Hi Guys,
Would you mind sharing a bit of your knowledge about the construction of a planar tweeter so it may help me make good choices on how I will repair them ?
The actual drivers are TEC(hnology) LS Plan (more that 10 years old)
http://www.ebacoustic.it/prod_images/LSplan.pdf
You will see below a picture of the parts
1- I noticed that each magnet was covered with a thin foil that look like aluminum, is that just about preventing from corrosion or do that play a part on the behavior of the driver ?
2- Each magnet under that foil have some kind of amalgam semi solid powder like and magnetized, what is it ? Magnet wearing off ?
3- besides measuring that the impedance of the diaphragm is according to the nominal impedance of the datasheets is there any other factor to take into account ?
Thanks
Would you mind sharing a bit of your knowledge about the construction of a planar tweeter so it may help me make good choices on how I will repair them ?
The actual drivers are TEC(hnology) LS Plan (more that 10 years old)
http://www.ebacoustic.it/prod_images/LSplan.pdf
You will see below a picture of the parts
1- I noticed that each magnet was covered with a thin foil that look like aluminum, is that just about preventing from corrosion or do that play a part on the behavior of the driver ?
2- Each magnet under that foil have some kind of amalgam semi solid powder like and magnetized, what is it ? Magnet wearing off ?
3- besides measuring that the impedance of the diaphragm is according to the nominal impedance of the datasheets is there any other factor to take into account ?
Thanks
Is it acting as a shim, maybe a bump stop?
Sounds like it may be powdered, insulated iron, or ferrite.
Pliability, insulation?
1 - Since the magnetic field gets very quickly weaker with the distance, it's absolutely impossible these were shims. Any added distance from the diaphragm to the magnets would introduce a greatly reduced spl and vastly increase the unlinear distortion since the difference between the magnet motor strength between the positive and negative excursion would differ a LOT.
That was the coating of the neodymium magnets which should have prevented the corrosion of the magnets. It was either damaged mechanically or by chemical influence (salt water air close? Aggressive cleaning agents?), a small damage/opening is enough and as a result it practically completely corroded (destroyed) the magnet.
2 - That are the remnants of the ND magnet sticks. The suggestion it might be a ferrite magnet is ridiculous and absolutely impossible, because these do not need any protection from corrosion, they are already oxidized (Fe3O4) and do not react to air moisture anymore.
3 - To suspect the insulation being defective would mean the diaphragm would have gathered conductive material between the circuit paths. According to the photo, there is NONE foreign material present. Which means, that theory is completely destroyed. However, the folds in the diaphragm can not be corrected to the former performance, even if the rest of the driver would be restored, a high increase of distortion will show in a more or less wide frequency range.
Conclusion: You'd have to replace the ND magnet sticks with the exact same dimensions, magnetic flux and field shape. It's next to impossible to get these, let alone for cheap. The damaged diaphragm (look at all these folds and waves!) cannot be salvaged either with acceptable distortion and response, which means, even if you'd find the parts, it would be a lot more expensive than getting new or used drivers. Even if you did, to assemble them again to the factory low tolerances would be impossible.
Or in other words: Sorry but it's a total loss. Magnetostats aren't that expensive anymore, buy different ones or AMTs.
That was the coating of the neodymium magnets which should have prevented the corrosion of the magnets. It was either damaged mechanically or by chemical influence (salt water air close? Aggressive cleaning agents?), a small damage/opening is enough and as a result it practically completely corroded (destroyed) the magnet.
2 - That are the remnants of the ND magnet sticks. The suggestion it might be a ferrite magnet is ridiculous and absolutely impossible, because these do not need any protection from corrosion, they are already oxidized (Fe3O4) and do not react to air moisture anymore.
3 - To suspect the insulation being defective would mean the diaphragm would have gathered conductive material between the circuit paths. According to the photo, there is NONE foreign material present. Which means, that theory is completely destroyed. However, the folds in the diaphragm can not be corrected to the former performance, even if the rest of the driver would be restored, a high increase of distortion will show in a more or less wide frequency range.
Conclusion: You'd have to replace the ND magnet sticks with the exact same dimensions, magnetic flux and field shape. It's next to impossible to get these, let alone for cheap. The damaged diaphragm (look at all these folds and waves!) cannot be salvaged either with acceptable distortion and response, which means, even if you'd find the parts, it would be a lot more expensive than getting new or used drivers. Even if you did, to assemble them again to the factory low tolerances would be impossible.
