Hi
As not having any experience in magnet circuit design or FEM simu, I would like to ask for your thoughts and help about an alternative magnet circuit for AMT's
Modern AMT's are basically built as simple "Hamburger's" – two iron plates (front and back) with vertical neodymium bar's over the full height in between (one each side).
The iron plates are perforated like known to allow the diaphragm to breath through the iron plates.
An alternative way to built the magnet circuit would be to use horizontal neodymium bars instead of the iron plates at front and back and close the magnet circuit by a heavy U-shape iron at the sides.
The U-shape iron would have to be "outside" adding depth compared to the usual design and also would have the disadvantage to generate more parasitic losses IMO.
The upside possibly could be stronger field due to more magnet material in use and possibly more equally distributed field over the width of the diaphragm.
What you think?
Michael
As not having any experience in magnet circuit design or FEM simu, I would like to ask for your thoughts and help about an alternative magnet circuit for AMT's
Modern AMT's are basically built as simple "Hamburger's" – two iron plates (front and back) with vertical neodymium bar's over the full height in between (one each side).
The iron plates are perforated like known to allow the diaphragm to breath through the iron plates.
An alternative way to built the magnet circuit would be to use horizontal neodymium bars instead of the iron plates at front and back and close the magnet circuit by a heavy U-shape iron at the sides.
The U-shape iron would have to be "outside" adding depth compared to the usual design and also would have the disadvantage to generate more parasitic losses IMO.
The upside possibly could be stronger field due to more magnet material in use and possibly more equally distributed field over the width of the diaphragm.
What you think?
Michael
Last edited:
the second configuration you describe ( multiple magnet bars with U shaped iron at the sides ) will be MUCH BETTER than the traditional one.
it will be much better for the same reason why neo-radial motors in dynamic drivers are better than traditional ones. neo-radial motors achieve over 90% field utilization while regular motors can be as bad as 30% or worse.
you will in fact be able to use LESS neodymium and still produce stronger field.
i was actually thinking about it just yesterday. i was thinking if i was doing an AMT thats how i would do it.
Thanks for your encouraging words.
Looking up the term "neo-radial motors" brought me to some interesting threads about motor design.
What I learned is that there are several neodymium grades to choose from and also different grades of iron I have to do "the optimally pick" for the purpose. Not sure though, how critically this choices may be?
At a first glance it seems the filed could be less strong with wrong material choice but materials would not affect linearity of field over gap width - but possibly there are other adverse effects to be considered besides resulting field strength when choosing materials used?
Michael
Looking up the term "neo-radial motors" brought me to some interesting threads about motor design.
What I learned is that there are several neodymium grades to choose from and also different grades of iron I have to do "the optimally pick" for the purpose. Not sure though, how critically this choices may be?
At a first glance it seems the filed could be less strong with wrong material choice but materials would not affect linearity of field over gap width - but possibly there are other adverse effects to be considered besides resulting field strength when choosing materials used?
Michael
Last edited:
frankly i don't know what exactly is the difference between different neodymium grades but i don't think it will be critical or even important as long as it is in fact a neodymium magnet and magnetic steel ( some steel is not magnetic, like certain stainless steels ).
sometimes in a neo-radial subwoofer magnets can fail when subjected to extreme heat from the voice coils to which they are directly adjacent. in such an application it DOES matter what magnet is used. i have seen pictures of the magnets which have failed - they had their protective coating peel off of them. also a magnet can demagnetize at a temperature.
for some extreme designs more exotic materials like AlNiCo, Permendur, Telar, S.L.M.M ( which appears to be ferrite ? ) may be in order but for an AMT i don't think it should matter.
in a normal driver the voice coil is a solenoid so it can generate a relatively strong field to which the steel will react resulting in undesirable effects which can be mitigated by using different materials ( although that's not the best way ). also in some applications like ultra-high-end dome & compression tweeters it is very desirable to have maximum gap field strength so if you can use a material which has a higher saturation point than steel that may be a good idea.
but as far as i can tell in an AMT neither of those should be a concern. the field from the diaphragm should mostly cancel itself out due to the conductor running back and forth. and the magnetic field requirement for ribbons is not nearly as great as for lets say compression drivers. and the magnets aren't likely to see any serious temperatures.
