To suspend or not to suspend... that is a question !
Burnt, I absolutely love your concept here for the speaker stands! Do you think this mainly works because your panels are tall and skinny?
If the panels were relatively wider do you think it would “dampen” the panels too much?
I’m potentially considering trying something like this now, inspired by these pics, with 4 * 2 panels versus the original plan of having the panels float freely from cheap clothing racks on wheels.
I guess you want to stick to wood from a resonance perspective, or do you think that using metal “L” brackets might also work ?
Covid-willing, my wood and exciters arrives tomorrow ! Crossing fingers and can’t wait !
Burnt, I absolutely love your concept here for the speaker stands! Do you think this mainly works because your panels are tall and skinny?
If the panels were relatively wider do you think it would “dampen” the panels too much?
I’m potentially considering trying something like this now, inspired by these pics, with 4 * 2 panels versus the original plan of having the panels float freely from cheap clothing racks on wheels.
I guess you want to stick to wood from a resonance perspective, or do you think that using metal “L” brackets might also work ?
Covid-willing, my wood and exciters arrives tomorrow ! Crossing fingers and can’t wait !
Burnt,
At least for the case of your simple tall panels, it makes sense that the beam formulas should work. Those panels, being held only at the one end, are basically cantilevered beams.
That said, the beam formulas are really just simplified versions of the plate formulas, and if you can find a plate formula reflecting the correct boundary conditions it should give very similar results to the beam equations. The trouble is that the most easily found plate formulas are for plates with simple supports, or possibly cantilevered supports all around.
Eric
Burnt,
Great to see some REW results! Can you clarify your mic position comments? When you say "left hand corner" or "right hand corner", do you mean on the left or right edge of the panel adjacent to the exciter, i.e. about a third from the bottom? Or do you actually mean the panel corners where the 300 and 1200 mm edges meet?
And which panels are these? The "simple" ones shown in your recent photos or other?
What did you set the windows at to minimize room effects?
I'd be interested to see an "on axis" measurement, at about 1 meter or similar, if you've done that.
Eric
I’m a dummy Eric... but I canz math a little !
By saying that the “beam formulae should work”, are you saying they are not exactly DMLs or does it have a different meaning ?
(BTW, thank you for all your contributions to this thread, I’ve learned a lot from your writings.)
Blimey chaps, give a bloke a chance to walk the dog! No that is not a euphemism, shame on all of you.
In no particular order.
Eric I agree the tall panels are closer to a beam so beam formula should work. I need to do more work on the panel formulae available to see if I can find a hybrid that will give answers closer to what I am measuring . Thinking about that one.
Also the REW measurements were just to answer @Spedges request for a close miked measurement over the exciter area. I included the corners for comparison as he wants to share his measurements and explore contribution from the exciter and from the panel as two separate contributions. He has some ideas about fine tuning which we will be exploring.
The corner measurements are 50mm from the panel surface and 50mm in from each edge on the bottom of the 600 by 1200 panel. I hope that is clear. Once I am more confident I have my REW settings right I am tempted to make gated measurements of the panel at several "logical" locations i.e. corners, mid points and four places along the long axis in the middle of the panel and same with the short axis
I have other measurements for the Varaize panels taken at the listening position. This location smoothes out the response significantly. I will do a set measuring at 1 M for you as soon as I am able, some time tomorrow now.
The panels are the original French ones- not the tall panels. I will be using REW to measure some tall panels as soon as I can get some built after the Lock-in ends. Could be another month here. . I am going for a 1:4 ratio, a 1:5 ratio and a 1:6 ratio. I don't expect the extreme ratio to work and expect it to mark the outlier extreme.
CovidCrazy
Glad you like the tall designs. The panels are as Eric said a clamped cantilever. The panels as stuck to the stands using industrial grade double sided tape, three strips each strip 50mm /2inches wide. They have been in use for over two months now with no sign of this crudely simple method failing, but two months is nothing really. I am afraid I can't really predict what will happen if you make them wider but the literature tells us the bigger the panels the more modes you have and thats a good thing. I strongly suspect that a 2*4 panel will work in DML mode and you will not get a 'beam' effect as my 2*4 panels don't exhibit the effect.
I have used L brackets on some test DML's I made some time back for easy materials testing. i isolated the panel from the bracket using elastomeric film like the sort of sheets you get for damping loudspeaker cabinets, it worked well.
As usual its a case of suck it and see if you go off script, but I have found the DML principle of opperation to be very tolerant of design decisions.
