Hello everyone,
With the help of many people here I have finished building a few subs over the past years and I thought it was time to graduate to the higher octaves. Instead of just jumping in and building a well documented DIY speaker from for example Troels I thought it might be better to first understand how all the measurements work, how they correlate (and don't correlate) to what I hear and how all the pieces fit together.
Since I own a pair of monitor audio silver S2, I like building cabinets and they are notorious for having a very simple unbraced cabinet, why not start there?
I intend to do some measurements and try to figure out what they mean. Hopefully you guys can keep me from making too many mistakes or misinterpreting the results. I am willing to perform any measurements within the capabilities of my Umik, just ask. I'll try and get some soundclips from the different cabs later.
I Like the speakers a lot better with the foam plug in the bassport, so the replacement cabinets do not have a port. All measurements for the original cab were taken with the plug in the bassport.
review with some measurements from stereophile Monitor Audio Silver S2 loudspeaker | Stereophile.com
So I made 2 designs, the first is a more standard enclosure with internal dimensions according to the golden ratio, triple bracing and double (36mm) front side and rear walls. Front baffle dimensions identical to the original. I glued 1cm felt to all walls except front baffle and put in some polyester mat at the top, back and bottom.
The second is an attempt to build something resembling the nautaloss enclosure from XRK in MDF. http://www.diyaudio.com/forums/full-range/247598-nautaloss-ref-monitor.html
I thought about using foamcore or that other foam that is getting popular with the appartment builders but I have a basement equipped to make these things in MDF and i can get high quality 0.72g/cm³ 18mm MDF for 7€/m² (they even cut the larger pieces for me at 0.2mm precision for almost no money and they are only 500m away)
I glued 1cm felt to all inside walls for the first 3/4 turn and only to the spiral arms and not the sidewalls for the subsequent turns. Polyester mat was added to all internal walls and an extra piece in the mouth behind the driver.
The crossover goes in a sealed cavity in the lower back, the tweeter goes in the sealed cavity in the front. And yes, routing those ridiculous faceplate shapes was a bitch.
With the help of many people here I have finished building a few subs over the past years and I thought it was time to graduate to the higher octaves. Instead of just jumping in and building a well documented DIY speaker from for example Troels I thought it might be better to first understand how all the measurements work, how they correlate (and don't correlate) to what I hear and how all the pieces fit together.
Since I own a pair of monitor audio silver S2, I like building cabinets and they are notorious for having a very simple unbraced cabinet, why not start there?
I intend to do some measurements and try to figure out what they mean. Hopefully you guys can keep me from making too many mistakes or misinterpreting the results. I am willing to perform any measurements within the capabilities of my Umik, just ask. I'll try and get some soundclips from the different cabs later.
I Like the speakers a lot better with the foam plug in the bassport, so the replacement cabinets do not have a port. All measurements for the original cab were taken with the plug in the bassport.
review with some measurements from stereophile Monitor Audio Silver S2 loudspeaker | Stereophile.com
An externally hosted image should be here but it was not working when we last tested it.
So I made 2 designs, the first is a more standard enclosure with internal dimensions according to the golden ratio, triple bracing and double (36mm) front side and rear walls. Front baffle dimensions identical to the original. I glued 1cm felt to all walls except front baffle and put in some polyester mat at the top, back and bottom.
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
The second is an attempt to build something resembling the nautaloss enclosure from XRK in MDF. http://www.diyaudio.com/forums/full-range/247598-nautaloss-ref-monitor.html
I thought about using foamcore or that other foam that is getting popular with the appartment builders but I have a basement equipped to make these things in MDF and i can get high quality 0.72g/cm³ 18mm MDF for 7€/m² (they even cut the larger pieces for me at 0.2mm precision for almost no money and they are only 500m away)
I glued 1cm felt to all inside walls for the first 3/4 turn and only to the spiral arms and not the sidewalls for the subsequent turns. Polyester mat was added to all internal walls and an extra piece in the mouth behind the driver.
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
The crossover goes in a sealed cavity in the lower back, the tweeter goes in the sealed cavity in the front. And yes, routing those ridiculous faceplate shapes was a bitch.
Last edited:
Listening impressions:
Original VS new boxes: Oh my god, I assumed I would be able to hear a difference but this big? Everything is more defined, organs actually sound like an organ instead of a flat tone, I can hear the initial air rushing in. It is so easy to hear all these details I didn't even know were there.