Agree that thermal effects for *magnet materials* will be of lesser concern in tweeter application and the large radiation surface that goes along with the motor structure of AMT.
What I would like to concentrate on is that field strength should not significantly drop below what is current standard in order not to scarify max SPL.
Give a certain field strength, IMO max SPL is basically limited by heat produced by current flow and subsequently the max. temperature the foil can withstand (at least in the upper frequency department).
So - keeping efficiency at high level is a goal - not necessary to make mayor progress tho..
Seems the iron grades 1008 (1010 12L14) are good choices from what I've read but have no idea how these are called in European names / types and where to source from?
Second thing I'm concerned of is field uniformity over the width of the diaphragm. Any pleat is basically independently moved (quite in contrary to a cone speaker) - so it is desirable to have good control over SPL uniformity of all pleats.
Michael
What I would like to concentrate on is that field strength should not significantly drop below what is current standard in order not to scarify max SPL.
Give a certain field strength, IMO max SPL is basically limited by heat produced by current flow and subsequently the max. temperature the foil can withstand (at least in the upper frequency department).
So - keeping efficiency at high level is a goal - not necessary to make mayor progress tho..
Seems the iron grades 1008 (1010 12L14) are good choices from what I've read but have no idea how these are called in European names / types and where to source from?
Second thing I'm concerned of is field uniformity over the width of the diaphragm. Any pleat is basically independently moved (quite in contrary to a cone speaker) - so it is desirable to have good control over SPL uniformity of all pleats.
Michael
Last edited:
I'm not the one to be telling this to people but i will say it anyway:
you think too much. just do it.
you aren't performing open heart surgery here. what's the worst thing that can happen ?
i can assure you that less solid research went into Copenhagen treaty than you already put into this AMT.
also performance can always be improved and therefore irrelevant unless you are in a competition.
you think too much. just do it.
you aren't performing open heart surgery here. what's the worst thing that can happen ?
i can assure you that less solid research went into Copenhagen treaty than you already put into this AMT.
also performance can always be improved and therefore irrelevant unless you are in a competition.
Last edited:
bad habits die slow...
considering myself as a "passionate thinker" - I enjoy to look at topics form as many sides as possible..
Have looked up today how to measure field uniformity and was surprised that ready available IC are so cheap and versatile (even buffered voltage output proportional and calibrated to field strength - wow!)
Last time I was knocking together some magnetic measuring equipment there were only mere Hall elements available....
My web search regarding sources for small quantity of low carbon steel was not really successful I'm afraid (all I came up with was the name "ARMCO" steel - available in several qualities)
Possibly I should follow your advice and go with what is called St37 here - the most common, cheap and unspeced (!) steel around - problem is - I'll never know what I trade in performance wise...
Michael
considering myself as a "passionate thinker"
- I enjoy to look at topics form as many sides as possible..Have looked up today how to measure field uniformity and was surprised that ready available IC are so cheap and versatile (even buffered voltage output proportional and calibrated to field strength - wow!)
Last time I was knocking together some magnetic measuring equipment there were only mere Hall elements available....
My web search regarding sources for small quantity of low carbon steel was not really successful I'm afraid (all I came up with was the name "ARMCO" steel - available in several qualities)
Possibly I should follow your advice and go with what is called St37 here - the most common, cheap and unspeced (!) steel around - problem is - I'll never know what I trade in performance wise...
