone CDs etc. How do the mods sound on my dirvers, in my cabs, in my room from my lsitening position with my ears is all that really counts for me.
Thinking on what Dave has said I do feel that many hi-fi speakers I have heard that are supposedly pretty flat in the response curve etc sound well ........flat....dead...um uninvolving. I like a little excitement.
For the most part I feel the majority of the tweaks we have discussed are highly likely to work better than the std driver, they make good sense in many ways and a manufacturer would likely implement many of them if productions costs and complications were no issue.
If you are able to get a smoother air flow behind the cone so it better balances that in front it seems logical it should help make the driver more responsive, if you can stop reflection that are out of phase coming back through the cone from the rear, it must help purify the sound somewhat, if the cone flexes less at the junction of the surround it ought to produce better bass when pushed etc. I don't feel there is any snake oil in any of this and compared to some of the illogical stuff you hear about cables. plugs, caps, etc it way up there on the credibility stakes.
Many mods may be questionable, many won't work, but ....and this is my approach if over time I change one thing at a time and then A/B with the unmodded drivers and listen in my room from my position I can clearly hear the differences and make my decisions from that point.
This may not be the most scientific method but hey it works for me.
Now Its not for me or anyone else to say well we should all do this or that, but many mods will probably reap benefits for all sorts of applications and if the drivers are cheap enough why not try.
I should add that with every change I make I get my wife, my son and a my daughter to have a listen and critique and friends as well, they all groan when asked to the music room, but interestingly there is usually a consensus about what works and what doesn't
I do this because, well I don't want to walk too far up my own garden path.
The great thing about forums like this is they get us thinking and examining and just from this thread I have come up with a few further Ideas that I feel will further refine my drivers, so I am very greatful to those who planted the seeds.
So since I started this I'll throw my thoughts into the mix.
Now for my part I'm not the least bit interested in breaking out the microphones and setting up tests that way, I don't deny its valide and I have avtually made my own test tone CDs etc. How do the mods sound on my dirvers, in my cabs, in my room from my lsitening position with my ears is all that really counts for me.
Thinking on what Dave has said I do feel that many hi-fi speakers I have heard that are supposedly pretty flat in the response curve etc sound well ........flat....dead...um uninvolving. I like a little excitement.
For the most part I feel the majority of the tweaks we have discussed are highly likely to work better than the std driver, they make good sense in many ways and a manufacturer would likely implement many of them if productions costs and complications were no issue.
If you are able to get a smoother air flow behind the cone so it better balances that in front it seems logical it should help make the driver more responsive, if you can stop reflection that are out of phase coming back through the cone from the rear, it must help purify the sound somewhat, if the cone flexes less at the junction of the surround it ought to produce better bass when pushed etc. I don't feel there is any snake oil in any of this and compared to some of the illogical stuff you hear about cables. plugs, caps, etc it way up there on the credibility stakes.
Many mods may be questionable, many won't work, but ....and this is my approach if over time I change one thing at a time and then A/B with the unmodded drivers and listen in my room from my position I can clearly hear the differences and make my decisions from that point.
This may not be the most scientific method but hey it works for me.
Now Its not for me or anyone else to say well we should all do this or that, but many mods will probably reap benefits for all sorts of applications and if the drivers are cheap enough why not try.
I should add that with every change I make I get my wife, my son and a my daughter to have a listen and critique and friends as well, they all groan when asked to the music room, but interestingly there is usually a consensus about what works and what doesn't
I do this because, well I don't want to walk too far up my own garden path.
The great thing about forums like this is they get us thinking and examining and just from this thread I have come up with a few further Ideas that I feel will further refine my drivers, so I am very greatful to those who planted the seeds.
MarkMcK said:
Proof of what?
First let's get something out of the way: I am trained as an experimental physicist. I've made measurement - and the analysis of those measurements - my life. I don't believe in magic, I believe in the laws of physics.
I also understand the limitations of any given measurement. More correctly, I understand that the numbers are not responsible for our interpretations; they're just numbers. The significance of the numbers is purely a consequence of human interpretation. In other words, hard science is not *always* as hard as we might like to believe it is. We need to be careful ...