Braced cabinet VS Spiral: Hmm this is very difficult for me, there is some difference but it is very difficult to hear and pinpoint. After going through more then 40 CDs and downloading some extra Flacs I only found one song (Koyaanisqatsi - Phillip Glass) were I'm sure i would be able to identify which is which in a blind test. At some point the voice gets some backup by a choir, this is barely audible with the braced cabinet but very clear with the spiral cabinet.
I have invited some friends to come over on thursday for a blind ABX listening test. 2 are classical musicians (1 violin, the other piano and guitar) 1 is a metal fan
and the other still uses an 80s all in 1 stereotower system with the speakers that came with the system... So quite a divers set of ears.
Original VS new boxes: Oh my god, I assumed I would be able to hear a difference but this big? Everything is more defined, organs actually sound like an organ instead of a flat tone, I can hear the initial air rushing in. It is so easy to hear all these details I didn't even know were there.
Braced cabinet VS Spiral: Hmm this is very difficult for me, there is some difference but it is very difficult to hear and pinpoint. After going through more then 40 CDs and downloading some extra Flacs I only found one song (Koyaanisqatsi - Phillip Glass) were I'm sure i would be able to identify which is which in a blind test. At some point the voice gets some backup by a choir, this is barely audible with the braced cabinet but very clear with the spiral cabinet.
I have invited some friends to come over on thursday for a blind ABX listening test. 2 are classical musicians (1 violin, the other piano and guitar) 1 is a metal fan
and the other still uses an 80s all in 1 stereotower system with the speakers that came with the system... So quite a divers set of ears.Now, onto the measurements. I measure impedance using REW and a homemade impedance jig. I use a Umic and REW for the frequency respons measurements.
The impedance trace from the original box looks quite good. There is a minor bump at 240Hz and a little wiggle at 1300.
Internal dimensions for the box: 166mmx234mmx324mm
If we then look at the braced box, driver resonance is a bit lower so the cabinet volume including all the damping is a bit larger then the original. The minor bump at 240 becomes a lot larger and shifts to 200, there appears to be a new minor bump at 500, no change at 1300.
Internal dimensions for the box: 148mmx238mmx384mm
For the spiral box, the driver resonance is even lower, the bump around 200 gets a lot larger and seems to have split in 2. Nothing at 500 same as original cabinet. At 1300 the little wiggle is gone and the peak has shifted a bit higher.
Internal dimensions for the box: mouth 185mmx148mm, golden ratio spiral so every 1/4 turn the opening reduces with about 1.62. Opening sizes every 1/4 turn: 185, 135, 84, 50, 30, 20, 16. The path length through the middle of the spiral is about 87cm.
The impedance trace from the original box looks quite good. There is a minor bump at 240Hz and a little wiggle at 1300.
Internal dimensions for the box: 166mmx234mmx324mm
If we then look at the braced box, driver resonance is a bit lower so the cabinet volume including all the damping is a bit larger then the original. The minor bump at 240 becomes a lot larger and shifts to 200, there appears to be a new minor bump at 500, no change at 1300.
Internal dimensions for the box: 148mmx238mmx384mm
For the spiral box, the driver resonance is even lower, the bump around 200 gets a lot larger and seems to have split in 2. Nothing at 500 same as original cabinet. At 1300 the little wiggle is gone and the peak has shifted a bit higher.
Internal dimensions for the box: mouth 185mmx148mm, golden ratio spiral so every 1/4 turn the opening reduces with about 1.62. Opening sizes every 1/4 turn: 185, 135, 84, 50, 30, 20, 16. The path length through the middle of the spiral is about 87cm.
Did you try different materials for damping? I'd suggest glass fibre insulation if it's a sealed enclosure. Cover the sides (and braces) with wool felt.
Can't figure out what you used on the sides but it looks like something stiffer than a felt.
I never liked those blanket type damping materials. Gave that away to my brother in law who uses it as a filter material for his pond.
Can't figure out what you used on the sides but it looks like something stiffer than a felt.
I never liked those blanket type damping materials. Gave that away to my brother in law who uses it as a filter material for his pond.
And the frequency respons measurements:
Measured at 0.5m normal listening volume
Measured at 1m normal listening volume
Measured at 1m distance very loud (i had to put my fingers in my ears for this one sitting at 2.5m)
An externally hosted image should be here but it was not working when we last tested it.
Measured at 0.5m normal listening volume
An externally hosted image should be here but it was not working when we last tested it.
Measured at 1m normal listening volume
An externally hosted image should be here but it was not working when we last tested it.