Or in other words: Sorry but it's a total loss. Magnetostats aren't that expensive anymore, buy different ones or AMTs.
I didn't realize it's a dipole, thanks to mention that important fact. That will be more expensive than I thought I'm afraid. I'll list AMTs and ribbons too.
The Mundorf 17D2.2 would be a good dipole candidate. It's an air motion transformer and dipole, higher dynamic than the magnetostat. Not cheap though. Mundorf got other dipole AMTs but they are much more expensive.
It's also an option to use 2 non-dipol drivers, one to the front, the other playing to the back with reversed polarity. Two each of these would be needed per side.
The OA AMT-50 (very cheap and good sound, lower spl though, should be crossed over high, >5k)
AMT_OEM (can likely modified to a dipole)
Mivoc KFT 130 M (very good magnetostat, cheap, very high spl because of the horn, I think it can be removed)
Monacor RBT-1000 (Magnetostat, expensive)
Monacor RBT-95 (Magnetostat)
Monacor RBT-10 (real ribbon, they are only sold in pairs, the price is therefore for two of them.)
Visaton MHT 12 (expensive but excellent!)
There are also others which I don't remember right now.
The Mundorf 17D2.2 would be a good dipole candidate. It's an air motion transformer and dipole, higher dynamic than the magnetostat. Not cheap though. Mundorf got other dipole AMTs but they are much more expensive.
It's also an option to use 2 non-dipol drivers, one to the front, the other playing to the back with reversed polarity. Two each of these would be needed per side.
The OA AMT-50 (very cheap and good sound, lower spl though, should be crossed over high, >5k)
AMT_OEM (can likely modified to a dipole)
Mivoc KFT 130 M (very good magnetostat, cheap, very high spl because of the horn, I think it can be removed)
Monacor RBT-1000 (Magnetostat, expensive)
Monacor RBT-95 (Magnetostat)
Monacor RBT-10 (real ribbon, they are only sold in pairs, the price is therefore for two of them.)
Visaton MHT 12 (expensive but excellent!)
There are also others which I don't remember right now.
That is already a very good list to start with, thanks for posting it!
Indeed I could double the tweeters, but the cost double also unfortunately, and it don't fit the actual design
I have put the tweeter together and unfortunately the impedance is bad, there is a small dent in the path of the diaphragm that I could not see without a magnifying glass, so scraped they are.
Indeed I could double the tweeters, but the cost double also unfortunately, and it don't fit the actual design
I have put the tweeter together and unfortunately the impedance is bad, there is a small dent in the path of the diaphragm that I could not see without a magnifying glass, so scraped they are.
-while there are some really cheap dipole planar's.. it's best to avoid them (Dayton has one).
There are 2 cheap'ish ones I'd consider, one of which needs modification (as in the back/enclosure removed):
Beston RT002-A, and
Radian LT-2 (which needs the back removed).
I'd probably choose the Radian for its greater surface area and likely lower distortion (typically making the crossover a bit easier), though other than 5-6 kHz 2nd order peak - the Beston isn't bad and does have greater horizontal dispersion + a natural "inset" from the waveguide (which can help with the crossover).
Both will need considerable *"break-in" where the film layer becomes more compliant.
*Note: this is not like "break-in" of a typical dynamic driver where you are stressing the excursion potential, rather it's more about time and heat. The best way to see this effect is a freq. graph withOUT smoothing where the newer/less break-in generally has more exaggerated peaks and dips (..if fewer overall).
Both will need a very good resistor and cap. components to achieve their best potential, absolutely more than the driver itself, and this is one area where you can do subjectively better AND spend less money by making your own resistor (..though a good cap is still hideously expensive).
Both should also provide an easier cross-over (filter + eq. for both low-pass and high-pass) than the LS tweeter spec..
There are 2 cheap'ish ones I'd consider, one of which needs modification (as in the back/enclosure removed):
Beston RT002-A, and
Radian LT-2 (which needs the back removed).
I'd probably choose the Radian for its greater surface area and likely lower distortion (typically making the crossover a bit easier), though other than 5-6 kHz 2nd order peak - the Beston isn't bad and does have greater horizontal dispersion + a natural "inset" from the waveguide (which can help with the crossover).
