Hello Christian.
I have not as yet had the pleasure of reading BZs paper .
But coil buckling could be another reason for coil and foot damage.
Especially on hard surfaces or high power levels. And definitely both.
But I am not sure about it being the cause of hf problems?
I will force myself to read another one of his papers 🙄
Steve.
I think it was pixel1 that had the problem with 1 in 4 damage.
But yes, as you can see one exciter is under more stress than the other.
But which is more likely to overheat is the question?
If three exciters are mounted on mutual + phase points and one is mounted on - phase points on the panel , all four exciters will not be working together in the low frequencies.
One or more could be having problems depending on positioning.
If on a very bendy panel ,the exciters will have to cope with the waves , and rolling each exciter coil foot as it rides the waves.
Even if the exciters are rigidly mounted the coils will be forced to roll with the panel waved.
The black thing, is the capacitor that came with the exciter.
I think this question was also asked before.
The green wires bypass the capacitors, the red wires are not connected.
Steve.
So, I must ask >
Do you have a solution or recommendation to address these problems,
especially concerning twin exciter placement on a light 600mm X 600mm panel ?
( remembering that I will be removing frequencies below 350Hz )
Mister audio.
What is the light material you are using for your panels ?
XO at 350hz will remove many of the problems in the low end and reduce overheating.
I use the NXT position for my single exciter, but have not used the other three recommended positions for multiple exciters.
I have no need or desire for more exciters on my panels in my room , and I only use 10watt exciters.
I use 3/7 by 4/9 position.
I have posted the other positions, but I can't get my hands on them at the moment.
Steve.
What is the light material you are using for your panels ?
XO at 350hz will remove many of the problems in the low end and reduce overheating.
I use the NXT position for my single exciter, but have not used the other three recommended positions for multiple exciters.
I have no need or desire for more exciters on my panels in my room , and I only use 10watt exciters.
I use 3/7 by 4/9 position.
I have posted the other positions, but I can't get my hands on them at the moment.
Steve.
Similar size to what I have used for all sound daily for many months now. NXT exciters.
Active panel size, they are in a frame. 463mmx x 408.
Exciters 2 on a slightly offset diagonal in respect to the width. Offset ~25mm or so on one short side and less on the other, 10mm maybe. They are on ~38mm ctrs and wired in series for 8ohm.
Spec, they are used with a woofer, xovers in that.
😉 While I find that the some level of bass range of the male voices does come out of the woofer when I stick my ear by it disturbing it's not noticeable in use. Probably phase matched to the panels. 160Hz is a level that relates to what some think is ok for stereo use. Also if main speakers handled to say 40Hz it doesn't leave much room for a sensible woofer.
The panels can be used on their own as rears for surround sound. Maybe why they have protection devices in series with the exciters - tiny, their resistance increases if they get hot. An amp might send too much bass to them.
Sound. I notice what all do with DML plus a lack of colouration as I call it. Different brands of speakers tend to have their own sound. I like accuracy. I bought these a fair old time ago. The people who sell stuff get to hear all sorts. General opinion of these - they pull well over their price point. A lot over. Mind you bought from a store that used to well undercut all at times. 😉 Those days have gone,
They are nothing like the panels discussed here. Back enclosed with a somewhat stiffened thin vacuum formed cover. There is also a foam resting on the back of the panel. Oval cut out to clear the exciters. The foam isn't thin. Not enough measurements done to say more. 😉 I'm using them. The panels are by no means flimsy yet efficiency is pretty good. Power handling - a very tricky area.Sine test a speaker at stated max wattage levels and probably say bye bye to it.
I've got a bit wrapped up at the moment with trying to design a 2way with the same levels of performance in a woofer like sized box as this set up uses / how big does it need to be. Not done this before so a learning curve especially in the xover area. Also with the software that is needed.
Just a rough calculation : 2x500W into 8 drivers = 125W each. With a 4Ohm impedance it is 5.6A. With a force factor 3 to 4NA (or Tm) it is something like a 20N force each exciter applies so 20N to be transmitted by the voice coil former. It is the force applied from a 2liter bottle attached to the panel by its thin neck.
More advanced people in material resistance could push this calculation a bit further up to evaluating the stress in the voice coil former...
@Leob @Andre Bellwood as you are dismounting exciters, might be interesting to share the voice coil former thickness.
