After studying my economy I've realised that I'll have to live with my Mirage FRX-3 speakers for quite a while. But I'm not quite satisfied with them. The higher frequencies har restrained - either by the metal tweeter, the baffle (?) or by the crossover. Two of these factors I cannot change.
Measurements using a WM60AY based microphone show reasonable linearity - but without a calibration file. The WM60 is not particular linear above 10khz. The sensivity is rising, thus the response of the FRX-3 is NOT linear above 10khz, but is falling instead. Details are lost, soundstage is not presented properly... maybe a way of controling an often aggresive nature of metal domes?
Anyway, I finally decided to take one of the speakers apart. As it seems, several potential improvements are available. I've got some suggestions, but I would like to hear your opinion. Unfortunately I don't have a digital camera, and I cannot get my hands onto one in the next two weeks, so I'll have to describe, what I see:
- The front baffle is made of some strange sort of plastic. Rather thin, I'd say. Definetly not somewhat I would pay for today, but at that I didn't give it a thought.
- The back of the speakers is likewise made of the same thin moulded plastic. Shame on me. These two issues cannot be corrected. No improvement available.
- The crossover is as expected in a 250$ a piece speaker (didn't pay more than 150$): Ordinary wire-wound resistors, nothing unsual here, air-core inductor for the tweeter and metal core for the bass/midrange and electrolytics capacitors (!) for both drivers. Obviously there's room for some improvement here. The inductors are of a poor quality. The wire is extraordinarily thin. And the capacitors - especially the ones for the tweeter section - should not be electrolytics. I'll post a schematic whenever I get around to it.
- The speakers are heavily stuffed in the upper part around the port. The stuffing may even reduce the air flow to the port. Not good. In the lower part there's no linning on the sides. Not good. At the back there's reasonably thick stuffing.
- NO bracing at all! Mirage claims in the user manual that the speakers are well-braced. Certainly not the truth. And the walls are rather thin too.
The obvious improvements are related to the crossover. The electrolytics will be changed to polypropylene except for the one in the impedance correction filter. Both inductors could be changed to better types - preferably air-core, but that depends on the values. I would want to reduce the stuffing around the port opening and instead put some on the sidewalls in the lower part of the box.
I'm thinking about putting "lead rubber"* (translation from danish - don't know what it's called in english) and linning the side walls and the bottom with some damping material too. Bracing is not easy to do, and it will reduce the box volume. Maybe linning on the sidewalls will provide enough "extra volume"? I wouldn't want to count on it, though.
Any suggestion is welcome. The speaker can be found at www.miragespeakers.com in the archive.
* UPDATE: I think it's something like bitumen with polymer. I've just decided not to use it due to serious health risks. Probably fine, if the enclosure is sealed, but not ported. Any alternative without hazardous properties?
Measurements using a WM60AY based microphone show reasonable linearity - but without a calibration file. The WM60 is not particular linear above 10khz. The sensivity is rising, thus the response of the FRX-3 is NOT linear above 10khz, but is falling instead. Details are lost, soundstage is not presented properly... maybe a way of controling an often aggresive nature of metal domes?
Anyway, I finally decided to take one of the speakers apart. As it seems, several potential improvements are available. I've got some suggestions, but I would like to hear your opinion. Unfortunately I don't have a digital camera, and I cannot get my hands onto one in the next two weeks, so I'll have to describe, what I see:
- The front baffle is made of some strange sort of plastic. Rather thin, I'd say. Definetly not somewhat I would pay for today, but at that I didn't give it a thought.
- The back of the speakers is likewise made of the same thin moulded plastic. Shame on me. These two issues cannot be corrected. No improvement available.
- The crossover is as expected in a 250$ a piece speaker (didn't pay more than 150$): Ordinary wire-wound resistors, nothing unsual here, air-core inductor for the tweeter and metal core for the bass/midrange and electrolytics capacitors (!) for both drivers. Obviously there's room for some improvement here. The inductors are of a poor quality. The wire is extraordinarily thin. And the capacitors - especially the ones for the tweeter section - should not be electrolytics. I'll post a schematic whenever I get around to it.