Michael
Last edited:
i can tell u right now if you go with the design you described field uniformity left-right ( as looking at a tweeter typically installed in a baffle ) will be almost perfect.
on the other hand if we do a vertical cross section down the middle of that AMT perpendicular to baffle surface most likely it will look like some kind of nightmare. not sure what you can do about it or most importantly whether it would make any difference ?
frankly i don't think field uniformity at all matters in AMT. the push-pull design inherently cancels out just about everything which is the entire point of AMT but not many people realize this.
distortion is a result of nonuniform gain or efficiency, not field. it doesn't matter what the field looks like as long as overall gain or efficiency is uniform.
in a typical speaker each voice coil turn moves through a VERY nonuniform field but the design is such that contributions form different voice coil turns average out to almost constant BL
on the other hand if we do a vertical cross section down the middle of that AMT perpendicular to baffle surface most likely it will look like some kind of nightmare. not sure what you can do about it or most importantly whether it would make any difference ?
frankly i don't think field uniformity at all matters in AMT. the push-pull design inherently cancels out just about everything which is the entire point of AMT but not many people realize this.
distortion is a result of nonuniform gain or efficiency, not field. it doesn't matter what the field looks like as long as overall gain or efficiency is uniform.
in a typical speaker each voice coil turn moves through a VERY nonuniform field but the design is such that contributions form different voice coil turns average out to almost constant BL
Last edited:
Yeah - AMT has its merits - where others have to develop all kinds of compensation patents X^2 and so on - we get it for free!
Quite a different topic but something I have an eye on:
vertical field "nightmare" (due to perforation needed) isn't really my concern - simply as one and the same pleat is affected and it will sum up to whatever acceleration / movement this very pleat finally will have.
On the other hand looking at field uniformity over the *width* of the AMT diaphragm is an entirely different story.
Any non-uniformity here will result in pleats radiating different SPL compared to all others. This may or may not be desirable as I had to find out most obviously with different versions of the NEO3-W speaker in my dipole horn.
In short, SPL distribution over the width *strongly* influences polar pattern of a diaphragm.
So, to control and vary this spec to my desire is an interesting exercise - if can be done at all.
Michael
Quite a different topic but something I have an eye on:
vertical field "nightmare" (due to perforation needed) isn't really my concern - simply as one and the same pleat is affected and it will sum up to whatever acceleration / movement this very pleat finally will have.
On the other hand looking at field uniformity over the *width* of the AMT diaphragm is an entirely different story.
Any non-uniformity here will result in pleats radiating different SPL compared to all others. This may or may not be desirable as I had to find out most obviously with different versions of the NEO3-W speaker in my dipole horn.
In short, SPL distribution over the width *strongly* influences polar pattern of a diaphragm.
So, to control and vary this spec to my desire is an interesting exercise - if can be done at all.
Michael
Last edited:
not particularly interesting to me but yes it can very easily be done.
if you really think this is interesting you should definitely model it.
i am betting that you will not need any correction but should you decide to do it you would only need to bend one of the metal plates to vary the width of the gap and the field will shift to where the gap is narrower.
on the other hand having the magnet line the gap as in your design as well as neo-radial design has the inherent property of making the field tend towards uniform so it should be less responsive to bending.
magnetic field is like water. to describe something like the shape of the water drop mathematically is very complex but the human brain through EXPERIENCE can easily predict this shape without performing any calculations !
if you log the same amount of experience with modeling magnetic fields as you have with dealing with water you will just as easily be able to predict how the field will flow ( using only your imagination and intuition, no math of any kind ) as you can now predict how water would flow if you lets say accidentally spilled your drink on the table.
and if you can't ? then i will be better than you hahahahaha ! ! !
anyway, definitely try to model it. i don't think you can model this design EXACTLY in FEMM but you can model some approximations which will give you the intuitive feel you need.
Got your point.
Neodymium bars are possibly hard to bend but some curved iron would be easy to glue on...
But all thats for further steps if at all....
First interest is in nailing a good starting point to avoid too many "built iterations" - if possibly.
Given the neodymium bars would be roughly 6x6mm, what could be needed for the iron U-shape as a cross section and how much overlap could be needed?
Is there a rule of thumb which iron versus magnet material cross section ration should be ok?