A good example is the notion of THD in amplifiers. It's easy to measure (hard science.) It's even easier to conclude that low THD must sound more accurate. Notice that I didn't say low THD must sound 'better' which opens all sorts speculation about personal tastes and questions of euphonic distortions, etc. No, I said it's easy to conclude that low THD must sound more accurate.
Well, subjective 'accuracy' is still subjective. It's easy (and reasonable) to dismiss one persons testimony that amplifier A sounds more accurate than B even though A produces measurably greater THD. It's not quite as easy when I'm the one doing the listening, but I'm still open to discussion... On the other hand, if MANY people listen to the two amps, and a large majority respond by saying something like "WOW! It's like Satchmo was standing right there in front of me playing! I could hear him rocking back and forth on his hips." Well, it makes you wonder about the conclusion that lower THD must result in subjectively better accuracy.
THD is just an example of a measurement that turned out to not correlate very well with subjectively good sound. Albeit, it's an easy target, but it's still just an example. The measurement is proof of something, I suppose, but not of good sound.
Having said all of that. What kind of proof are you looking for that any given driver modification is an improvement?
Or, to point the question the other way around, what proof do you have that a mod that improves measured IR or FR makes the driver sound better? Will it bring Satchmo into my living room? Can you prove it?
-- Dave
A few random ramblings.
Just thinking about speakers with flat freq response, I suppose that the whole concept of flat response fails to take into account the interaction with the room and our own hearing, which is far from linear, nor would it take into account the loss of HF response we get as we age and a whole raft of other issues related to the box design, listening position, source response etc.
I imagine a speaker with huge peaks and troughs would probably sound pretty rough, but a speaker that would actually sound great in a specific environment would probably need to have a response curve that is optimised for that room......bring on the equalizers!
Anyhow end result aiming for flat response is probably not only impossible but really a sure route to banality, unless of course a equalizer is used to twist the final sound.
A good friend of mine who is in his late 50s and has spent his life an audio tech reckons that modern drivers have tended to lower efficiency and more linear response, but at the cost of dynamics and detail....basically he feels the magic is sucked out of them. Maybe he has something. He tells marvellous stories of whole audatoriums and picture theatres filled with great sound on the smell of a few watts and full/wide range drivers with field coils etc....he almost gets teary eyed!
Just thinking about speakers with flat freq response, I suppose that the whole concept of flat response fails to take into account the interaction with the room and our own hearing, which is far from linear, nor would it take into account the loss of HF response we get as we age and a whole raft of other issues related to the box design, listening position, source response etc.
I imagine a speaker with huge peaks and troughs would probably sound pretty rough, but a speaker that would actually sound great in a specific environment would probably need to have a response curve that is optimised for that room......bring on the equalizers!
Anyhow end result aiming for flat response is probably not only impossible but really a sure route to banality, unless of course a equalizer is used to twist the final sound.
A good friend of mine who is in his late 50s and has spent his life an audio tech reckons that modern drivers have tended to lower efficiency and more linear response, but at the cost of dynamics and detail....basically he feels the magic is sucked out of them. Maybe he has something. He tells marvellous stories of whole audatoriums and picture theatres filled with great sound on the smell of a few watts and full/wide range drivers with field coils etc....he almost gets teary eyed!
And now for another potential tweeky poo.
Listening to my modded drivers, they definitely work better with the modded dust cap, rather than no dust cap at all and a phase plug, but this opens up another possible problem.
The reflection off the back of the dust cap most likely bounces of the pole piece and muddies the highs a little as it comes back through the vent holes.
I imagine the answer is to pad the top of the pole piece but this is not possible as getting the padding through the vent holes and into position would be well nigh impossible.
So my idea is cut the cap out, then pad the pole and make a new cap and araldite it in. I think the cap could be made from aluminium shim, which is rubbed into shape in the neck of a bottle (I have done this before so I know that part works), then have the vents cut into it.
The little support legs on the cap could be padded so only the tiny centre dome actually produces sound.
If your not sure what I mean with the cap look back at the pics I psoted at the start of this thread.
Listening to my modded drivers, they definitely work better with the modded dust cap, rather than no dust cap at all and a phase plug, but this opens up another possible problem.
