I now you said it wouldn't be easy. I thought from an earlier conversation about the current drive that a higher impedance was preferable to the standard low Dcr impedance of a speaker set up for voltage drive. I though you wanted as flat an impedance curve as possible to go with that higher impedance. With the current drive I am thinking a closed box would be better than a reflex box. In my case getting a lot of wire in the gap is easy, the gap is 4X the coil length.
Circlomanen,
Something just doesn't sound right about your description of the device. Just because you have many magnets distributed about the outer diameter doesn't mean it will have a high magnetic energy. Without a flux concentration method the magnets are only as strong as the amount of surface area you have if you don't have some kind of return circuit besides using air as the return path. And having a couple of very large cross sectional wire seems weird to me. I would think without many windings your field coil would put out very little field strength, current be damned in that situation
Something just doesn't sound right about your description of the device. Just because you have many magnets distributed about the outer diameter doesn't mean it will have a high magnetic energy. Without a flux concentration method the magnets are only as strong as the amount of surface area you have if you don't have some kind of return circuit besides using air as the return path. And having a couple of very large cross sectional wire seems weird to me. I would think without many windings your field coil would put out very little field strength, current be damned in that situation
150 turns of wire around a 12 mm pvc pipe could shoot a small Nd magnet several meters by just lightly tapping the wire leads to a 9 volt battery. More turns and a longer coil and a 12 volt car battery with low internal resistance should be able to shoot the magnet much longer.
The very thick solid strip cu-coil was partly meant as a joke playing with the annual copper production of Bolivia. I did not mean that litteraly. The basic principle of using a very low DCR fixed voicecoil is correct though.
The reason this is not used commersially is probably the high cost of 10 - 30 kg copper wire.
To expensive to manufacture commersially but not for the diy crowd.
The very thick solid strip cu-coil was partly meant as a joke playing with the annual copper production of Bolivia. I did not mean that litteraly. The basic principle of using a very low DCR fixed voicecoil is correct though.
The reason this is not used commersially is probably the high cost of 10 - 30 kg copper wire.
To expensive to manufacture commersially but not for the diy crowd.
It's not Current or Voltage that Amps have a problem with .. its Power.
Any amp with a given PSU and number of output devices can only deliver so many Volts and so many Amps .. with some relation for values between the 2 maximums. This doesn't change if it is a Voltage or Current source.
A higher impedance means you need more Volts but less Current. But if you take an amp and convert it to a Current Source, it will still be limited by its PSU & output devices. ie it will make the same amount of noise though final THD may be less with Current drive.
What Don and I propose is that you don't think of 8R or 4R speakers .. but what you need to drive at the frequencies of interest.
If you are designing a drive unit, I would ask what amp you are going to use and optimise the unit/box for it.
I think Don explains it well in his paper. He recommends an amp with much higher Voltage capability to take advantage of the Impedance peak. This holds whether the amp is a Current or Voltage source.
Yes. For current drive, a closed box is much easier.
___________________
The "flat an impedance curve as possible" is a red herring. Merilainen et al spout it cos it is difficult to get frequency response flat with current drive without it. Adding extra bits to flatten the impedance curve takes away a lot of the Current Drive goodness.
But you are using sophisticated Keele/Lee EQ so you can keep the current drive to the unit as pure as God intended 🙂
___________________
Circlomanen, the Eqns for Moving Coil are the same for Moving Magnet speakers. The important factor is STILL Bl^2 / Rdc.
The last time I did detailed calculations, it was still better (for a speaker) to use Moving Coil .. even with the latest NeFeB mixes.
Wharfedale even have a patent for something like what you describe but it was still a toy 😀
But technology is changing fast. eg NeFeB has abysmal temperature characteristics but some new expensive mixes are much better.
___________________
Current amps may not mind driving a SC with no signal but they DO when giving full output. 😱 Still limited by PSU & output devices.
And its not Current or Voltage that heats coils up. It's Power. To make the same amount of noise, you have to put in the same power.
The opposite of Thermal Runaway in a speaker voice coil is Compression .. things get quieter when the speaker heats up. The lack of compression is one of the big advantages of Current Drive .. especially at LF. The David Birt IOA paper I mentioned, discusses how you deal with thermal runaway.
Last edited:
The best way to decrease the risk of hot voicecoils and thermal runaway is to increase Bl and efficiency. A higher efficiency means more of the applied power becomes noise or sweet music and less power becomes heat. Add to that you also need less power for any given spl. It is a win win situation, exept for those who wants to make a living selling cheap Bl-deficient drivers.
Currentdrive is very easily used with closed box, open baffles and BIBs (large konical taper back loaded horns).