Measured at 1m distance very loud (i had to put my fingers in my ears for this one sitting at 2.5m)
An externally hosted image should be here but it was not working when we last tested it.
I used felt mats from Kwantum, cheapest i could find at 4.5€/m² for 5mm thickness https://www.kwantum.be/vloer-ondervloeren-isolatie-ondertapijt-vilt-snijvast-0130600 and some polyester mats from conrad. Dempingsmateriaal in de Conrad online shop | 333999
It is impossible to get farther then 3/4 turn in the spiral once it is glued and with the removable baffle i didn't also want a removable sidepanel. So I can't change damping further in te spiral.
Last edited:
I wasn't happy with the legibility of the original graphs so for future posts color coding will be:
Red for the original cabinet
Blue for the new cabinet
Green for the spiral cabinet
Dark 1m loud listening volume
Medium 0.5m normal listening volume
Light 1m normal listening volume
Impedance measurements:
Frequency respons measurements:
Red for the original cabinet
Blue for the new cabinet
Green for the spiral cabinet
Dark 1m loud listening volume
Medium 0.5m normal listening volume
Light 1m normal listening volume
Impedance measurements:
Frequency respons measurements:
Distortion measurements:
First the original cab:
quiet (I don't have a calibrated dB meter and I'm not sure how far the Umik is off) quiet in this case for testing THD. Normal listening volume.
Ooh, looks pretty
normal: what i would call loud listening volume.
at 1m
at 50cm
Loud: put my fingers in my ears at 2.5m away while testing, loud.
Then the new cab
Normal 1m
Normal 0.5m
Loud 1m
And finally the spiral cab:
Normal 1m
Normal 50cm
Loud 1m
First the original cab:
quiet (I don't have a calibrated dB meter and I'm not sure how far the Umik is off) quiet in this case for testing THD. Normal listening volume.
Ooh, looks pretty
normal: what i would call loud listening volume.
at 1m
at 50cm
Loud: put my fingers in my ears at 2.5m away while testing, loud.
Then the new cab
Normal 1m
Normal 0.5m
Loud 1m
And finally the spiral cab:
Normal 1m
Normal 50cm
Loud 1m
@ Wesayso
I'm happy to try different things and modify the enclosures further, just stating the origin of the damping materials used now and limitations of the spiral enclosure for modifications.
I did stuff the damping mats tighter when going in the small sections of the spiral. Trying to explain as best as i can going from the woofer: the first 3/4 turn is accessible (last 3/4 part doable but difficult due to bracing), the next 1/4 turn is not accessible, the next and final full turn is allready "full" with damping material. So I can still modify most of the missing bits.
I still have nightmares about the horrible glass wool insulation i used to insulate the attic at my moms place, would rockwool be a suitable replacement?
Do you have a good (cheap) source for wool felt cause all the stuff I found was super expensive.
I'm happy to try different things and modify the enclosures further, just stating the origin of the damping materials used now and limitations of the spiral enclosure for modifications.
I did stuff the damping mats tighter when going in the small sections of the spiral. Trying to explain as best as i can going from the woofer: the first 3/4 turn is accessible (last 3/4 part doable but difficult due to bracing), the next 1/4 turn is not accessible, the next and final full turn is allready "full" with damping material. So I can still modify most of the missing bits.
I still have nightmares about the horrible glass wool insulation i used to insulate the attic at my moms place, would rockwool be a suitable replacement?
Do you have a good (cheap) source for wool felt cause all the stuff I found was super expensive.
With this experiment i want to find out how much is too much when it comes to enclosures and to get as much experience with listening and measuring as possible. I don't mind building complex enclosures (the spiral takes 5 times longer to build then the braced box) but then it needs to be better or why bother.
The extended plan is:
1) to get a minidsp and go active to replace the crossover
2) get a pair of SS 10F/8414g10 for a 3-way to get rid of the nasty distortion between 1k and 3k
2.5) maybe do some other cool stuff with the 10F depending on cash before i
3) buy a pair of R2604/832000 and 22W/8534G00 (or RS225-4 for sealed crossed to sub) and make an active Troels SS 3-way classic sort off.
I'm hoping to document as much of it as possible on this site seeing how you, XRK and so many others inspired me to finally take the plunge and go build my own speakers. These different cabinets seemed like a cheap fun way to learn a lot and might be interesting to other beginners, at least for me it was difficult to find some good data on the effect of the quality of the cabinet.