The reflection off the back of the dust cap most likely bounces of the pole piece and muddies the highs a little as it comes back through the vent holes.
I imagine the answer is to pad the top of the pole piece but this is not possible as getting the padding through the vent holes and into position would be well nigh impossible.
So my idea is cut the cap out, then pad the pole and make a new cap and araldite it in. I think the cap could be made from aluminium shim, which is rubbed into shape in the neck of a bottle (I have done this before so I know that part works), then have the vents cut into it.
The little support legs on the cap could be padded so only the tiny centre dome actually produces sound.
If your not sure what I mean with the cap look back at the pics I psoted at the start of this thread.
My measuring kit isn't ready to go... I suppose driver measuring software would be a different thread, but perhaps I could be directed to the virtually no cost solutions?
I find myself agreeing that it would be nice to see a bunch of measurements regarding these ideas. But if nobody has performed these measurements I would still like to hear people's mod ideas. Before "general wisdom" can be systematically evaluated, it must be collected...
Zero One: It seems like cutting holes in the dustcap would produce rear wave cancellation from the parts that remain, also it would vent whatever volume behind the spider and produce some sort of output out of phase with regular cone motion. You mentioned initially doing these dustcap mods because "I feel the hard dust cap also interacts with the whizzer to produce phase problems and harsh treble in some cases." As an alternative to these rather in-depth dustcap mods, have you tried felting over the dustcap? Maybe something like the $.98 tweak inside the whizzer? I suppose this might add too much mass, but it might leave you with fewer secondary effects to correct. Just thinking about alternative solutions...
I find myself agreeing that it would be nice to see a bunch of measurements regarding these ideas. But if nobody has performed these measurements I would still like to hear people's mod ideas. Before "general wisdom" can be systematically evaluated, it must be collected...
Zero One: It seems like cutting holes in the dustcap would produce rear wave cancellation from the parts that remain, also it would vent whatever volume behind the spider and produce some sort of output out of phase with regular cone motion. You mentioned initially doing these dustcap mods because "I feel the hard dust cap also interacts with the whizzer to produce phase problems and harsh treble in some cases." As an alternative to these rather in-depth dustcap mods, have you tried felting over the dustcap? Maybe something like the $.98 tweak inside the whizzer? I suppose this might add too much mass, but it might leave you with fewer secondary effects to correct. Just thinking about alternative solutions...
AdamThorne said:
I suggest you download ARTA.
The [free] demo version is fully functional, only lacking the possibility to save compleate measurement files to disk.
Using a PC with built-in soundcard and a microphone you can start practicing today!
Then you need a decent microphone.
The virtually no cost solution is a Panasonic WM-61A electret capsule at $2 a piece. The mic needs a battery and a couple of resistors for power feeding, as described here: Powering microphones
There are a few threads on the subject in the Loudspeaker section.
SveinB
Does any of this software measure or determine what effect a bass note will have on a cord played on a 12 string guitar?
The software I have used only produces a single frequency, with the option of different waveforms. Most of the frequencies it does produce are not musically related. It will do 41, but not 82. And that is what it says it is. How do you adjust or check the oscillators in the PC, which are the reference? They are all x-tal locked, and drift like crazy.
Is there any that will do a 2 or more tone test to check that a speaker can faithfully reproduce a note high in the register while reproducing a note down low keeping all the harmonics at the right levels.
Ears that have listened to a lot of music can, especially if they have been trained on live music.
Apart from that, try using square waves and a very good mike.
As for using a $2 mic, you still have to have it calibrated against a reference sound source, and know it's slew capabilities.
I say good onya Zero One. Us Ausies do use fencing wire to achieve all manner of things. It's your ears you have to keep happy. If some one doubts your success, it's their problem.
Geoff.
The software I have used only produces a single frequency, with the option of different waveforms. Most of the frequencies it does produce are not musically related. It will do 41, but not 82. And that is what it says it is. How do you adjust or check the oscillators in the PC, which are the reference? They are all x-tal locked, and drift like crazy.
Is there any that will do a 2 or more tone test to check that a speaker can faithfully reproduce a note high in the register while reproducing a note down low keeping all the harmonics at the right levels.
