Comparison seems to suggest the shape close to the back of the driver could be affecting the highs and the depth might be affecting the lows.
[PS. I've moved the plots to the same level. The *.jpg image format is not really conducive to plots and line drawings. gif is one example of a type that is.]
[PS. I've moved the plots to the same level. The *.jpg image format is not really conducive to plots and line drawings. gif is one example of a type that is.]
Attachments
I tried to look for information if anybody has done any metalwork to the Beyma TPL 150 or 200. Like drilling some holes to the front/back metal bars to reduce reflection etc? Or maybe adding some material to the front face of the bars to make them more 'aerodynamic'.
Last edited:
Of course. But the magnet field does not disappear if you drill a couple of holes to the bars. Just little bit reduces sensitivity. So I guess nobody has done it. I think I will try it.
The holes at the front already weaken the magnet field. Expect a drop in efficiency if you drill big holes into it. If the holes are too small you'll have a resonator effect though. Offset the holes from the ones on the front so the magnet field stays halfway intact.
I have three of these so I can play around littlebit 😀
So first I attach the virgin frame to my cnc machine frontside up.
An externally hosted image should be here but it was not working when we last tested it.
Started to drill. Decided to put 3 medium size holes to each bar. Didn't want to go overboard.
An externally hosted image should be here but it was not working when we last tested it.
Then added a 45 degree chamfer to the slots. And why 45? Had this cutting tool already, so just used that one.
An externally hosted image should be here but it was not working when we last tested it.
Finished and wire brushed the top to clean the burrs. But the most laborious part was to clean the magnet gap inside. The magnets are super strong and what goes in wants to stay there. And when you get the big ones out there is a lof of small metal dust there which I removed by using Gorilla tape with strong glue. Took hours to do a perfect job. 🙄
An externally hosted image should be here but it was not working when we last tested it.
Time for some measurements. I don't plan to use this with the original horn in the speakers, but it was a good idea to put it on for sensitivity, etc. measurements for now.
An externally hosted image should be here but it was not working when we last tested it.
So how much do you think the sensitivity dropped?
Machismo,
Did you put the same 45 degree chamfer on the underside as well as the front of this plate?
This reminds me of the "Fazor" phase plug that Audeze started using in their planar magnetic headphones for similar benefits which should help to reduce interference.
However Audeze did this by adding the phase plug elements on top of their existing magnets, rather than cutting away the actual magnet material.
I use Beyma TPL150H tweeters myself. I'd be interested to see your before / after measurements for this mod.
Did you put the same 45 degree chamfer on the underside as well as the front of this plate?
This reminds me of the "Fazor" phase plug that Audeze started using in their planar magnetic headphones for similar benefits which should help to reduce interference.
However Audeze did this by adding the phase plug elements on top of their existing magnets, rather than cutting away the actual magnet material.
I use Beyma TPL150H tweeters myself. I'd be interested to see your before / after measurements for this mod.
Removing the front plate plates with the size of those Nd magnets would be a huge undertaking. I doubt they were removed, based on his description of fighting with the metal shavings.
Sensitivity dropped exactly 1 db, which is good. 2 db would have been a little too much for my taste.. Of course it depends on what you compare it to. If some change really is beneficial to the sound quality and you can't find a product from the market which would be similar, then you just must accept even a 3 db drop.
I didn't know how the changes will effect the frequency response, but I did pretty much guess, that making the front more open and uniform cannot make it sound more worse. And it didn't. I haven't had time to listen very much, but the first tests were promising. Also the measurements showed a surprise.
At 19.5 khz the responses are still even, but then my version extends all the way to almost 22 khz. Now, since I'm not 15 anymore, I can't hear this high, but it's an indication, that the playback might be more natural a bit lower also.
Btw that dip that you see on both responses at around 15 khz, it's not there anymore, when you take the microphone more far away from the wave guide.
Also it's good to note, that the back chamber was not attached during the measurement. I haven't used that in years because it's so bad. Instead, there was some damping material and behind that a pillow as you might see from the previous posts picture.
I didn't know how the changes will effect the frequency response, but I did pretty much guess, that making the front more open and uniform cannot make it sound more worse. And it didn't. I haven't had time to listen very much, but the first tests were promising. Also the measurements showed a surprise.
At 19.5 khz the responses are still even, but then my version extends all the way to almost 22 khz. Now, since I'm not 15 anymore, I can't hear this high, but it's an indication, that the playback might be more natural a bit lower also.
Btw that dip that you see on both responses at around 15 khz, it's not there anymore, when you take the microphone more far away from the wave guide.
Also it's good to note, that the back chamber was not attached during the measurement. I haven't used that in years because it's so bad. Instead, there was some damping material and behind that a pillow as you might see from the previous posts picture.