The extended plan is:
1) to get a minidsp and go active to replace the crossover
2) get a pair of SS 10F/8414g10 for a 3-way to get rid of the nasty distortion between 1k and 3k
2.5) maybe do some other cool stuff with the 10F depending on cash before i
3) buy a pair of R2604/832000 and 22W/8534G00 (or RS225-4 for sealed crossed to sub) and make an active Troels SS 3-way classic sort off.
I'm hoping to document as much of it as possible on this site seeing how you, XRK and so many others inspired me to finally take the plunge and go build my own speakers. These different cabinets seemed like a cheap fun way to learn a lot and might be interesting to other beginners, at least for me it was difficult to find some good data on the effect of the quality of the cabinet.
I have been trying to record the speakers using the umik-1 and audacity but I only get horrible results. Recording was done in the same way as the frequnecy sweeps (placement etc)
View attachment Trifonic new box.asc
I ordered some polyester mats from an online pillow shop guess they really do make it for use as filter. I had no idea. It does feel more rugged then the conrad stuff less and thicker fibres.
View attachment Trifonic new box.asc
I ordered some polyester mats from an online pillow shop guess they really do make it for use as filter. I had no idea. It does feel more rugged then the conrad stuff less and thicker fibres.
I guess you could try Rockwool. No problem... Since you're relatively close you might try ordering wool felt from here:
Wolvilt 15 m2 op rol 5 mm dik - Te koop | Tweedehands.nl
It's an old add but has the contact info. It definitely is real sheep wool, the smell confirms it
.I didn't move the stand or any other furniture between these and yesterdays measurements in order to keep them as comparable as possible.
Impedance measurements of the spiral enclosure with added poly (2 pieces of 15x80cm folded in 3 in the first half turn)
A resonance shift due to the larger apparent enclosure and the peak at 200 has lowered in amplitude but becomes a lot wider. The shift and weird shape confirms my suspicion that it could be a 1/2 wave from the spiral length. 340m/s divided by 0.87m is 390Hz. In any case it wasn't possible to build a longer spiral out of MDF to get the halfwave resonance out of the passband. (3m long spiral with only 12liter internal volume...) When comparing the braced cabinet to the spiral while listening I did not notice any weird stuff around 200Hz but it could be possible that the difference I heard with phillip glass is more due to a drop in respons caused by the halfwave then an increase in detail.
Respons measurement of spiral enclosure with added poly:
Loud 1m
Normal 1m
Normal 0.5m
Seems that the suckout around 190Hz has become a bit smaller
Impedance measurements of the spiral enclosure with added poly (2 pieces of 15x80cm folded in 3 in the first half turn)
A resonance shift due to the larger apparent enclosure and the peak at 200 has lowered in amplitude but becomes a lot wider. The shift and weird shape confirms my suspicion that it could be a 1/2 wave from the spiral length. 340m/s divided by 0.87m is 390Hz. In any case it wasn't possible to build a longer spiral out of MDF to get the halfwave resonance out of the passband. (3m long spiral with only 12liter internal volume...) When comparing the braced cabinet to the spiral while listening I did not notice any weird stuff around 200Hz but it could be possible that the difference I heard with phillip glass is more due to a drop in respons caused by the halfwave then an increase in detail.
Respons measurement of spiral enclosure with added poly:
Loud 1m
Normal 1m
Normal 0.5m
Seems that the suckout around 190Hz has become a bit smaller
Last edited:
Further adding polyester mat to the enclosure from 1/2 turn till 3/4 turn (top back of the enclosure and the deepest i can still get stuff in and maybe out) 2 sheets of 13x70cm folded in 4 were added (quite densely packed)
The peaks have dropped further, I think it looks quite good now actually.
The resonance peak didn't lower any further but it did get bigger. Can anyone explain why this might be? I always thought that the peaks would get smaller and shift lower when adding stuffing like it had been doing up till now.
The peaks have dropped further, I think it looks quite good now actually.
The resonance peak didn't lower any further but it did get bigger. Can anyone explain why this might be? I always thought that the peaks would get smaller and shift lower when adding stuffing like it had been doing up till now.
After further listening while switching between the spiral enclosure and the braced one I came across a few other songs where things seemed to be a bit clearer with the spiral enclosure. It's difficult to put in words for me but it happened with 4 different songs that when I switched to the spiral enclosure I heard something new I didn't know was there before. The distinct different claps in the clapping in "Dog days are over - Florence and the machine", venue sounds in a live recording from "your latest trick - Dire straits" etc. Once identified I could also hear it with the braced enclosure, but it remained more clear with the spiral.