Burnt
P.S. Just remembered my son placed a large weight on the back of the stands when he moved them downstairs into the living room and said it improved the bass a little. He's a physicist and as sceptical as hell so I would trust that to be an accurate report.
In no particular order.
Eric I agree the tall panels are closer to a beam so beam formula should work. I need to do more work on the panel formulae available to see if I can find a hybrid that will give answers closer to what I am measuring . Thinking about that one.
Also the REW measurements were just to answer @Spedges request for a close miked measurement over the exciter area. I included the corners for comparison as he wants to share his measurements and explore contribution from the exciter and from the panel as two separate contributions. He has some ideas about fine tuning which we will be exploring.
The corner measurements are 50mm from the panel surface and 50mm in from each edge on the bottom of the 600 by 1200 panel. I hope that is clear. Once I am more confident I have my REW settings right I am tempted to make gated measurements of the panel at several "logical" locations i.e. corners, mid points and four places along the long axis in the middle of the panel and same with the short axis
I have other measurements for the Varaize panels taken at the listening position. This location smoothes out the response significantly. I will do a set measuring at 1 M for you as soon as I am able, some time tomorrow now.
The panels are the original French ones- not the tall panels. I will be using REW to measure some tall panels as soon as I can get some built after the Lock-in ends. Could be another month here. . I am going for a 1:4 ratio, a 1:5 ratio and a 1:6 ratio. I don't expect the extreme ratio to work and expect it to mark the outlier extreme.
CovidCrazy
Glad you like the tall designs. The panels are as Eric said a clamped cantilever. The panels as stuck to the stands using industrial grade double sided tape, three strips each strip 50mm /2inches wide. They have been in use for over two months now with no sign of this crudely simple method failing, but two months is nothing really. I am afraid I can't really predict what will happen if you make them wider but the literature tells us the bigger the panels the more modes you have and thats a good thing. I strongly suspect that a 2*4 panel will work in DML mode and you will not get a 'beam' effect as my 2*4 panels don't exhibit the effect.
I have used L brackets on some test DML's I made some time back for easy materials testing. i isolated the panel from the bracket using elastomeric film like the sort of sheets you get for damping loudspeaker cabinets, it worked well.
As usual its a case of suck it and see if you go off script, but I have found the DML principle of opperation to be very tolerant of design decisions.
Burnt
P.S. Just remembered my son placed a large weight on the back of the stands when he moved them downstairs into the living room and said it improved the bass a little. He's a physicist and as sceptical as hell so I would trust that to be an accurate report.
Last edited:
Attachments
Maybe this will help someone with narrow panels:US4426556A - Electrodynamic loudspeaker
- Google Patents
- Google Patents
I absolutely love this patient, thanks for the share.
Am pouring over its details. The one part that I find unsatisfying is that I can’t develop intuitions about behavior of the waves traveling through the surface that are not perpendicular to any edge.
This patent addresses waves that go along parallel to the long edge of the surface and essentially bounce right back reinforcing / cancelling themselves to form modes.
But surely there is vastly more energy traveling in a non-perpendicular fashion even in a thin plank. My understanding is that the angle of incidence always equals the angle of reflection.
So doesn’t the vast majority of the “ripple energy” radiating out from the exciter just bounce around completely randomly with no clear reinforcement? Does that render the fixation in this patent on the small proportion of energy that is in fact parallel to the long edge relatively immaterial?
(full disclosure: I am a statistician so think in predominantly statistical terms)
Thank you Burnt, we got her as a rescued dog. She was in horrible shape and badly abused...less than 40 lbs when we got her, now after 1.5 years she is at almost 80 lbs....full size the vet says.
BTW, shes only half lab retriever, the other half is Siberian husky...quite a mix and temperament
geosand
It can be useful for fans of panels wider:WO2007083127A2 - Bending wave acoustic device and method of making thereof
- Google Patents
- Google Patents
Burntcoil.
I've been very confused by the downward trend in the centre area from the 300hz to 10k point on your plots ,i started to look around to see if I could find better responses most showed similar drops after 2or 3k using the 32mm exciter,but then I looked back to page 149 and found urpgor's plot's ,which show a good response up to at least 15k? Using beech and poplar.
Even parts express plot shows minus 15db at 10k ,but that is using some kind of 1/2 inch foam core board ?
Am waiting to see the 1mrt and maybe the 3 MTR responses to see what happens there?
Have you any idea of what can be causing the discrepancy .