- The speakers are heavily stuffed in the upper part around the port. The stuffing may even reduce the air flow to the port. Not good. In the lower part there's no linning on the sides. Not good. At the back there's reasonably thick stuffing.
- NO bracing at all! Mirage claims in the user manual that the speakers are well-braced. Certainly not the truth. And the walls are rather thin too.
The obvious improvements are related to the crossover. The electrolytics will be changed to polypropylene except for the one in the impedance correction filter. Both inductors could be changed to better types - preferably air-core, but that depends on the values. I would want to reduce the stuffing around the port opening and instead put some on the sidewalls in the lower part of the box.
I'm thinking about putting "lead rubber"* (translation from danish - don't know what it's called in english) and linning the side walls and the bottom with some damping material too. Bracing is not easy to do, and it will reduce the box volume. Maybe linning on the sidewalls will provide enough "extra volume"? I wouldn't want to count on it, though.
Any suggestion is welcome. The speaker can be found at www.miragespeakers.com in the archive.
* UPDATE: I think it's something like bitumen with polymer. I've just decided not to use it due to serious health risks. Probably fine, if the enclosure is sealed, but not ported. Any alternative without hazardous properties?
Sealed IMO would not be fine with the rubber lead either because of the solvents used. They could cause the glues or any other material used inside the box to disintergrate
Id say the biggest difference will be made when you chage the electrolytic caps to good poly ones especially in the tweeter crossover.
With regards to the inductors id just replace them with thicker wire varients, but they dont need to be air core, yes they do sound better but with $250 speakers i dont think its worth it. Just some good gauge iron or ferrite cores. I'd leave the tweeter air core as it is. Iron core are better then ferrite though.
Id also point out that lots of crossover components together cost lots, so maybe ull be better off just replacing the tweeter caps and leaving at that, and save until you can build better speakers altogether.
Id say the biggest difference will be made when you chage the electrolytic caps to good poly ones especially in the tweeter crossover.
With regards to the inductors id just replace them with thicker wire varients, but they dont need to be air core, yes they do sound better but with $250 speakers i dont think its worth it. Just some good gauge iron or ferrite cores. I'd leave the tweeter air core as it is. Iron core are better then ferrite though.
Id also point out that lots of crossover components together cost lots, so maybe ull be better off just replacing the tweeter caps and leaving at that, and save until you can build better speakers altogether.
Yes, I know crossover components are expensive, so I might very well leave most of it as it is.
My description may not have been clear enough. The speaker is ported NOT sealed. And lead rubber (or lead-bitumen?) is really not an option. It's supposed to diffucult to get, since it's rather hazardous on long term basis.
People in Denmark often call bitumen polymer lead rubber, which it is not, as I've found out. There shouldn't be any problem in using bitumen polymer.
How about the lack of acoustic damping material on the sides?
My description may not have been clear enough. The speaker is ported NOT sealed. And lead rubber (or lead-bitumen?) is really not an option. It's supposed to diffucult to get, since it's rather hazardous on long term basis.
People in Denmark often call bitumen polymer lead rubber, which it is not, as I've found out. There shouldn't be any problem in using bitumen polymer.
How about the lack of acoustic damping material on the sides?
Improving a standard design, cheap or not, is fun and most times worthwile.
The changes in filter are obvious, important are coupling caps, decoupling not that big an issue.
A simple cross brace, one rod from top to bottom and one from side to side joined at crossing point (not symetrical higer or lower than middle) works well without changing internal volume that much. For the plastic back use polyuretane potting compound, have the box level on its back and pour in one 250g pack and let cure. The baffle can be greatly improve by attaching rods, did this on a friend B&O speakers, cut 2-3 pieces of solid wood rods same widht as inside box. Find suitable places above and below woofer and drill holes from front.
Use a good windowsealing rubber list on the rods, and atach from inside with rubber towards baffle and screw down tight with screws from outside.
The changes in filter are obvious, important are coupling caps, decoupling not that big an issue.