Currentdrive is very easily used with closed box, open baffles and BIBs (large konical taper back loaded horns).
Yes to all that.
Got any impedance curves of BIBs suitable for current drive? Or just links to these?
If such a risk occurs, it means that one is demanding too much to the driver.
I look for data in real domestic hifi conditions where the voice coil temperature
degrades the quality. It it happens, the cause is of badly dimensioned devices regarding the required SPL output.
http://youtu.be/WkZJfTlsvfE
A crude low quality video of my BIB. The microphone was in a bad condition since i accidently dropped the camera into a lake.
My BIB was powered by a simple no negative global feedback inverted Triadtron. It has around 12 to 15 ohm out-z. One watt output. A simple14,7 volt powersupply. Two IRF9530 fets. 0.6 amp current runing through the fets.
If you read the monumental thread about the BIB in the fullrange forum of diuaudio, you can see several people praising the BIB with high Q drivers. I still want a high Bl for the efficiency and transient respons. My B&C 10ps26 has a very low Qms which makes them suitable for current drive. A BIB is a backloaded horn without a back chamber behind the driver. It has a wide bandwidth hornloading of the driver which helps linearise the respons from a the driver when driven from a high impedance source.
Cheers, Johannes
A crude low quality video of my BIB. The microphone was in a bad condition since i accidently dropped the camera into a lake.
My BIB was powered by a simple no negative global feedback inverted Triadtron. It has around 12 to 15 ohm out-z. One watt output. A simple14,7 volt powersupply. Two IRF9530 fets. 0.6 amp current runing through the fets.
If you read the monumental thread about the BIB in the fullrange forum of diuaudio, you can see several people praising the BIB with high Q drivers. I still want a high Bl for the efficiency and transient respons. My B&C 10ps26 has a very low Qms which makes them suitable for current drive. A BIB is a backloaded horn without a back chamber behind the driver. It has a wide bandwidth hornloading of the driver which helps linearise the respons from a the driver when driven from a high impedance source.
Cheers, Johannes
Which BiB thread is this?
I've looked at several hundred posts in diyaudio.
is probably the most useful.
Have you got Impedance curves for your BiB?
I've considerable experience with Paralines with the original Elac, Wharfedale Super8RSDD and Lowther PM6 from the 70's. The Lowther was the best. Modern Lowther units aren't as good.
The most sophisticated Voigt cabinets were designed by Roger Loughton at Castle Acoustics. Roger used to be one of my engineers.
Those who like Voigts with high Q units must like loadsa boom boom bass 🙂
I'm not sure Current Drive is good for BiBs (which I gather is a species of Voigt) They need a LOT of electrodynamic damping if they are not to boom.
http://www.diyaudio.com/forums/full...anyone-have-those-fostex-craft-handbooks.html
This is the very large main thread about the BIB.
Almost any deviation from the clean conical horn will create problems with current drive. It is the wide bandwidth (both below and above the Fs)of the conical horn that keeps the cone in control when current driving a driver in the BIB.
My BIBs does not sound boomy when driven from a z-out of 470 ohm.
It is important to calculate the BIB with current drive in mind.
I will post some simulations when my main computer is running again.
Cheers,
Johannes
This is the very large main thread about the BIB.
Almost any deviation from the clean conical horn will create problems with current drive. It is the wide bandwidth (both below and above the Fs)of the conical horn that keeps the cone in control when current driving a driver in the BIB.
My BIBs does not sound boomy when driven from a z-out of 470 ohm.
It is important to calculate the BIB with current drive in mind.
I will post some simulations when my main computer is running again.
Cheers,
Johannes
All, the trouble a troubled mind can get into! I may never build a DIY speaker, but I am tempted to build a toy railgun like was just described. I have a 12 volt car battery just begging for the next mad scientist experiment!
Use a 10 ohm power resistor between the battery and a large bank of capacitors, and a switch between the capacitors and the coil in the railgun. This way you don't risk destroying the battery. Remember to use very thick wire between the capacitors and the coil, for maximum performance. There is no such thing as to large capacitor, coil or wire...
Cheers,
Johannes
I am sorry, but no actual measurements. The best i can do is Hornresp simulations.
This is my BIB as it is built (within a couple of centimeters).
The last section of the horn (L34) is the the corner-expansion outside the actual horn, crudely measured/guesstimated with a with a yardstick. It is not very exact but it is a better representation of the actual horn path and expansion then if omitted. It is much nearer actual measurements of the spl-response, then only the actual horn simulated without the corner expansion is.