Both will need considerable *"break-in" where the film layer becomes more compliant.
*Note: this is not like "break-in" of a typical dynamic driver where you are stressing the excursion potential, rather it's more about time and heat. The best way to see this effect is a freq. graph withOUT smoothing where the newer/less break-in generally has more exaggerated peaks and dips (..if fewer overall).
Both will need a very good resistor and cap. components to achieve their best potential, absolutely more than the driver itself, and this is one area where you can do subjectively better AND spend less money by making your own resistor (..though a good cap is still hideously expensive).
Both should also provide an easier cross-over (filter + eq. for both low-pass and high-pass) than the LS tweeter spec..
Last edited:
CAUTION: a lot of those AMT's cannot be easily modified for dipole use because of the way the terminals are placed ON the rear chamber.
In contrast the terminals of the Radian are BELOW the rear chamber, so it should be easier to modify.
Added CAUTION:
Almost all film-type tweeters should NEVER have their connections soldered to. Doing so destroys the etching and the film it's on unless you have some sort of near "instant" soldering iron that has exceptional short heat application and resulting fast cooling (..that industry uses).
In contrast the terminals of the Radian are BELOW the rear chamber, so it should be easier to modify.
Added CAUTION:
Almost all film-type tweeters should NEVER have their connections soldered to. Doing so destroys the etching and the film it's on unless you have some sort of near "instant" soldering iron that has exceptional short heat application and resulting fast cooling (..that industry uses).
Last edited:
The Omnes Audio AMT 50 are very cheap (25€) and are very shallow, just 22mm according to the data sheet, it would only add the thickness of the baffle and only maybe catching the eye when looking at it from the side. It doesn't even have to be mounted at exactly the same height, reducing the thickness to just the 22mm of its thickness. With the offset of the driver height (64mm, one above the other vertically), you won't be able to tell its offset, the back radiated sound can't be located precisely because you only hear it after one or several reflections. I cannot understand your arguments there, the reasons are just wrong. I'm fine with you not wanting any of the listed drivers, but don't bring up wrong reasons please because then nobody can actually help you to get what you really want.
The contacts can be cut out of the back chamber and mounted on a self-made holder. No need for soldering them.
You can solder these with a metal plate below the film, a drop of water between and low melting temperature solder. It's still very difficult, I would not recommend it even for experienced electronic veterans but it's possible.
But there are much easier solutions for that: You can use electric conductive tape or electric conductive glue. That works for the connecting points, it does not work well or very long on the moving part of the diaphragm though. And you can't do that where the mounting frame sits because it doesn't lay flat anymore and a lot of distortion is the result. The perfectly flat fit of the diaphragm is generally the biggest reasons of diverting response and high distortion, even if it's comming straight from the factory.
Thanks ScottG,
You'll always surprise me with brands I never heard of, nice
The two have indeed a pricing that is more manageable than most I have seen
The Beston, I have found a site in Denmark that sell them
I can't help but notice that the logo look like a brand for kids toys and they may have it right, big kids playing with toys
The Specs indicate 92db and the graph show it most of the time around 90dB, but on the following website the measurement show it a lot higher (at my hearing range) Test Bench: Beston Technology RT002A Ribbon Tweeter | audioXpress
What do you think about that ?
The Radian, it seem a lot more difficult to find, I have found on reference in a website that I am not very found of the service and they may be available in a few weeks to month
The efficiency is very high but that seem to be with a wave-guide, so I guess that without it should be lower but I suppose I'll have to attenuate it a lot if I want to cross with the 8ohm/90db mid ?
Unfortunately I'm still not experienced enough to understand how they are better for the crossover than the previous LSPlan
Don't charge me I didn't raise the red flag ICG (a.k.a the Bull
)I'm not rejecting everything anyway, I actually not had enough time yet to look in detail at all of them. One sure thing I don't want to do is heavy customization of the tweeters, I have enough to do already so if there are no easy way to make them dipole I prefer not to consider them for this project.
As for your proposition for two tweeters by cabinet it seem I didn't understand you well, I thought that you meant one in the front of the cabinet and one at the back, while in fact is back to back, and in that case it would be possible, especially for the cheaper one that you mention.