A simple cross brace, one rod from top to bottom and one from side to side joined at crossing point (not symetrical higer or lower than middle) works well without changing internal volume that much. For the plastic back use polyuretane potting compound, have the box level on its back and pour in one 250g pack and let cure. The baffle can be greatly improve by attaching rods, did this on a friend B&O speakers, cut 2-3 pieces of solid wood rods same widht as inside box. Find suitable places above and below woofer and drill holes from front.
Use a good windowsealing rubber list on the rods, and atach from inside with rubber towards baffle and screw down tight with screws from outside.
Suitable compounds from 3M or Electrolube will reach about 50 degrees centigrade while curing. You wont be able to take it away
after curing. Also !!! Do not drill, file or perform any abrasive work on polyurethane as the dust is NOT healthy.
Instead of the potting You can do the same rod atachment on the back as on the front. Holes will be there.. so not realy reversible.
Swapping out crossover components might not help as much as you expect. For instance, the crossover design may take into account the resistance of the skinny wire inductors. Swapping in high quality low resistance inductors may do more harm than good. I can't think of a problem with swapping out the capitors (but maybe somebody else can). I would also be wary of changing the stuffing, again because it could affect the effective box volume and the design response.
The one thing I think you could do with low risk is try to fix the box to tame resonances. You could even go as far as building a whole new properly braced box. I would not change the baffle dimensions as this could affect the response, in particular the baffle step correction.
The one thing I think you could do with low risk is try to fix the box to tame resonances. You could even go as far as building a whole new properly braced box. I would not change the baffle dimensions as this could affect the response, in particular the baffle step correction.
I'll just advise going over the top is not a good idea.
Keep your inductors, replace the capacitors.
Industrial grade vinyl floor tiles are good for damping in layers.
If the front and back have lots of ribbing "filling the holes" is not a good idea IMO.
The mass stiffness ratio will decrease. Any filling should have high damping properties.
Move the stuffing away from the port.
A stuffed area does not need foam lining, conversely
an area with a port and no stuffing does, ideally.
Detuning the port by foam lining will give you an extra half an
octave of real bass extension with good placement in real rooms.
Modifications to the drive units are possible, cheaper the drive
units generally the more effective the drive unit mods are.
sreten.
Keep your inductors, replace the capacitors.
Industrial grade vinyl floor tiles are good for damping in layers.
If the front and back have lots of ribbing "filling the holes" is not a good idea IMO.
The mass stiffness ratio will decrease. Any filling should have high damping properties.
Move the stuffing away from the port.
A stuffed area does not need foam lining, conversely
an area with a port and no stuffing does, ideally.
Detuning the port by foam lining will give you an extra half an
octave of real bass extension with good placement in real rooms.
Modifications to the drive units are possible, cheaper the drive
units generally the more effective the drive unit mods are.
sreten.
Ok, thanks. I'll just change the capacitors, then. I'll try to improve the stiffness of the box and the damping.
Sreten, when you mention driver modifications - were you thinking something specific? Are there any easy mods (little risk) with a good chance of a noticeable improvement? The parameters must not be changed in any way.
Sreten, when you mention driver modifications - were you thinking something specific? Are there any easy mods (little risk) with a good chance of a noticeable improvement? The parameters must not be changed in any way.
Funny you should post this. I was groing impatient with how long it was taking me to complete my own speaker project, so I recently opened up my cheap ($200-$300 per pair) pioneer tower speakers and was relatively shocked by what I found. These are S-H453F-K's if you want to look them up at www.pioneerelectronics.com . Firstly, the cabinet, which virtually rings when you knock on it, is made of the crappiest, low-density particle board I've ever seen. There was no bracing and absolutely no damping material inside. The woofer was at the very top, the passive radiator at the bottom, and the tweeter and dummy midrange in between. The "midrange" is a two-inch polymer cone with no surround (relying only on the flexibility of the diaphram to move, and measured a shocking DCR of 27Ohms! So I'm guessing it's about a 32-ohm speaker with what could barely pass as a refridgerator magnet on the back. playing it by itself (with the .22uF series capacitor attached to it) revealed that it did indeed make sound, but requiring full volume on my gainclone, and the sound that did escape was quite objectionable noise, which I attribute partially to the harshness of the overall sound.