Christian
Got another reply from soundimports:
They have not addressed my question regarding warranty yet, but it seems like it should be quite clear cut when the item has 5 year warranty, but yet they say that the design "can create problems over extended periods of use".
Last edited:
Interesting! At least 125w peak should be within the rating, which is 40w RMS. Typically peak refers to 4x RMS, so that would mean 160w peak.
The panel material I intend to use is called (I believe) EPS > extruded polystyrene 10mm.
It has a very thin layer of aluminum bonded to both sides. Its manufactured use is insulation.
I have a thought that some of the problems people are having are caused by (always rectangular)
panels, 'hanging low' with no bottom restraint, causing a flapping/flex that stresses the exciter at mounting point.
My panels will have flexible attachment at top and bottom.
Last edited:
Mister audio.
That is XPS and will be heavily damped by the Ali foil on each side.
So the efficiency will probably be lower.
I prefer not to have a square panel , but it seems to work with some materials.
You are XO at 350hz so wobble, hopefully should not be too much of a problem.
Steve
That is XPS and will be heavily damped by the Ali foil on each side.
So the efficiency will probably be lower.
I prefer not to have a square panel , but it seems to work with some materials.
You are XO at 350hz so wobble, hopefully should not be too much of a problem.
Steve
And that explains fully why the drivers voice coils were not even slightly discoloured eventhough I thought I was over-driving them. It also suggests that the 40WRMS rating is accurate and conservative .
I became interested in speaker power ratings for various reason. I came across this concerning IEC tests
https://www.scan-speak.dk/datasheet/tech/Scan-Speak_Technote02_Powertest.pdf
With a cone speaker we get an idea of efficiency due to SPL info at 1 or 2w. On an exciter SPL is stated as not available but a plot is given by Dayton when it's on a 12" square piece of foam. I don't think this gives a clear idea of efficiency once they are mounted on some other form of panel. Also the pink noise profile used in the IEC method is filtered.
https://www.scan-speak.dk/datasheet/tech/Scan-Speak_Technote02_Powertest.pdf
With a cone speaker we get an idea of efficiency due to SPL info at 1 or 2w. On an exciter SPL is stated as not available but a plot is given by Dayton when it's on a 12" square piece of foam. I don't think this gives a clear idea of efficiency once they are mounted on some other form of panel. Also the pink noise profile used in the IEC method is filtered.
The problem is that efficiency is directly related to the type of panel material. Lower densities will give higher efficiency than high density panels, but density also effects frequency response and not just efficiency. In fact, almost everything about the panel—density, thickness and edge terrmination, effects frequency response and efficiency.
And if you roll off the bass end, then power handling also increases (that's something I hadn't thought of! My PA panels are rolled off at about 150Hz, which also explains why the voices were not burnt, and suggests that driver power handling could possibly increase due to LF being blocked). Rolling off the bass end has nothing to do with VC excursion, but is simply a natural function of signal magnitude at low frequencies.
As an aside...
If the driver is not braced or supported, then the natural action of the magnet at mechanical resonance should provide better VC airflow, and therefore better cooling and power handling than for the same signal level but with a braced driver.
I was wondering if people with problems should look at what driver wattage they are actually using. Sound levels may not relate as easily than they do with cone speakers etc. I did manage to find details of the IEC filter but no details on how it's used - sections grounded as needed maybe.
The other aspect is just how speakers or exciters are tested. Certain IEC aren't used by all.
I should add that the filter may be out of date and graphical image showing the response may show a range of what they accept
Here's the frequency response of that filter.
I'm not sure what a 'very low impedance' is in terms of line levels (something like a line or headphone output stage?) Low impedance between stages generally assumes 600 ohms. But VERY low impedance? Could be a power amplifier output stage.
The circuit I've simulated above is driven by an ideal source, (not shown) probably 0.1ohms, and feeds into a 10k load (not shown.)
I can only guess. I'd say low means in respect to the input impedance of the filter. The other end high to avoid interfering with it's operation. Bit like the input impedance of an amp, same in respect to the output impedance of a pre amp.
Use though. The impression I have is that various filter shapes are needed according to the speakers intended F range, low, mid and hi. The Seas pdf gives a clue. The filter has an in and an out. The only way I see the circuit being used to achieve that is switch the triangles to ground as needed. Otherwise leave floating. A noddy look at the first RC network gives -3dB at 21Hz.