But then I'd have to conceive a short chamber behind the one on the front panel and then add the one at his back in another panel and run the wires in that chamber, but it's possible.
But I'd still prefer if possible to have a ready made dipole, simplify an already complex thing for me.
The Test Bench for the Beston is for the NON-dipolar version. Same basic driver but with a rear chamber. When you remove the rear chamber there is a low freq. extension increase (below 1.5 kHz) and spl at around 2.5 kHz DECREASES. Unfortunatly non-linear distortion increases at these lower freq.s. Basically that increasing (rising) response to about 14 kHz will need to be "shelved"/eq.ed-out and you'll need to add a bit of resistance to get it down to the average 89-90 db 1 watt 1 meter from its eq'ed average nearing 92-93 db (seen at 5-6 kHz in the Test Bench measurement): which should put you closer to 8 ohms.
The Radian has the same lower freq. effect as the Beston, but also has the horn-load effect that further increases pressure (near 2 kHz). In other words remove the back AND the horn and 2 kHz should be much closer to 92 db. Again, add a bit of resistance to that. Maybe some eq.. (but maybe not).
The problem with the LS Plan driver was very limited lower freq. extension (even with high distortion). That makes for very few crossover options: largely requiring a steep(er) high-pass for that driver and a resulting steep(er) filter for the Fostex: depending on the selected crossover freq.. With the Beston and the Radian you still have to crossover relatively high in freq. (because of rising non-linear distortion as freq.s lower), but you can probably do so with a lower order slope because of the increased lower freq. extension.
As for TLHP - eh, it's not like you got burned by them - instead it was their choice in shipping Co. (..and even then it did get to you, if belated).
More troubling is the fact that it's not in stock. You could of course purchase it direct from Radian (..though I've no idea what the shipping and customs would be). They also sell through Thomann - so you could email them as well.
-anyway, I'd still hold-off on that until after hearing it with just the Fostex.
Of course there are other avenues to approach there as well: including monopole tweeters of varying type in a standard design, and even potentially just using a monopole tweeter as an "ambiance fill" a'la Wilson Audio Alexandria XLF.
Last edited:
Thanks for the clarifications ScottG,
It is as clear as today sky, sunny with a few clouds
As for TLHP - eh, it's not like you got burned by them - instead it was their choice in shipping Co. (..and even then it did get to you, if belated).
More troubling is the fact that it's not in stock. You could of course purchase it direct from Radian (..though I've no idea what the shipping and customs would be). They also sell through Thomann - so you could email them as well.
-anyway, I'd still hold-off on that until after hearing it with just the Fostex.
You are right as usual, I'll stop worrying about that tweeter aspect and get on with the work
Yes sure I'm quiet flexible to changes, sometimes caused by unforeseen events I'll see what I will end up on that path and if not satisfied I'll deviate
Don't charge me I didn't raise the red flag ICG (a.k.a the Bull
heh. Guilty as charged?
The modifications needed to change some of the drivers range from very easy to practically impossible because of the magnet structure and mechanical construction. The AMTs being there much easier to work with because they forgive a lot more because the diaphragm is always in a frame and you won't change the compilance of the foil with different pressure etc. But yeah, I can see why you want to avoid that and there's nothing wrong with that.
Why don't you post a picture of the speaker (also back) and the position in the room so giving a suggestion or a completely different solution is much easier?
The Monacor RBT-1000 is identical.
That's the general problem of magnetostats, they can't do excursion but lower frequencies automatically require exactly that. The horn/WG ofcourse was increasing the spl in some range, which means, the level goes down too and the distortion rises to reach the same spl because it needs more power (and excursion) to do the same.
Yes, that's always a problem with magnetostats, AMTs are much easier regarding that and, on top of that, they got a lot better dynamic and much slower distortion rise. But to replace the original driver isn't that demanding, the old (destroyed) tweeter couldn't be used below 5k either.
Yes, that's a viable option. And: The back driver does not have to be the same as the front one, it just has to behave similarily. Even a slightly different response is still okay, the back radiated sound isn't only reflected, it's partly absorbed. A 2nd driver on the back (or, if close to the wall, on top of the speaker) got also the advantage you could change the level of the back sound.
heh. Guilty as charged?
Why don't you post a picture of the speaker (also back) and the position in the room so giving a suggestion or a completely different solution is much easier?