After the ABX testing tomorrow I'll know more.
After the ABX testing tomorrow I'll know more.
I also added dense stuffing to the braced cabinet:
As expected the resonance lowers in amplitude and shifts lower in frequency. The little peak at 200Hz stays about the same but shifts a bit lower in frequency.
I have no idea what that peak is for the braced cabinet, I can't link it to any internal dimensions and it isn't present with the original cabinet with similar internal dimensions.
As expected the resonance lowers in amplitude and shifts lower in frequency. The little peak at 200Hz stays about the same but shifts a bit lower in frequency.
I have no idea what that peak is for the braced cabinet, I can't link it to any internal dimensions and it isn't present with the original cabinet with similar internal dimensions.
ABX test: 10 different 30 second mono sound clips, 1 test per clip. 1 speaker upside down on top of the other. I put some blankets on the floor to minimise reflection. Listening distance 4.5 meters.
Original cabinet vs braced cabinet, 4 listeners.
6/7/7/1 correct out of 10
It's possible the violin player (guessed 1 correct out of 10) guessed wrong on purpuse to mess with me. He had this big wide grinn on his face when I read the results and I've known him for almost 20 years.
In any case, the results are what they are and we had a lot of fun.
We proceeded to compare the braced cabinet to the spiral but didn't do an ABX test considering the results of the first so I put the speakers next to each other. No difference reported at 4.5 meters. A small hard to place difference at 2 meters listening distance. 1 listener (violin player) specified a difference in the treble, with the spiral being a bit more detailed. He was very happy when I showed him the impedance plot afterwards with the hump at 1.3k dissapearing.
Original cabinet vs braced cabinet, 4 listeners.
6/7/7/1 correct out of 10
It's possible the violin player (guessed 1 correct out of 10) guessed wrong on purpuse to mess with me. He had this big wide grinn on his face when I read the results and I've known him for almost 20 years.
In any case, the results are what they are and we had a lot of fun.
We proceeded to compare the braced cabinet to the spiral but didn't do an ABX test considering the results of the first so I put the speakers next to each other. No difference reported at 4.5 meters. A small hard to place difference at 2 meters listening distance. 1 listener (violin player) specified a difference in the treble, with the spiral being a bit more detailed. He was very happy when I showed him the impedance plot afterwards with the hump at 1.3k dissapearing.
@ Jay: it is what it is.
@ Cocoapuffs: glad you like it
I changed the wiring for the spiral enclosure, the crossover was a bit too far away and the woofer wire was kinda stretched. So I got some silver coated copper wire to change it. I also completely sealed the wire hole for the tweeter and crossover cavity with silicone afterwards.
The large impedance peak shifts a bit to the right indicating a decrease in volume for the cab. Some weird things going on with the little hump at 200Hz.
Because the hump is also present and much stronger in the braced cabinet I made some measurements while changing the position of the spiral on the stand.
I used books on the stand to raise the spiral enclosure in order to have the drivers on the same height for all previous measurements. This time I also measured with the books removed and with the enclosure upside down. Looks like the books had more influence then just height because the whole assembly is a lot more wobbly. With the books gone and the spiral upside down the woofer is approximately the same height from the floor as with the braced cabinet and the peak returns.
The original cab didn't have the impedance peak but I only measured it with books, it is also much much lighter then the other cabs perhaps changing coupling to the floor
@ Cocoapuffs: glad you like it
I changed the wiring for the spiral enclosure, the crossover was a bit too far away and the woofer wire was kinda stretched. So I got some silver coated copper wire to change it. I also completely sealed the wire hole for the tweeter and crossover cavity with silicone afterwards.
The large impedance peak shifts a bit to the right indicating a decrease in volume for the cab. Some weird things going on with the little hump at 200Hz.
Because the hump is also present and much stronger in the braced cabinet I made some measurements while changing the position of the spiral on the stand.
I used books on the stand to raise the spiral enclosure in order to have the drivers on the same height for all previous measurements. This time I also measured with the books removed and with the enclosure upside down. Looks like the books had more influence then just height because the whole assembly is a lot more wobbly. With the books gone and the spiral upside down the woofer is approximately the same height from the floor as with the braced cabinet and the peak returns.
The original cab didn't have the impedance peak but I only measured it with books, it is also much much lighter then the other cabs perhaps changing coupling to the floor
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.