If i remember correctly, you can short the outputs of an aleph (maybe only the X) with no problems. A low voltage version may help there.
I havn't had a chance to go through LineSource's thread but I will definately do it when I get some spare time.
I havn't had a chance to go through LineSource's thread but I will definately do it when I get some spare time.
If we tried to run a ribbon without a transformer wouldn`t this
be a good place to use a power follower, as we might not need
any voltage gain just a heap of current. One other potintial
problem here is the speaker cables. Not only any voltage loss from resistance but the effects of any inductance will be much
more noticable.
Woody
be a good place to use a power follower, as we might not need
any voltage gain just a heap of current. One other potintial
problem here is the speaker cables. Not only any voltage loss from resistance but the effects of any inductance will be much
more noticable.
Woody
Good point. The cable could easily end up contributing a substantial portion of of a direct-drive speaker's Re. Perhaps it would be best to make the ribbons "active" and skip speaker wires altogether.
Hi all,
I am in total agreement. If you are willing to undertake the additional work, an amplifier specifically designed to drive the ribbon would be the "best" way to go.
I don't want to eliminate all the possible ribbon builders who are not also amplifier builders. For them, a transformer is an easier fit to their existing modes of amplification.
There is a world of design possibilities, each with advantages and disadvantages. Like what has been shown with the magnetic circuit simulations, sometimes the gains of a particular path are large and sometimes they are small. Direct drive is better than transformer coupling, but I can also testify from personal experience that, when properly designed, transformer coupling is not going to be the make or break variable.
Mark
I am in total agreement. If you are willing to undertake the additional work, an amplifier specifically designed to drive the ribbon would be the "best" way to go.
I don't want to eliminate all the possible ribbon builders who are not also amplifier builders. For them, a transformer is an easier fit to their existing modes of amplification.
There is a world of design possibilities, each with advantages and disadvantages. Like what has been shown with the magnetic circuit simulations, sometimes the gains of a particular path are large and sometimes they are small. Direct drive is better than transformer coupling, but I can also testify from personal experience that, when properly designed, transformer coupling is not going to be the make or break variable.
Mark
This is about the only "make your own driver" thread that seems like it could really work! The new cheap magnets are part of the reason- the cheap steel for the "yoke?" also is a good break...
The special amp has to continue to be an option!
The special amp has to continue to be an option!
Hi,
Here some numbers as an indication. The DC resistance of the ribbon of my DK30’s is app. 20 mohms. Making an amp that directly drives this is almost impossible with reasonable efficiency and low distortion. The turns ratio of the DK30 transformer is 1:20. This makes a neat 20^2 * 20 e-3 = 8 ohms load for the amp. The transformer has to come in at 1 kHz and considering it is 2nd order filtered at above 2 kHz it can be a small ferrite transformer directly mounted on the tweeter frame.
But you can buy special ready made transformers for ribbon tweeters at http://users.argonet.co.uk/business/hsdawson/price.html for £25.- each.
Cheers 😉
Here some numbers as an indication. The DC resistance of the ribbon of my DK30’s is app. 20 mohms. Making an amp that directly drives this is almost impossible with reasonable efficiency and low distortion. The turns ratio of the DK30 transformer is 1:20. This makes a neat 20^2 * 20 e-3 = 8 ohms load for the amp. The transformer has to come in at 1 kHz and considering it is 2nd order filtered at above 2 kHz it can be a small ferrite transformer directly mounted on the tweeter frame.
But you can buy special ready made transformers for ribbon tweeters at http://users.argonet.co.uk/business/hsdawson/price.html for £25.- each.
Cheers 😉
Your numbers seem right. It was too good to be true that the ribbon resistance wuold be up around half an ohm.
Nice resources too....
For a tube amp, couldn't you make or have made a special output transformer that has a ratio to drive ultra low impedences rather than a second transformer on the driver?
Nice resources too....
For a tube amp, couldn't you make or have made a special output transformer that has a ratio to drive ultra low impedences rather than a second transformer on the driver?
This site Ribbon speakers - DIY shows a small hand wound ferrite toroid transformer that the author says sounds great.
Regards
James
Regards
James
Hi,
Certainly. You'll have to filter somewhere though: at the amp input or using an active x-over , for instance.