Hehe natürlich
If you want to have a look at the project and it's progress (or lack of) :
Advices on First Crossover Design (VituixCAD2)
It goes pretty wild and the title does not illustrate anymore the content, on that specific page you will find the room and general appearance of the cabinet
Ah, okay. So this is the room and the placement?
That is very problematic for a dipole, the distance to the back is way too short, and a box like, relatively narrow opening (probably less than 30°?) to the back reduces the advantage of the dipole (great, wide, 'free', roomy impression). In such a situation, a top 2nd tweeter would provide these qualities in a much better way. I would suggest, build a test enclosure of the speaker, test the sound with an open back (or 2nd tweeter to the back) and closed and put a 2nd tweeter into a board, ~15x20cm. You can experiment with the angle/position and how you like it best. For such an experiment you don't need a dipole nor a foil tweeter, to get an impression, a cheap tweeter or fullrange will suffice.
While that will maybe solve the problems in the highs, you still will have the boxy sound in the mids. It would be a lot better to have just a baffle above the bass and a reflector board at ~45°. You should really, REALLY try that first before you invest a lot of money into the tweeters. A fullrange would be good for such a test, cheap and can go low enough to play the mids too. A good candidate for that would be the FRS 8 (4 or 8 Ohm version) or the FRS 5 X (very good). If these aren't loud enough, the R 10 S maybe is. I tried to look at other fullranges than the Visaton but they are either much more expensive or not loud enough. Anyway, the R 10 S misses a bit sparkle at the top but it's just for testing and despite being very cheap, still measures surprisingly good, well, apart from a bit missing at the top end. You could do another project with the fullranges or sell them, not much losses there unlike buying planar tweeters.
Ah, okay. So this is the room and the placement?
That is very problematic for a dipole, the distance to the back is way too short, and a box like, relatively narrow opening (probably less than 30°?) to the back reduces the advantage of the dipole (great, wide, 'free', roomy impression). In such a situation, a top 2nd tweeter would provide these qualities in a much better way. I would suggest, build a test enclosure of the speaker, test the sound with an open back (or 2nd tweeter to the back) and closed and put a 2nd tweeter into a board, ~15x20cm. You can experiment with the angle/position and how you like it best. For such an experiment you don't need a dipole nor a foil tweeter, to get an impression, a cheap tweeter or fullrange will suffice.
While that will maybe solve the problems in the highs, you still will have the boxy sound in the mids. It would be a lot better to have just a baffle above the bass and a reflector board at ~45°. You should really, REALLY try that first before you invest a lot of money into the tweeters. A fullrange would be good for such a test, cheap and can go low enough to play the mids too. A good candidate for that would be the FRS 8 (4 or 8 Ohm version) or the FRS 5 X (very good). If these aren't loud enough, the R 10 S maybe is. I tried to look at other fullranges than the Visaton but they are either much more expensive or not loud enough. Anyway, the R 10 S misses a bit sparkle at the top but it's just for testing and despite being very cheap, still measures surprisingly good, well, apart from a bit missing at the top end. You could do another project with the fullranges or sell them, not much losses there unlike buying planar tweeters.
Hi ICG, thanks for taking the time.
It's is probably difficult to catch-up on all the posts on that project thread I completely understand, so to resume there are already some other factors to take into account.
- Yes that's the room I have to work with at the moment and the speakers you see there are my Cabass, ScottG already informed me of this issue and advised me to make a stand with casters to get away from the wall and find a better listening position.
- At the moment the drivers included in the project that I own already are the following :
Woofer Gradient Acoustics AXP-08
Will be ported in an individual chamber
Full-range : Fostex FF85WK
In his own "tunnel" on top with an open back
At the moment the tweeters are an option not a requirement now the LS plan are out of the picture, I have not yet made frequency response measurements and tests maybe I don't even need a tweeter in the end but it's good to evaluate options just in case.
Would you mind elaborating what you mean on the following :
By narrowness do you mean the conduit on the cabinet? What is less than 30° ?
At the left of the left cabinet and right of the right cabinet?
- I take note of the Visaton, they are indeed very affordable, could be great to make some other tests in the future, I also had a look at the FRS 5 XWP - 8 Ohm inexpensive and have a very wide polar patern but not much T&S parameters advertised unfortunately.
