Oh, I think you will find it is corrugated
Not much, but looking closely there is a small pattern
At least it looked so to me
On the latest one it looked like company logo
btw, RAAL is a pure ribbon, but SA is/was a planar
Fore some reason the foil on film types are planars and doesnt seem to be the best around
Maybe because it demands fore a special magnet design with poleplate in front of diaphragm
But they do take a lot more power
Last edited:
You are quite correct.
I assumed the surface geometry was an attempt to increase the surface area (ie heat transfer ability) of the membrane. However, it surely has an effect on the eigenmodes of the membrane.
Last edited:
With regards to the duct in which the ribbon is located.
If we assume a purely rectangular duct geometry, the ribbon will initially launch a plane wave, which will transform into a spherical wave with respect to time. As the wave propagates down the rectangular duct, an abrupt transformation of the wavefront will occur once it reaches the end of the duct. diffraction and Higher Order Modes (reactive forces) may result from this I believe. This is what is causing me trouble with regards to optimization.
For flat pistons, they are affected by the global pressure over the surface. However, ribbons, being elastic membranes, are affected by local pressure over the surface. I haven't done any simulations, but the reactive nature of the aperture may have an effect on the ability of the ribbon to reproduce an impulse.
If a "waveguide" with a small roundover was used, it should provide a linear rate of expansion and can provide a vector that is orthogonal to the initial vector, however this seems like an obvious solution and may not offer a significant benefit.
If the width of the exit aperture of the duct/waveguide is restricted to 1/2 wavelength of the highest reproduced frequency (8.5mm for 20khz), pressure may be uniform with respect to the width of the aperture. If the depth of the duct/waveguide is restricted to 1/4 wavelength (4.25mm for 20khz), the reflections at the exit aperture cannot combine in phase to produce a resonance. I believe this would result in an acoustic impedance that would be primarily resistive. The optimal aperture should minimize reactance and allow resistance to approach a constant value. Could generalizations be made in this case? Would the shape of the waveguide be significant with regards to the response of the membrane if the following restrictions are assumed?
If we assume a purely rectangular duct geometry, the ribbon will initially launch a plane wave, which will transform into a spherical wave with respect to time. As the wave propagates down the rectangular duct, an abrupt transformation of the wavefront will occur once it reaches the end of the duct. diffraction and Higher Order Modes (reactive forces) may result from this I believe. This is what is causing me trouble with regards to optimization.
For flat pistons, they are affected by the global pressure over the surface. However, ribbons, being elastic membranes, are affected by local pressure over the surface. I haven't done any simulations, but the reactive nature of the aperture may have an effect on the ability of the ribbon to reproduce an impulse.
If a "waveguide" with a small roundover was used, it should provide a linear rate of expansion and can provide a vector that is orthogonal to the initial vector, however this seems like an obvious solution and may not offer a significant benefit.
If the width of the exit aperture of the duct/waveguide is restricted to 1/2 wavelength of the highest reproduced frequency (8.5mm for 20khz), pressure may be uniform with respect to the width of the aperture. If the depth of the duct/waveguide is restricted to 1/4 wavelength (4.25mm for 20khz), the reflections at the exit aperture cannot combine in phase to produce a resonance. I believe this would result in an acoustic impedance that would be primarily resistive. The optimal aperture should minimize reactance and allow resistance to approach a constant value. Could generalizations be made in this case? Would the shape of the waveguide be significant with regards to the response of the membrane if the following restrictions are assumed?
Last edited:
Typical ribbon tweeters are around 0R1, better have low, low, low DC off-set.
I looked at the problem about 12 years ago and decided not to do the project (for a ribbon manufacturer).
The transformer is usually a piece of thin-wall tubing (looks like 10ga wire), one, maybe two turns around the core for the secondary.
Takes about 20A peak to get 20W RMS into 0R1.
I looked at the problem about 12 years ago and decided not to do the project (for a ribbon manufacturer).
The transformer is usually a piece of thin-wall tubing (looks like 10ga wire), one, maybe two turns around the core for the secondary.
Takes about 20A peak to get 20W RMS into 0R1.
Last edited:
my suggestion was to use a flat diaphram ( full range ) and only corrugate the ends to be tensioned .
tweeters seem to benefit from a corrugated diaphragm ...