The rest of the crossover consisted of a 4uF NPE cap in series with the tweeter, no low-pass filter (which is ok in theory, but I can't measure the response to say for certain), and the most puzzing part, a 1.3uF ceramic cap in series with the whole thing. I'm certainly no electrical engineer, but from various tables and such, I think that with an 8-ohm load, that should make something around a 15k+ high pass filter, which seems to be a paradox since I did actually hear sound from the speakers. Anyway, on to the free modifications I made...
The only damping material I could line the cabinet walls with was polyfil (which I actually scavenged from an old pillow, lol). This actually turned out to be more of a stuffing than a lining, though. I completely disconnected the midrange, and removed the 1.3uF mystery cap altogether and replaced the 22ga(!!!) wire connecting the woofer with 14-ga monster cable soldered on both ends. I also moved the midwoofer to the bottom, and put the PR on the top. The result was quite surprising. The bass is much more controlled, and goes much deeper. I think I actually turned it into a transmission line of sorts by adding the stuffing and moving the driver and passive radiator. The harshness is gone, and vocals, pianos, etc. sound much more real. I plan to add an active baffle step compensator in front of my gainclone to make up for the lack of it in the speaker. But, from my experience with these crappy speakers, it seems possible to make significant improvements to cheap commercial speakers without spending any money at all.
The rest of the crossover consisted of a 4uF NPE cap in series with the tweeter, no low-pass filter (which is ok in theory, but I can't measure the response to say for certain), and the most puzzing part, a 1.3uF ceramic cap in series with the whole thing. I'm certainly no electrical engineer, but from various tables and such, I think that with an 8-ohm load, that should make something around a 15k+ high pass filter, which seems to be a paradox since I did actually hear sound from the speakers. Anyway, on to the free modifications I made...
The only damping material I could line the cabinet walls with was polyfil (which I actually scavenged from an old pillow, lol). This actually turned out to be more of a stuffing than a lining, though. I completely disconnected the midrange, and removed the 1.3uF mystery cap altogether and replaced the 22ga(!!!) wire connecting the woofer with 14-ga monster cable soldered on both ends. I also moved the midwoofer to the bottom, and put the PR on the top. The result was quite surprising. The bass is much more controlled, and goes much deeper. I think I actually turned it into a transmission line of sorts by adding the stuffing and moving the driver and passive radiator. The harshness is gone, and vocals, pianos, etc. sound much more real. I plan to add an active baffle step compensator in front of my gainclone to make up for the lack of it in the speaker. But, from my experience with these crappy speakers, it seems possible to make significant improvements to cheap commercial speakers without spending any money at all.
wow my cheap no name speakers sound better built than that..
Guess you cant trust brand names.
They were a surround setup I bought at A&B a while ago (5.0).
Actually the towers are made nicely, the surrounds and center are crap. Center has a very small volume, and cant go deep at all. Electrolytics on the tweeters thats it..
Towers are braced very well. Fair bit of stuffing in the top and bottom sections. Crossover is a 3 way with some okay parts, but not designed very well (inductor is too low of a rating).. But no idea if the speakers are any good, a fabric dome, some mids, and a sub..
I can take some pictures if you want of the crossover or speakers. Planning to sell them and build new ones though.
brand is "SEL"
I like your modifications though, basically get better sound for free
Guess you cant trust brand names.
They were a surround setup I bought at A&B a while ago (5.0).
Actually the towers are made nicely, the surrounds and center are crap. Center has a very small volume, and cant go deep at all. Electrolytics on the tweeters thats it..
Towers are braced very well. Fair bit of stuffing in the top and bottom sections. Crossover is a 3 way with some okay parts, but not designed very well (inductor is too low of a rating).. But no idea if the speakers are any good, a fabric dome, some mids, and a sub..
I can take some pictures if you want of the crossover or speakers. Planning to sell them and build new ones though.
brand is "SEL"
I like your modifications though, basically get better sound for free
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.