- The next step in my workflow is to make an IEC baffle to do gated measurements, once that is done I will hopefully have what I need to proceed with the test cabinet.
Last edited:
I believe this is not to be taken out of the context of my project, he just mean that planar tweeter are expensive compared to to using full-range to test in this design, it's not a general caution to have.
No. It is a caution to buy them and being disappointed you can't use them (or the way you've planned it) and therefore having to take a financial loss in the one or other way.
I'm sorry, I did not read it all. Yes, it's very time consuming, not very reader-friendly to extract the informations and gather them all on my own. Maybe you could post a summary in that thread? That would make it much easier for others to catch up, post their opinions and suggestions.
Good. Dipoles need room and a free area behind them. They start to work good with ~1m distance to the next wall, minimum distance ~60cm, with shorter distances it's usually a lot worse than sealed.
Well, they got a very good efficiency but they got a very low linear excursion, not more than ~1,5-2 mm Xmax. I don't know how much dynamic you need but if you want it louder, they are better suited as a mid or low-mid for a 3 or 4-way speaker, they are more designed like fullrange drivers (think about the BG-20). If you want to louder, a lowcut at around the tuning frequency will help a lot to reduce the excursion.
Well, that's a big problem, the FF85WK got only 86.5dB/W(m), the Gradient is 93, so the bass driver is 6-7dB louder, in the bass it goes down to maybe 90, 89dB. If it is as close to the wall as in the drawing, it will stay at ~91-92dB. You can't get that to fit passive since you can't increase the level of the Fostex passively. No objection if you are using a DSP though.
Yes. That not only limits the backwards radiated angle, it brings the dipole opening far more back, the energy bilance within the room will be very awkward because the back radiated sound will be reflected by only a very small surface and angles, it will be strong from some angles and from others hardly anything is comming back. If you build it that way, it needs much more distance to the back to get that great, wide roomy impression of a dipole. In such a situation it is much better to build a open baffle above the bass or a [ shaped baffle/frame there (bass still ported).
No. That would give you the short reflections which make the location quite difficult and would muddy down a lot of details. It takes down not only the depth but also the left-right positoning. The reflection should go upwards, to the ceiling to compensate for the too short distance to the back. That's not ideal but a lot better with the speakers close to the wall than without it. And you can easily (and practically for free) experiment with it, change the angle and find out which way you like it. You cannot do that with the planned design. A test box for the woofer and a separate baffle and/or box for the FR keeps you a lot more options open and, while costing money for the mid-high 'enclosure' wood, will likely save you a lot in the end.
The direct sound is very important for the details, the short reflections let the impulse reach the ear within a few ms, making it difficult to differenciate between them, it's perceived as unprecise, the details dwindle for the ear. The longer reflections (with longer runtimes) are important for the room impression but they also add up to the frequency response. That explains i.e. why an uneven absorption is bad for the stage impression. I've seen a video which explains it quite well, I think it was from Toole. I'll post it if I find it again. He didn't talk about dipole, but most can be applied to them too though.
I've got 4 of the FRS 5 X and they are absolutely amazing for the money. Detailed, very homogene and good dispersion pattern, easy to work with regarding the crossover. They are pretty close to good dome tweeters, a tad less details maybe but with a much wider frequency range. I haven't seen better FR of that size, let alone that price.
The German magazine 'Klang & Ton' did a lot of speakers with them, I think maybe 8 (which include 3 versions of one speaker but with completely different crossover and concepts).
You don't need an IEC baffle. Or a IEC baffle to measure gated. That's counter productive because that will be nothing like in the speaker. The only use for that is if you want to get tons of drivers and build your own databank. It's much better to measure the drivers in the speaker because then you can work directly with it. You'd have to measure it in the speaker afterwards anyway. And with a gated measurement you'll not be able to get real results with a dipole because you're filtering out the half of the energy, the perceived response will be a lot different from the gated measurement. You need that measurement to build the basic crossover (or setting of the DSP) but you will have to work from there with the response in the room. Dipoles aren't easy. You'll likely have the best results if you measure gated and then find a good position for the speakers by listening, then equalize them with REW or similar programs. It doesn't look like you have much space to get different, alternative speaker positions, it may happen you can't find possible/good sounding positions in that room.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.