Hmmm....big 1 Ohm Apogees driven straight from a tube amp....
Cheers,😉
Certainly. You'll have to filter somewhere though: at the amp input or using an active x-over , for instance.
Hmmm....big 1 Ohm Apogees driven straight from a tube amp....
Cheers,😉
Greetings to all diy designers,
Just a hint for the amp people. No almost purely resistive load is impossible to driver. And regardless of load, a watt is still a watt.
Amplifiers have miniumum load because of current doubling as loads decrease (power also increases). If 20 watts is your amplification goal, then very low power supply rails will make for an extremely stable and happy amplifier even driving a .01 ohm load. If you want to have a high damping factor amplifier, then lots of low voltage outputs run in parallel with heavy gage (or is it gauge) wiring running through the power supply, output stage, and finally to the ribbon are needed.
Neither is a particularly difficult design problem. For the Pass group, the challenge is the power supply. While the amplification circuit is simple, the power supply is not. Here, however, is where the battery group has the edge. With the low power supply rails, a dedicated ribbon drive amplifier is the thing when running off of high capacity car batteries (with suitable bypassing and charging circuits).
Good designing,
Mark
Just a hint for the amp people. No almost purely resistive load is impossible to driver. And regardless of load, a watt is still a watt.
Amplifiers have miniumum load because of current doubling as loads decrease (power also increases). If 20 watts is your amplification goal, then very low power supply rails will make for an extremely stable and happy amplifier even driving a .01 ohm load. If you want to have a high damping factor amplifier, then lots of low voltage outputs run in parallel with heavy gage (or is it gauge) wiring running through the power supply, output stage, and finally to the ribbon are needed.
Neither is a particularly difficult design problem. For the Pass group, the challenge is the power supply. While the amplification circuit is simple, the power supply is not. Here, however, is where the battery group has the edge. With the low power supply rails, a dedicated ribbon drive amplifier is the thing when running off of high capacity car batteries (with suitable bypassing and charging circuits).
Good designing,
Mark
Yes, speaking of Pass, here's the white paper regarding his First Watt current amp. Very interesting comments about ribbon tweeters in it. Apparently damping isn't that important for the ribbon, (makes sense) and no matching trans required.
http://www.firstwatt.com/current_source_amps_1.htm
I'm sure someday we will have a DIY version of the amp
http://www.firstwatt.com/current_source_amps_1.htm
I'm sure someday we will have a DIY version of the amp
Frost,
Any updates on your progress? How close are you to testing or listening?
Anyway, a design consideration is the change in field strength in the air gap. Obviously it will fall off and thus change across the width of the ribbon.
I do not know how many commercial ribbons use this technique, but by moving the ribbon to the front of the gap and then floating (magnetically) twin flux carriers in front of the gap on either side of the ribbon, you can produce nearly flat flux strength across the ribbon.
While these results are a simulation, the ribbon tweeter does exist as proof of concept and I can post acoustic performance results. The ribbon is small and is only usable down to 3 kHz. There is no reason it cannot scale.
This implementation uses a small chunk of ceramic magnet, so the field strength is low. You also have to give up some field strength to make it even across the gap. As always, design is mostly about trade-offs and which variable matters most.
Good designing and building,
Mark
Any updates on your progress? How close are you to testing or listening?
Anyway, a design consideration is the change in field strength in the air gap. Obviously it will fall off and thus change across the width of the ribbon.
I do not know how many commercial ribbons use this technique, but by moving the ribbon to the front of the gap and then floating (magnetically) twin flux carriers in front of the gap on either side of the ribbon, you can produce nearly flat flux strength across the ribbon.
While these results are a simulation, the ribbon tweeter does exist as proof of concept and I can post acoustic performance results. The ribbon is small and is only usable down to 3 kHz. There is no reason it cannot scale.
This implementation uses a small chunk of ceramic magnet, so the field strength is low. You also have to give up some field strength to make it even across the gap. As always, design is mostly about trade-offs and which variable matters most.
Good designing and building,
Mark
An externally hosted image should be here but it was not working when we last tested it.
Has anyone looking at making ribbon speakers checked out a book by Justus v verhagen in the following link:
http://www.hificollective.co.uk/books/bk5007.html
And do they know if its worth spending over £20 on a book with less than 100 pages?