Measure SPL on a DML panel though? On a cone it has a sort point source for the sound. Artificial but used in the phase area and this allows mic's to be used mm from the dust cap to get around LF room reflections etc. A DML radiates from all over it's surface. You theoretically need to capture the lot some how or the other. Maybe a single pure tone such as 5 - 600Hz might work out while also comparing what loudness you would expect to hear from the level that is read. Then measure watts in.
Use though. The impression I have is that various filter shapes are needed according to the speakers intended F range, low, mid and hi. The Seas pdf gives a clue. The filter has an in and an out. The only way I see the circuit being used to achieve that is switch the triangles to ground as needed. Otherwise leave floating. A noddy look at the first RC network gives -3dB at 21Hz.
Measure SPL on a DML panel though? On a cone it has a sort point source for the sound. Artificial but used in the phase area and this allows mic's to be used mm from the dust cap to get around LF room reflections etc. A DML radiates from all over it's surface. You theoretically need to capture the lot some how or the other. Maybe a single pure tone such as 5 - 600Hz might work out while also comparing what loudness you would expect to hear from the level that is read. Then measure watts in.
As I think most are aware, there are numerous IEC standards and filters.
Like the filters used for "A weighting" and "B weighting" for S/N ratio measurements.
It is because the above filter is passive and has a very specific 'curve' that the input
needs to be driven by very low impedance and the output needs to see much higher impedance.
If these 'rules' weren't followed, the 'curve' would change. If it was an active filter these 'rules' wouldn't apply.
It has always amazed me that IEC provide pure passive circuits for their filters. Very clear and unambiguous.
In a more 'real world' example, the source drive could be anything below a few ohms and the output driving anything above 2Kohms.
These various IEC standards are mainly used to 'generalize' measurements > like finding a median value.
Like the filters used for "A weighting" and "B weighting" for S/N ratio measurements.
It is because the above filter is passive and has a very specific 'curve' that the input
needs to be driven by very low impedance and the output needs to see much higher impedance.
If these 'rules' weren't followed, the 'curve' would change. If it was an active filter these 'rules' wouldn't apply.
It has always amazed me that IEC provide pure passive circuits for their filters. Very clear and unambiguous.
This is the exactly perfect way to measure the filter.
In a more 'real world' example, the source drive could be anything below a few ohms and the output driving anything above 2Kohms.
These various IEC standards are mainly used to 'generalize' measurements > like finding a median value.
I'm sort of suggesting measure at some theoretical volume you aught to be able to reach and measure wattage in sine style at a suitable frequency.
Using the ear as well. I'm curious as to why a frequency sweep plot of a DML may look pretty bad but sound fine. Suggests there is something wrong with the usual measuring technique when used on DML's. However a sine wave doesn't relate well to music. Pink noise is closer. This seems to be the reason for using it. It can contain very low and high frequencies so best restricted to a range that suits what it drives or the other F's will just cause confusion. Same if white was used.
Using the ear as well. I'm curious as to why a frequency sweep plot of a DML may look pretty bad but sound fine. Suggests there is something wrong with the usual measuring technique when used on DML's. However a sine wave doesn't relate well to music. Pink noise is closer. This seems to be the reason for using it. It can contain very low and high frequencies so best restricted to a range that suits what it drives or the other F's will just cause confusion. Same if white was used.
When I was experimenting with bracing, the holes and documentation made so little sense I emailed parts express about it. The tech explained that the single threaded hole in the back was to accommodate a particular vendor application and wouldn't expect it to generally be useful. Denied any knowledge of voice coils warping due to weight, or of any need to brace for sound improvement.
Last edited:
Happy Holidays everyone. Picked up a Christmas present for myself at Home Depot. Over the years I've had to have bought at least 10 of these for various projects. If some of the main contributors here think 3-ply birch is where it's at, then I've got to see for myself.
A nice of piece as I've ever found at HD.
You can see the three layers with the side view. This was the only board available where you could tell it had 3 layers.
I suppose my plan is to cut it in half, but, before I do anything, curious what the birch-ply veterans have to say. Cut in half? buy a second? are these better with a frame?
A nice of piece as I've ever found at HD.
You can see the three layers with the side view. This was the only board available where you could tell it had 3 layers.
I suppose my plan is to cut it in half, but, before I do anything, curious what the birch-ply veterans have to say. Cut in half? buy a second? are these better with a frame?