Ears that have listened to a lot of music can, especially if they have been trained on live music.
Apart from that, try using square waves and a very good mike.
As for using a $2 mic, you still have to have it calibrated against a reference sound source, and know it's slew capabilities.
I say good onya Zero One. Us Ausies do use fencing wire to achieve all manner of things. It's your ears you have to keep happy. If some one doubts your success, it's their problem.
Geoff.
Hi Adam
(It seems like cutting holes in the dustcap would produce rear wave cancellation from the parts that remain, also it would vent whatever volume behind the spider and produce some sort of output out of phase with regular cone motion. You mentioned initially doing these dustcap mods because "I feel the hard dust cap also interacts with the whizzer to produce phase problems and harsh treble in some cases." As an alternative to these rather in-depth dustcap mods, have you tried felting over the dustcap? Maybe something like the $.98 tweak inside the whizzer? I suppose this might add too much mass, but it might leave you with fewer secondary effects to correct. Just thinking about alternative solutions.)
Thanks for your suggestion here, I have tried this early on but found with these particular drivers it killed the highs stone dead, removing the cap and phasing plugging was a little better but still not ideal.
It seems these drivers need a cap, but the mods I have done seem to work really well though it is a little painful to implement.
I have also tried the felt on some other wide range and mid/bass drivers and actually got good results, cleaning the sound up nicely so it certainly works in some cases as far as I can tell.
I suspect the type of mods you can get away with has a lot to do with the strength of the magnet, it appears from what I have read elsewhere that the stronger the magnet the more forward the mids and highs and the more recessed the bass (generally), such speakers would probably benefit from a little extra weight and padding. No doubt it would impact on efficiency though and high end extension but it might serve to balance the system better.
These drivers have fairly modest magnets so that may account for the impact that any extra weight has on the sound. They also have light paper formers which I think are probably reinforced somewhat by the dust cap.
Overall the differences in sound by using the mods to the dustcap are pretty substantial so I feel it is definitely an area worth further investigation as it is unlikely I have yet hit on the ideal solution.
No doubt with input from many on this great forum we can find great solutions to many issues.
Thanks Adam and all the best.
Zero One.
(It seems like cutting holes in the dustcap would produce rear wave cancellation from the parts that remain, also it would vent whatever volume behind the spider and produce some sort of output out of phase with regular cone motion. You mentioned initially doing these dustcap mods because "I feel the hard dust cap also interacts with the whizzer to produce phase problems and harsh treble in some cases." As an alternative to these rather in-depth dustcap mods, have you tried felting over the dustcap? Maybe something like the $.98 tweak inside the whizzer? I suppose this might add too much mass, but it might leave you with fewer secondary effects to correct. Just thinking about alternative solutions.)
Thanks for your suggestion here, I have tried this early on but found with these particular drivers it killed the highs stone dead, removing the cap and phasing plugging was a little better but still not ideal.
It seems these drivers need a cap, but the mods I have done seem to work really well though it is a little painful to implement.
I have also tried the felt on some other wide range and mid/bass drivers and actually got good results, cleaning the sound up nicely so it certainly works in some cases as far as I can tell.
I suspect the type of mods you can get away with has a lot to do with the strength of the magnet, it appears from what I have read elsewhere that the stronger the magnet the more forward the mids and highs and the more recessed the bass (generally), such speakers would probably benefit from a little extra weight and padding. No doubt it would impact on efficiency though and high end extension but it might serve to balance the system better.
These drivers have fairly modest magnets so that may account for the impact that any extra weight has on the sound. They also have light paper formers which I think are probably reinforced somewhat by the dust cap.
Overall the differences in sound by using the mods to the dustcap are pretty substantial so I feel it is definitely an area worth further investigation as it is unlikely I have yet hit on the ideal solution.
No doubt with input from many on this great forum we can find great solutions to many issues.
Thanks Adam and all the best.
Zero One.
"The changes to the cone total less than two % of the cone's area. This makes the modification a high skill endeavor."