FEMM certainly would help - but frankly I don't have the guts
Michael
FEMM certainly would help - but frankly I don't have the guts
Michael
it's easy.
step 1: google for "FEMM" download and install.
step 2: find the tutorial PDF in the downloaded package and go through it step by step.
congratulations ! you now know how to use FEMM ! takes about 2 minutes for step 1 and another 30 minutes for step 2.
i'll help you. just try it.
HELLO WORLD !
Thanks for your support and for encouraging me!
Pretty picture I have to say – but not really reasonable - what you think?
Why is there so minimal field in the U-shapes for example?
Michael
An externally hosted image should be here but it was not working when we last tested it.
Thanks for your support and for encouraging me!
Pretty picture I have to say – but not really reasonable - what you think?
Why is there so minimal field in the U-shapes for example?
Michael
Last edited:
Adding some "air" around the construction makes the field lines in the U-shape more reasonable but also makes the field in the magnet bars and in the gap looking strange ?
Michael
An externally hosted image should be here but it was not working when we last tested it.
Michael
...how the current AMT are built:
Aarrgh ?
*This* looks scary reasonable....
Do I already have to draw the final conclusion my alternative magnet circuit is way off ?
Michael
An externally hosted image should be here but it was not working when we last tested it.
Aarrgh ?
*This* looks scary reasonable....
Do I already have to draw the final conclusion my alternative magnet circuit is way off ?
Michael
...seems I would have to extend the U-shape to completely surround the neodymium bars to make the magnetic circuit work :
Quick learning with such tools – wow...
No idea if that's worth the effort, though.
Michael
An externally hosted image should be here but it was not working when we last tested it.
Quick learning with such tools – wow...
No idea if that's worth the effort, though.
Michael
I've already have seen above AMT arrangement somewhere
I guess its been from Neil Davis - have to look it up...
#######
Yes - here:
http://www.diyaudio.com/forums/planars-exotics/149226-neil-davis-heil-amt.html
Not exactly the same magnetic circuit but in its appearance close to...
Michael
I guess its been from Neil Davis - have to look it up...
#######
Yes - here:
http://www.diyaudio.com/forums/planars-exotics/149226-neil-davis-heil-amt.html
Not exactly the same magnetic circuit but in its appearance close to...
Michael
Last edited:
there you go
that's the design i was talking about all along. the field is very uniform just as i said it would be.
i didn't realize you wanted to leave a hole in the middle. you can't have that hole because you need to have a high-permeability return path for the field. but you probably realize this by now.
the one thing you may still want to do with FEMM is learn how to make the PLOT of the field. it won't really show you anything that you cannot already see from these pictures you posted but it will show field strength with much better resolution. for example if you want to see the difference between this design and the traditional one the plot will actually have the resolution to show the difference if there is any.
i don't remember the menu names but you basically need two points. then in one of the windows where diagrams are shown there is an icon with red line or something which you draw between the points. then you select "plot XY" or something. basically it plots the field along the red line.
for example here:
http://www.diyaudio.com/forums/planars-exotics/155042-i-wrong.html
i got the field to be so uniform that you can only really see how uniform it is in the plots.
also something you may not have thought about. one way to think of efficiency of a motor is to count the magnetic lines going through the magnet and the number of lines going through the gap. if all of the lines that go through the magnet also go through the gap the design is very efficient. this is more or less the inherent property of neo-radial motors and why i recommended this configuration to you.
on the other hand if you look at the "traditional" motor some of the lines coming out of the magnet short-circuit back to the magnet without ever reaching the magnetic gap where the ribbon is. this is why i said with this design you are likely to get more useful field out of less neodymium.
the downside will be that the design might be less acoustically transparent than the "traditional" one due to more stuff hanging in front of the ribbon.
on the other hand if you look at the "traditional" motor some of the lines coming out of the magnet short-circuit back to the magnet without ever reaching the magnetic gap where the ribbon is. this is why i said with this design you are likely to get more useful field out of less neodymium.
the downside will be that the design might be less acoustically transparent than the "traditional" one due to more stuff hanging in front of the ribbon.
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.