Here an example of a circuit idea from me: a CSPP (cross shunt push pull respective circlotron) power follower in current control mode. There are two disadvantages:
1) 2500 watts loss power for 40 watts output power at 0,1 ohms load (pure class A up to approximately 20 watt output power)
2) voltage gain only 0,13 times instead 1 times (unity gain) - therefore the needed voltage gain stage of the front end must have 7,5 times more voltage gain than by normal voltage controlled output power buffer (I would choice arround 60 - 100 times)
If you have the appropriate heatsinks, proper power supply and high quality double cascode voltage gain stage, you must get actually ultimate sound quality
Satisfactory sonic quality even at normal class A / B topology? I can't beleave this.
Here the circuit and the simulation results (please note by THD interpretation: 0db = 1Vss and +6db = 2Vss)
I think if you tension for a resonance out of band you'll be at the limit of the membrane in many ways. I think you'll wind up with piles of ripped foil. Just the means of stabilizing/calibrating the force, assuming it can be held, might drive you crazy.
That looks fantastic!
How much would the components of such an amplifier cost?
Perhaps water cooling could be used as a potential solution to deal with the enormous amounts of heat produced by the amplifier.
Any thoughts?
A water loop may be the only way to go since the amount of air required would be so high fan noise would kill any advantage you had in the sonic department unless you could plant the radiator remote or make it large enough for natural convection. I've made a passive radiator out of PVC pipe and brass nipples for a class A amp that worked surprisingly well. But finally electric is not an economical way to heat the home, so even in winter that amount of loss isn't practical. The clothes dryer plan is of course a joke even if the sink temperature could be allowed to rise high enough for that to work.
I haven't realize such a project, but the main costs of material causes the transformer (for each channel 2x54V~/60A), the capacitors (at whole 4x1F/100V) and the heatsinks, if you don't like the diy of this one.
That looks fantastic!
How much would the components of such an amplifier cost?
If I have more exactly data of your ribbon speakers (wanted frequency range, efficiency and impedance) I can guess the nessecary costs
With regards to power handling,
A heat transfer analysis would be appropriate for establishing thermal limits. I recently spoke with one of the professors in the physics department, he shared the opinion that most of the heat would migrate through the clamped contact interface (conduction), rather than through the air.
Andrew has proposed forced convection as a potential solution, however I believe that may contribute excessive noise.
I can't think of many options for further optimizing the transducers thermal limits, without negatively effecting other aspects of the design. The obvious solution of maximizing the area of the contact interface + thermal mass of the structure does not appear to have many disadvantages. Have members experimented with high thermal conductivity epoxies as a bonding element at the contact interface?
Last edited:
Corrugating does appear to be a interesting solution. However, the forces acting on the membrane from the expansion and contraction of the corrugations may induce non-linearities.
If the reactive force of the membrane was minimized, the transducer's ability to track a waveform would improve. However, minimizing the stiffness of the membrane may induce an oscillating bias of the membrane position related to room mode coupling. It's response may become chaotic with excessive LF content and possibly result in transducer failure due to excessive excursion.
I have seen designs which attempted to solve this problem by segmenting the membrane, however this solution appears puzzling to me. The problem is not singularly the result of membrane length, but rather membrane compliance. Segmenting the membrane simply lowers the total compliance.
If the membrane is segmented (decreased compliance), why would they corrugate the individual membranes (increased compliance) within the segmented line? Why not determine the optimal compliance and engineer a full length membrane that meets that target?
A single membrane would appear to be superior to multiple membranes with regards to energy distribution wrt the height of the line.
Last edited:
Aluminum 7475-T7351 looks interesting.
Fatigue Strength = 220 MPa
Electrical Resistivity = .00000416 ohm-cm
Thermal Conductivity = 163 W/m-K
MatWeb - The Online Materials Information Resource
Are there any other alloys members may consider suitable for such an application or would like to see tested?
Any thoughts?
Fatigue Strength = 220 MPa
Electrical Resistivity = .00000416 ohm-cm
Thermal Conductivity = 163 W/m-K
MatWeb - The Online Materials Information Resource
Are there any other alloys members may consider suitable for such an application or would like to see tested?
Any thoughts?
If there are 8 ohms load impedance, that would mean 84db/1W/m But in your case that means 84 db/8W/m (resp. 5.3W/m by load of 1.5 ohms).
Normal speakers delivers between 90 db (medium efficiency) and 100 db (high efficiency) at 1W/m. You need therefore approx. 32W for 90 db, 128W for 96 db and 512W for 102 db. To avoid clipping effects during peak levels in dynamic aeras (drum solo e. g.) my rule is at least 10 times more power as avarage.
This means for your 84 db full range ribbon transducer arround 5000 watt undistorted peak output power.
My additional rule for ribbon transducers is pur class A, because cross over distortion must be avoid in all cases.
Before I continue the calculation, I must ask follow:
How many family homes you would like to provide this winter with heat?