Grey line is with normal voltage drive. Black line is with 15 ohm series-resistance. If designed for currentdrive i would have built a much larger BIB, simulated with 15 ohms or higher Rg.
My BIB sound best with 6 - 10 ohm resistance. They lack some of the deep bass with to high resistance.
Johannes: I read Wikipedia on what is a "railgun" and your magnet cannon is slightly different. A true railgun uses two parallel rails and a conductive projectile. What was described originally would be a large coil with a magnetic projectile (no metal contact). Wouldn't this require not a blast of high amperage, but rather at least a few seconds of high current to establish a magnetic field? A 12 V auto battery "might" stand a dead short for a few seconds. I'll try ... with some protection for myself!
Be careful with Glass mat(gel cell) car batteries. They explode from a dead short. They can't even be load tested with an AVR. Normal lead acid can be shorted momentarily with no problems though.
Be very careful dead shorting even a lead acid battery. If the temperature is elevated in the least or you have just charged the battery be careful of explosive gas at the battery. I still remember to this day one of my instructors telliing us what could happen with a dead short on a lead acid battery. He did that in class and needless to say it exploded on the spot.
Yes please use a 10 ohm (or more) resistor between the battery and the capacitors. There is a lot of power in a car battery. I used a 9 volt battery, and if we shot several times in a row the battery got hot to touch. You could always start with several 9 volt batteries parallel for more current.
Moving coil vs moving magnets.
If we look outside sound reproduction we see moving magnet being much more efficient then moving coils. Trains, electric cars, large industrial fans and equipment all use moving magnet design. Large coils can generate much more power without thermal problems then small thin coils inside a narrow gap. Bl is still "Tesla times meters (of wire in the coil)" and you can not make the magnetic fields much stronger inside the gap then modern drivers like the B&C 18IPAL.
Brushless DC electric motor - Wikipedia, the free encyclopedia
My son has a brushless RC car and the acceleration is spectacular. It is so aggressive it is hard to control the car. It usually ends up upside down on the neighbours lawn.
I still believe the moving magnet loudspeaker is the path forward for more Bl, efficiency and power density in loudspeaker design. It is also easier for DIY experiments. A very large suspension along the outer rim of the cone is much more suitable for a low Qms, very linear long travel design. For true current drive we want a Qms of around 0,3 to 0,7 or so, depending on the intended application.
Cheers,
Johannes
Moving coil vs moving magnets.
If we look outside sound reproduction we see moving magnet being much more efficient then moving coils. Trains, electric cars, large industrial fans and equipment all use moving magnet design. Large coils can generate much more power without thermal problems then small thin coils inside a narrow gap. Bl is still "Tesla times meters (of wire in the coil)" and you can not make the magnetic fields much stronger inside the gap then modern drivers like the B&C 18IPAL.
Brushless DC electric motor - Wikipedia, the free encyclopedia
My son has a brushless RC car and the acceleration is spectacular. It is so aggressive it is hard to control the car. It usually ends up upside down on the neighbours lawn.
I still believe the moving magnet loudspeaker is the path forward for more Bl, efficiency and power density in loudspeaker design. It is also easier for DIY experiments. A very large suspension along the outer rim of the cone is much more suitable for a low Qms, very linear long travel design. For true current drive we want a Qms of around 0,3 to 0,7 or so, depending on the intended application.
Cheers,
Johannes
Circlomanen, we look forward to the first practical speaker that uses the principles your expound ... or even a fully worked out theoretical design. 🙂
I don't think the 2 examples you quote qualify except VERY specialise applications for the Servo Motor unit over a very limited bandwidth.
I don't think the 2 examples you quote qualify except VERY specialise applications for the Servo Motor unit over a very limited bandwidth.
I hope i can build something like this. I am very interested in current drive and there is always a problem with "normal" high Qms (above 0,8) drivers and high z-out amps.
I will not build anything remotely serious and professional, since i have no way of scientific design a driver to desired specifications. I have been thinking about this for a year now, and i need to buy Nd magnets. I will use cotton cloth and Liquisole Liquisole - Casco to make the suspension along the rim. I wish i was working as an engineer at B&C or 18Sound with access to their equipment and testing facility, but as a uneducated guy living in Sweden i don't think that will happen.
Cheers,
Johannes
I will not build anything remotely serious and professional, since i have no way of scientific design a driver to desired specifications. I have been thinking about this for a year now, and i need to buy Nd magnets. I will use cotton cloth and Liquisole Liquisole - Casco to make the suspension along the rim. I wish i was working as an engineer at B&C or 18Sound with access to their equipment and testing facility, but as a uneducated guy living in Sweden i don't think that will happen.
Cheers,
Johannes