The 25mm exciter has a good response up to 20k so thought it might be the 32mm exciter having problems,but urpgor's plot's show otherwise?
Surely we need to get to the bottom of this problem ,what is going on?
Eq ing a 15db roll off is not a good place to start, surely eq should be a last resort.
Any answers ?
Steve
Hi Steve,
I will post a 1M and my seating position measurement (4 M) when I have completed the latest ones. ( I am learning how to use REW)
I know that the current 4M position FR I have is less 'droopy' which suggests the overall response is a result of the summed performance of the whole panel.
I would never attempt to EQ a 15dB drop. Its possible to choose a reference point that allows you to reduce some of the higher output and lift the lower output-very roughly, as a hypothetical example, a 7dB cut for the peaks with a 7dB lift for the lower response is a good compromise. Yes it reduces headroom but I find because the panels are so large I never get anywhere near overtaxing the exciters so thats not much of an issue.
I will post a 1M and my seating position measurement (4 M) when I have completed the latest ones. ( I am learning how to use REW)
I know that the current 4M position FR I have is less 'droopy' which suggests the overall response is a result of the summed performance of the whole panel.
I would never attempt to EQ a 15dB drop. Its possible to choose a reference point that allows you to reduce some of the higher output and lift the lower output-very roughly, as a hypothetical example, a 7dB cut for the peaks with a 7dB lift for the lower response is a good compromise. Yes it reduces headroom but I find because the panels are so large I never get anywhere near overtaxing the exciters so thats not much of an issue.
Hi everyone, I'm new here, but today I got my first Dayton Audio Exciter DAEX30HESF-4 and a whole bunch of materials to mess around with during the weekend. I got the last one available in Poland as because of the Covid next delivery will be in around a month or 2, but I wanted to point your attention to something that I've never seen before.
Many years ago Wharfedale produce picture speakers.
Grające obrazy WHARFEDALE para, chwilowa obniżka 8851204534 - Allegro.pl
This is on our "Polish Ebay" and they seem pretty cool. I'm wondering if anyone heard about them before?
Best
Raf
Many years ago Wharfedale produce picture speakers.
Grające obrazy WHARFEDALE para, chwilowa obniżka 8851204534 - Allegro.pl
This is on our "Polish Ebay" and they seem pretty cool. I'm wondering if anyone heard about them before?
Best
Raf
Burntcoil.
Look forward to seeing the other plots .
I don't mean to go on but you are missing the point ,according to urpgors plots you should not need to use eq at all,which is what is puzzling me.
Like urpgor I find ply panels to have a pretty good response up to 20k ,so if his exciter is putting hf into his panels so should yours .
What I would look into is what is happening inbetween the exciter and the panel ,something is blocking the signals from getting through.
Even though I glue my exciter to the panel and make sure of a solid firm contact,over time or from a sideways jolt they can come loose ,not that you would know by physically examining it,but it would show frequency response fall off on the daq .
If you are using the sticky ring that came with the exciter ,this would be my first suspect,either from not making good contact ,or not fit for purpose.
As for the response levelling out the further away from the panel,I would expect the LF to reduce below about 300 hz as you move further away into the room,but frequencies above this tend to hold firm well into the furthest corners of the room.So basically the low frequencies are coming down to the low levels of the rest of the panel.
Whatever the cause ,something is wrong and needs to be fixed,if all else fails try a 25mmm exciter.
The only other thing I can think of is a faulty mic or measurement?
As I said ,I don't mean to bang on ,but I can't look at that sort of response and think it's normal or in anyway acceptable.
Please don't take this the wrong way,I really am trying to help.
Steve
Look forward to seeing the other plots .
I don't mean to go on but you are missing the point ,according to urpgors plots you should not need to use eq at all,which is what is puzzling me.
Like urpgor I find ply panels to have a pretty good response up to 20k ,so if his exciter is putting hf into his panels so should yours .
What I would look into is what is happening inbetween the exciter and the panel ,something is blocking the signals from getting through.
Even though I glue my exciter to the panel and make sure of a solid firm contact,over time or from a sideways jolt they can come loose ,not that you would know by physically examining it,but it would show frequency response fall off on the daq .
If you are using the sticky ring that came with the exciter ,this would be my first suspect,either from not making good contact ,or not fit for purpose.
As for the response levelling out the further away from the panel,I would expect the LF to reduce below about 300 hz as you move further away into the room,but frequencies above this tend to hold firm well into the furthest corners of the room.So basically the low frequencies are coming down to the low levels of the rest of the panel.