Interesting. As soon as I read this, an image popped in my mind of my Velodyne aluminum cone 6" mids that I parted from the bookshelf series they came up with years ago that wasn't around very long. Looking at the rear of the cone - each one has a little 'x' marked with a pen and a nice ring-shaped glob (dime-sized) of silicone applied around the x. Looks like they pinpointed a nasty resonance mode of these metal cones and with high 'skill' added a gram or two of goo to beat it. Gosh I wish I knew how to do this for paper cones.
Interesting. As soon as I read this, an image popped in my mind of my Velodyne aluminum cone 6" mids that I parted from the bookshelf series they came up with years ago that wasn't around very long. Looking at the rear of the cone - each one has a little 'x' marked with a pen and a nice ring-shaped glob (dime-sized) of silicone applied around the x. Looks like they pinpointed a nasty resonance mode of these metal cones and with high 'skill' added a gram or two of goo to beat it. Gosh I wish I knew how to do this for paper cones.
InclinedPlane said:
Either finite element analysis & laser interferometry, or by guess and by golly.
good reading:
http://www.bwspeakers.com/downloadFile\technicalFeature\Nautilus800_white_paper.pdf
The appendices starting getting very interesting around page 40 or so.
Many of their materials are more exotic than most DIYers would likely experiment with (such as the more recent diamond tweeter domes), but the research approach should apply to all common cone materials.
It's probably fair to say the N- series incorporate some of the most interesting technical R&D and production engineering in a mainstream commercial loudspeaker. Power appetite and retail pricing structure aside, they're pretty decent sounding - and personally I find their aesthetic quite appealing.
(interesting. As soon as I read this, an image popped in my mind of my Velodyne aluminum cone 6" mids that I parted from the bookshelf series they came up with years ago that wasn't around very long. Looking at the rear of the cone - each one has a little 'x' marked with a pen and a nice ring-shaped glob (dime-sized) of silicone applied around the x. Looks like they pinpointed a nasty resonance mode of these metal cones and with high 'skill' added a gram or two of goo to beat it. Gosh I wish I knew how to do this for paper cones.)
That gets me wondering. If you take a small wooden skewer or similar and tap around mine the cones make different noises at different spots, this may help in finding the right spot if of course I knew what I was looking for.
Interstingly the coated cones (PVA glue) are more consistent acoss the entire cone, but then that probably means they now have one dominant cone resonsonance.
Anyhow I also noticed that the biggest difference occurs directly opposite the point where the braid is attached to the cone, I suppose the two attachments must have some effect on resonance.
Perhaps a spot opposite the braid and about a third in from the outer edge is a good starting point if you were to use one dab of silicone or something similar.
I can only imagine that using several dabs around the cone would raise the point, and I guess that in this case it would be best to not have them evenly spaced, so as to avoid one dominant point, but maybe I'm totally wrong. In the end it would all be a balancing act between the benefit and the additional weight.
With just one point used, possibly blue tac could be used to experiment and find the point through listening with test tones till you nail it then replacing it with silicone.
zero one
That gets me wondering. If you take a small wooden skewer or similar and tap around mine the cones make different noises at different spots, this may help in finding the right spot if of course I knew what I was looking for.
Interstingly the coated cones (PVA glue) are more consistent acoss the entire cone, but then that probably means they now have one dominant cone resonsonance.
Anyhow I also noticed that the biggest difference occurs directly opposite the point where the braid is attached to the cone, I suppose the two attachments must have some effect on resonance.
Perhaps a spot opposite the braid and about a third in from the outer edge is a good starting point if you were to use one dab of silicone or something similar.
I can only imagine that using several dabs around the cone would raise the point, and I guess that in this case it would be best to not have them evenly spaced, so as to avoid one dominant point, but maybe I'm totally wrong. In the end it would all be a balancing act between the benefit and the additional weight.
With just one point used, possibly blue tac could be used to experiment and find the point through listening with test tones till you nail it then replacing it with silicone.
zero one
That B&W paper was great. Their stuff is sweet, but cones n' domes with the crossover 'gremlins' just bore me. That's why I'm here, for better or worse.
I've only been aware of the fullrange side of the audio sickness for a short time but I think I've learned enough to understand the problem with existing 'affordable' drivers and that the proper correction of them is neither affordable nor simple. All I want is a Fostex FE206E type driver that I can plunk in a basic BR and play at 110db with no peaks and dips outside 5db and no resonances to ruin the fun. 