A couple of questions:
Is there a particular reason why Aluminium is used for the ribbons rather than something like copper? Especially as you need to have a transformer to get the correct impedance.
What is the average thickness of a ribbon? I've found some good looking 2'x4' sheets of copper at 0.3mm that I'm tempted to try and use.
Once I get paid and can buy my pile of 2x1x0.5" neos (at £4 each) I'll see what I can achieve with a real put it together and try it attitude. I'm thinking of getting about 20 mags and starting with a tweeter, then the next step is a 72 inch monster. I cant wait......
Nick.
http://www.hificollective.co.uk/books/bk5007.html
And do they know if its worth spending over £20 on a book with less than 100 pages?
A couple of questions:
Is there a particular reason why Aluminium is used for the ribbons rather than something like copper? Especially as you need to have a transformer to get the correct impedance.
What is the average thickness of a ribbon? I've found some good looking 2'x4' sheets of copper at 0.3mm that I'm tempted to try and use.
Once I get paid and can buy my pile of 2x1x0.5" neos (at £4 each) I'll see what I can achieve with a real put it together and try it attitude. I'm thinking of getting about 20 mags and starting with a tweeter, then the next step is a 72 inch monster. I cant wait......
Nick.
Aluminum is lighter and more electrically resistive than copper. Eminently more suitable.Grumpy_Git said:
I believe 0.3mm is too thick.
I assumed that 0.3mm would be too thick but best to ask. As for the resistivity, as we are using a transformer to get the appropriate impedance surely this is not too much of an issue?
N
N
The data on a Fountek driver says the ribbon is .009 mm thick
So about one hundredth of a millimeter.
So about one hundredth of a millimeter.
MarkMck: I think the Apogee Acoustics patents during the 1980's had dual-pole magnetic field design along the lines of your posted FEMM model.
The Magnepan tweeter ribbons are pure aluminum and about 2.5 microns thick.
The current Raven ribbons are pure aluminum and about 9 microns thick. This seems to be popular thickness.
5 microns is probably the thinnest pure aluminum that is robust enough for good reliability. Some non-magnetic alloys like 5056 have good strength and annealing properties which can be useful for a ribbon
aluminum
2.65 10-8 ohm m; or micro_ohm cm
2700 kg/m3
copper
1.7 10-8 ohm m; or micro_ohm cm
8920 kg/m3
A copper ribbon would be about 3.3x less power efficient than aluminum.
A=ribbon area meter^2
B=magnetic field in Telsa
%= percent of ribbon which conducts current(100=100%)
m= mass of ribbon in grams
Efficiency = A^2 * B^2 * % / m
spl 112.2 + 10*(LOG(efficiency)/LOG(10))
senS effeciency + LOG(8/resistance)/LOG(10)
The Magnepan tweeter ribbons are pure aluminum and about 2.5 microns thick.
The current Raven ribbons are pure aluminum and about 9 microns thick. This seems to be popular thickness.
5 microns is probably the thinnest pure aluminum that is robust enough for good reliability. Some non-magnetic alloys like 5056 have good strength and annealing properties which can be useful for a ribbon
aluminum
2.65 10-8 ohm m; or micro_ohm cm
2700 kg/m3
copper
1.7 10-8 ohm m; or micro_ohm cm
8920 kg/m3
A copper ribbon would be about 3.3x less power efficient than aluminum.
A=ribbon area meter^2
B=magnetic field in Telsa
%= percent of ribbon which conducts current(100=100%)
m= mass of ribbon in grams
Efficiency = A^2 * B^2 * % / m
spl 112.2 + 10*(LOG(efficiency)/LOG(10))
senS effeciency + LOG(8/resistance)/LOG(10)
Great ribbon info here
You may want to go through this site http://ldsg.snippets.org/ALSR/index.php3#LSOURCE where John Whittaker goes through his experiences with building and measuring ribbons.
Hige amount of info and very helpful chap.
Ruvane.
You may want to go through this site http://ldsg.snippets.org/ALSR/index.php3#LSOURCE where John Whittaker goes through his experiences with building and measuring ribbons.
Hige amount of info and very helpful chap.
Ruvane.
- Status
- Not open for further replies.