I suffered also, had a bit of hangover, had to rest. 🙄
No I didn't, I did just what is shown in the pictures. Don't want to even try to take this thing apart. Also, removing a lot of material from the inside near the membrane probably reduces sensitivity more than taking material from the outside. But I still might want to try to take a just a little bit in the future.
I haven't seen that one before but it's exactly what I had in mind when I started. Inside, in the magnet gap you can't do that but both front and the backside it would be possible. However the membrane is so deep under the metal that I thought it's best to make a cut instead. But on the back I could certainly see myself adding those fins.
Losing 1dB sensitivity is less than the 3dB you lose by removing Beyma's horn waveguide, and many folks are happy using the tweeter without the horn.
All that repeated drilling and routing on the CNC machine must have also heated up the thin metal strips of that front polar piece grille, and prolonged excess high temperatures can cause magnets to lose their magnetic charge. (Although for Neodymium magnets, the Curie temperature above which they would be permanently de-magnetised would be have to be 300°C to 400°C...!)
Nonetheless, if some sensitivity loss was due to de-magnetisation (from heat), it might be possible to actually restore it back again, if you could find somebody with a powerful pulse magnetiser machine to recharge your magnets for you, back to maximum magnetic saturation. They might even be stronger / better than stock afterwards!
Beyma's Nov 2010 AES paper talks a little on page 9-10 about different bevelling profiles for those grille strips, and they've published a comparison graph showing a slightly increased output (above 9kHz) for their bevelled version, which is interesting. Seems a shame this improvement didn't make it into the final product design.
Wonder if the trade-off between overall sensitivity drop vs possible increased output above 9kHz is beneficial - which of the two has the biggest net effect?
All that repeated drilling and routing on the CNC machine must have also heated up the thin metal strips of that front polar piece grille, and prolonged excess high temperatures can cause magnets to lose their magnetic charge. (Although for Neodymium magnets, the Curie temperature above which they would be permanently de-magnetised would be have to be 300°C to 400°C...!)
Nonetheless, if some sensitivity loss was due to de-magnetisation (from heat), it might be possible to actually restore it back again, if you could find somebody with a powerful pulse magnetiser machine to recharge your magnets for you, back to maximum magnetic saturation. They might even be stronger / better than stock afterwards!
Beyma's Nov 2010 AES paper talks a little on page 9-10 about different bevelling profiles for those grille strips, and they've published a comparison graph showing a slightly increased output (above 9kHz) for their bevelled version, which is interesting. Seems a shame this improvement didn't make it into the final product design.
Wonder if the trade-off between overall sensitivity drop vs possible increased output above 9kHz is beneficial - which of the two has the biggest net effect?
Last edited:
The metal didn't get hot at all, just a bit warm because, even if I used a CNC machine, I did the cuts manually without a program, which made the process so slow the heat had plenty of time to dissipate.
I think I found the paper you mentioned
https://www.beyma.com/wp-content/up...ributions_improvement_Pleated_Loudspeaker.pdf
Thanks.
I think I found the paper you mentioned
https://www.beyma.com/wp-content/up...ributions_improvement_Pleated_Loudspeaker.pdf
Thanks.
Have you had a chance to listen to music with the modded TPLs? Did you notice differences?
Yes I listened a little bit after the mod and it seemed less harsh. Then I realized I can listen just to the diaphragm itself when I place it next to the magnet, completely outside, in free air and turned the volume up to compensate the weaker magnetic field. That of course sounded more natural than behind the usual metal bars, but it still was not perfect to my ears. So it just might be that no matter how you drill you will not get past a certain point, because the quality of the diaphragm is the next limiting factor.
Last edited:
I live in two places, so I sent the tweeters for my other location for listening. And while there was plenty of cardboard, something sharp and apparently made of hardened steel had pierced the packet during shipping and hit the tweeter. Disaster! I thought. At least I had been smart enough to put the membrane to my backback instead inside the frame, because it would have been crushed on the sideways movement.
It was a lot of work to modify the tweeter so I decided to try to sort it out. And luckily my method worked perfectly. First I aligned it sideways,
then finally crushed the plates tightly together also.
And what was strange I decided to take a look of the other tweeter that didn't get hit in the shipping. It was also a bit off (on the left). So I repaired also that one. (on the right)
So it's possible these are not glued together at all from the factory. Maybe they rely on the really strong magnetic field. In any case, I didn't want them to move so easily in the future, so I applied some very thin and free flowing superglue which got sucked in to the gaps on all four seams.
It was a lot of work to modify the tweeter so I decided to try to sort it out. And luckily my method worked perfectly. First I aligned it sideways,
then finally crushed the plates tightly together also.
And what was strange I decided to take a look of the other tweeter that didn't get hit in the shipping. It was also a bit off (on the left). So I repaired also that one. (on the right)
So it's possible these are not glued together at all from the factory. Maybe they rely on the really strong magnetic field. In any case, I didn't want them to move so easily in the future, so I applied some very thin and free flowing superglue which got sucked in to the gaps on all four seams.