Whatever the cause ,something is wrong and needs to be fixed,if all else fails try a 25mmm exciter.
The only other thing I can think of is a faulty mic or measurement?
As I said ,I don't mean to bang on ,but I can't look at that sort of response and think it's normal or in anyway acceptable.
Please don't take this the wrong way,I really am trying to help.
Steve
OK, got my materials, and smacked together the prototypes today !
3mm Finnish Birch Airplane Ply 2 * 4 ft each.
Cheap Dayton 100w amp DTA-Pro - works great.
4 of the Dayton 24w exciters, 2 per panel, close together, placed in the PE recommended spot.
Cheap Dayton 1000l powered sub.
It sounds great, even thought I have not yet gotten to rounding the corners or suspending the speakers. Wish I had gotten a better sub, the DMLs deserve better.
The freq response is pretty flat, but has a dip around 160hz and this sub only kicks in around 140hz so that's a bummer... on paper... but to my ear it has been a joy so far.
Can't wait to put the finishing touches on it !
3mm Finnish Birch Airplane Ply 2 * 4 ft each.
Cheap Dayton 100w amp DTA-Pro - works great.
4 of the Dayton 24w exciters, 2 per panel, close together, placed in the PE recommended spot.
Cheap Dayton 1000l powered sub.
It sounds great, even thought I have not yet gotten to rounding the corners or suspending the speakers. Wish I had gotten a better sub, the DMLs deserve better.
The freq response is pretty flat, but has a dip around 160hz and this sub only kicks in around 140hz so that's a bummer... on paper... but to my ear it has been a joy so far.
Can't wait to put the finishing touches on it !
@Spedge
Hi Steve
The exciters are glued on so I doubt its a problem with the contact area between exciters and panel.
Its most likely the material and I really don't have a problem EQing out any deviation. I know you like to get a panel that needs minimal work and I agree with the philosophy, however I don't have the opportunity to try alternatives for a few weeks yet.
It is interesting that the FR profile I get is very similar to the XPS responses I have seen on the PartsExpress thread so it may be an exciter/ panel combination.
I will post the measurements as promised but until the lock down lifts here I will not be able to try different materials. I suggest we hook up with your thoughts on fine tuning then.
Thank you for the input, I do appreciate it but until France reopens I am kinda stuck.
Meanwhile I am listening to Maria Callas singing Il Trovatore, Act IV at La Scala and its giving me chills. One hell of voice that girl!
Burnt
Hi Steve
The exciters are glued on so I doubt its a problem with the contact area between exciters and panel.
Its most likely the material and I really don't have a problem EQing out any deviation. I know you like to get a panel that needs minimal work and I agree with the philosophy, however I don't have the opportunity to try alternatives for a few weeks yet.
It is interesting that the FR profile I get is very similar to the XPS responses I have seen on the PartsExpress thread so it may be an exciter/ panel combination.
I will post the measurements as promised but until the lock down lifts here I will not be able to try different materials. I suggest we hook up with your thoughts on fine tuning then.
Thank you for the input, I do appreciate it but until France reopens I am kinda stuck.
Meanwhile I am listening to Maria Callas singing Il Trovatore, Act IV at La Scala and its giving me chills. One hell of voice that girl!
Burnt
Last edited:
Raph1978.
The wharfedale panels were a bit plasticy and used if I remember rightly a thin hardboard panel ,the sub that came with it was a bit noisy and I had trouble trying to move it away far enough to blend it in,but in all it sounded ok.
At least wharfedale sort of made a serious attempt at producing a good sounding dml speaker,but could have done better.
If space was a problem it was great .
Covid.
I see space isn't a problem for you,my panels were similar to this 3x2ft and 6x2ft but using only one exciter,I eventually clamped them solidly onto a cheap 3x2inch stud work frame to maximise reflections and increase output with good results,probably should have used a heavier more solid wood ,but it was only an experiment.
Steve
The wharfedale panels were a bit plasticy and used if I remember rightly a thin hardboard panel ,the sub that came with it was a bit noisy and I had trouble trying to move it away far enough to blend it in,but in all it sounded ok.
At least wharfedale sort of made a serious attempt at producing a good sounding dml speaker,but could have done better.
If space was a problem it was great .
Covid.
I see space isn't a problem for you,my panels were similar to this 3x2ft and 6x2ft but using only one exciter,I eventually clamped them solidly onto a cheap 3x2inch stud work frame to maximise reflections and increase output with good results,probably should have used a heavier more solid wood ,but it was only an experiment.
Steve