Maybe I'll try gluing on some shallow strips of foam along the outside of these FE166Es here and who knows.... Ack, just shoot me.
I've only been aware of the fullrange side of the audio sickness
for a short time but I think I've learned enough to understand the problem with existing 'affordable' drivers and that the proper correction of them is neither affordable nor simple. All I want is a Fostex FE206E type driver that I can plunk in a basic BR and play at 110db with no peaks and dips outside 5db and no resonances to ruin the fun. Maybe I'll try gluing on some shallow strips of foam along the outside of these FE166Es here and who knows.... Ack, just shoot me.
Hi Zero one. Timely comments about the tapping of PVAed cones. I gave a 10" bass driver 2 coats today. Tapping the cone gives more of a tick than a thud. That indicates I have achieved a truer piston action. This is to be used as a bass helper for some OBs. But I wanted to extend the upper frequency limit so it blends with the OBs, rather than a crossover.
Have you ever seen high speed photographs of cones in action? The paper literally flaps in its own wind. The PVA reduces that. The cone resonances you talk of are the ones that introduce cone breakup.
But the cone needs to dampen the high frequencies as they extend out from the apex to extend freq response.
On a curvilinear cone, I apply more PVA around the apex than the perimeter. On a ribbed cone I may use 3 coats from the apex to the first rib, two from there out to halfway to the perimeter and a single coat for the rest.
BTW, if the driver is playing when you paint it, you can hear the paint dry.
It does make sense to balance the cone opposite the braids. Good observation.
Regards,
Geoff.
Have you ever seen high speed photographs of cones in action? The paper literally flaps in its own wind. The PVA reduces that. The cone resonances you talk of are the ones that introduce cone breakup.
But the cone needs to dampen the high frequencies as they extend out from the apex to extend freq response.
On a curvilinear cone, I apply more PVA around the apex than the perimeter. On a ribbed cone I may use 3 coats from the apex to the first rib, two from there out to halfway to the perimeter and a single coat for the rest.
BTW, if the driver is playing when you paint it, you can hear the paint dry.
It does make sense to balance the cone opposite the braids. Good observation.
Regards,
Geoff.
chrisb said:
http://www.freepatentsonline.com/5848173.pdf?s_id=2f61994ea4f7a048705f677355c229b7
You may need to make sign up to view this piece. Pioneer developed these 10 years ago. They sold as the s-cr505 model, or as I bought them in the sv-605 as a 5 piece sattelite and a powerd 10. I have loved and cherished this set for years untill I got back into DIY. The 5 1/8 linear drive, piston, surroundless, speaker is great. The tweeter was terrible. pioneer claimed that the woofer would go down to 55hz. They actually went fairly low, and at a rated 165 watts, they did well. Obviously the wattage is a bit overblown, they handle a 15 true watts well, but start to get out of hand. I have been toying with the idea of putting them in a BIB for a while, with a super tweeter. They were housed in an ABS box, with a port inside a port configuration. Sound was great, but on heavier lower bass tones you could hear that there was a bit too much resistance on the causing standing wave to become audible since they were not blocked by a surround, but I never heard them during normal listening. I do not want to mod these very badly since it is hard to find another set. I did find a website that sells replacement woofer for $85. So, if anyone wants to experiment...... I still need to sit down and get the t/s paramaters off them, but have only read how to, so it may take time. All the tools are readily available since I work with electronics. Enjoy the info. It is very thourough, and interesting. Shows all the tests with wire and bobbin lengths, gaps, and double spiders.
Geoff H said:
On a thick paper woofer, if you puzzlecoat the front and back you get a constrained layer sandwich (like a poor mans infinity (ceramic-AL-ceramic) or Focal (glass-foam-glass). When i am doing up the 12s i use in budget woofers (i have a cache of Fosters -- i'll see if i can pics if anyone is interested) i'll use 3 coats on the front & 2 on the back.
I call this adding a virtual whizzer cone. As well i have used a coat of dammar first to define the whizzer.
dave
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.