Progress!
First up, On initial testing, the Seos12 (Blue) looks more linear over the Faital LTH-102 (Red), especially as I don’t need the lower crossover point below ~1.5kHz.
Test 55-Litre Cabinet, tuned to 40Hz
Spent a good while playing around but could not get it to sound right in the mid/treble range, with the HF10AK and Seos 12 or LTH102, bass was/is superb. Really good punch if with a bit of resonance, still waiting on the speaker insulation to arrive, and that the bass cab is just screwed together for testing does not help.
The bean cans help to raise the down-firing ports off the floor, I've some Oak mid-century 15cm legs on their way!
Cabinet tuning at 40Hz and the impedance ripples don’t show up on on the raw driver impedance plot so that might relate to that resonance although nothing major on the plots further down in the 500-600Hz. range.
Ran the midrange much higher, up to around 6kHz and sounding a lot better, to take advantage of the higher crossover point, decided to give the Sica HFU 6x8 tweeters I have, these were bought a couple of years ago to match up with the Goodmans Axiom 301 project. It tests pretty well for a £30 tweeter unit! http://hifi-selbstbau.de/images/stories/chassis/sica_hfu_Z009501/Data_Z009501.htm
4th order crossover at 600Hz and 6.5kHz with just some EQ on the bass peaks (31Hz, 50Hz, 80Hz & 125Hz) with a mild -2dB drop in the midrange to flatten it a little.
Tried adding a small box on the side of midrange driver to lower the baffle step and raise the <700Hz region which helps but then creates a larger peak around 900Hz so I need to try a few different shapes on that I think.
Started adding a broad EQ drop in the mid/treble for a gradual roll off but need to do a bit more playing around first.
Individual driver responses at 1m on midrange (listening height) axis
And at the seating position
Not a bad start I think!
First up, On initial testing, the Seos12 (Blue) looks more linear over the Faital LTH-102 (Red), especially as I don’t need the lower crossover point below ~1.5kHz.
Test 55-Litre Cabinet, tuned to 40Hz
Spent a good while playing around but could not get it to sound right in the mid/treble range, with the HF10AK and Seos 12 or LTH102, bass was/is superb. Really good punch if with a bit of resonance, still waiting on the speaker insulation to arrive, and that the bass cab is just screwed together for testing does not help.
The bean cans help to raise the down-firing ports off the floor, I've some Oak mid-century 15cm legs on their way!
Cabinet tuning at 40Hz and the impedance ripples don’t show up on on the raw driver impedance plot so that might relate to that resonance although nothing major on the plots further down in the 500-600Hz. range.
Ran the midrange much higher, up to around 6kHz and sounding a lot better, to take advantage of the higher crossover point, decided to give the Sica HFU 6x8 tweeters I have, these were bought a couple of years ago to match up with the Goodmans Axiom 301 project. It tests pretty well for a £30 tweeter unit! http://hifi-selbstbau.de/images/stories/chassis/sica_hfu_Z009501/Data_Z009501.htm
4th order crossover at 600Hz and 6.5kHz with just some EQ on the bass peaks (31Hz, 50Hz, 80Hz & 125Hz) with a mild -2dB drop in the midrange to flatten it a little.
Tried adding a small box on the side of midrange driver to lower the baffle step and raise the <700Hz region which helps but then creates a larger peak around 900Hz so I need to try a few different shapes on that I think.
Started adding a broad EQ drop in the mid/treble for a gradual roll off but need to do a bit more playing around first.
Individual driver responses at 1m on midrange (listening height) axis
And at the seating position
Not a bad start I think!
Its the Butter Beans that are messing the sound, adding far to much wind🤣🤣
Looking forward to the final outcome. The HF10AK crossed at 6k should sound superb, almost like a BE dome. You should try different slopes and settings.
I have used them for years and are really very good. Volti Audio use them in the Vittora's
Looking forward to the final outcome. The HF10AK crossed at 6k should sound superb, almost like a BE dome. You should try different slopes and settings.
I have used them for years and are really very good. Volti Audio use them in the Vittora's
Not sure but appears in the manufacturer plots and other projects
https://faitalpro.com/products/files/HF10AK/8/HF10AK_response_8.gif
https://faitalpro.com/products/files/HF10AK/8/HF10AK_response_8.gif
horn throat = 34400/pi/2.54 = ~4311 Hz
1/2 WL circular eigenmode = 34400/2/4311 = ~2156 Hz
Added the legs and moved to the other side of the room so its a bit tidier while I chop and change between stuff.
Trying various combinations of angled baffle externders and found the remote control so fitted that!.
Trying various combinations of angled baffle externders and found the remote control so fitted that!.
I've been playing around with a couple of configurations.
I put the 2nd bass enclosure together and changed it to a slot port on both which is a bit neater, ended up with a port tuning frequency of 45Hz (~55 Litres). Also added some medium density foam on the cabinet walls and some loose fill lambs wool.
With the Veravox 5S with LTH102/HF10AK combo, I tried LR2 & LR4 slopes but could not get it to sound particulary good, crossing around 700Hz & 6.5kHz. Resulted in the plot below using Bessel 4th order filters sounded more integrated, (from the listening position) sounded great apart from some hardness in the treble which I thought is the 1kHz peak.
I also wanted to try the original configuration earlier in the thread with the Scanspeak D2004/6020 & 10F/8414. Using LR4 slopes at 500Hz & 4kHz I got a very well controlled and sounding result but still seemed to have some hardness in the treble so maybe its something else?
Like the first plot there is some EQ applied on the bass to tame large peaks around 35Hz & 75Hz and also a raised response in the 200Hz area. The foam helps flatten the response around 2-3kHz range when moving the cabinet back to time align them, I'm aware I can do that in DSP but trying things out!
Struggling to get a real good overall sound with either though. Any tips/pointers that might help?
I put the 2nd bass enclosure together and changed it to a slot port on both which is a bit neater, ended up with a port tuning frequency of 45Hz (~55 Litres). Also added some medium density foam on the cabinet walls and some loose fill lambs wool.
With the Veravox 5S with LTH102/HF10AK combo, I tried LR2 & LR4 slopes but could not get it to sound particulary good, crossing around 700Hz & 6.5kHz. Resulted in the plot below using Bessel 4th order filters sounded more integrated, (from the listening position) sounded great apart from some hardness in the treble which I thought is the 1kHz peak.
I also wanted to try the original configuration earlier in the thread with the Scanspeak D2004/6020 & 10F/8414. Using LR4 slopes at 500Hz & 4kHz I got a very well controlled and sounding result but still seemed to have some hardness in the treble so maybe its something else?
Like the first plot there is some EQ applied on the bass to tame large peaks around 35Hz & 75Hz and also a raised response in the 200Hz area. The foam helps flatten the response around 2-3kHz range when moving the cabinet back to time align them, I'm aware I can do that in DSP but trying things out!
Struggling to get a real good overall sound with either though. Any tips/pointers that might help?
could you turn off the tweeter and see if the harsh sound is still there as well as maybe try some different eq settings notching down different frequencies see if you can figure out where it's coming from
Still getting some high freqency issues (could well be me to be fair) but decided as I was not getting anywhere I would setup my Marshal Chong CM2 speakers (Seas Excel tweeter and Nextel W22 woofer) which have been my backup speakers for a while. They sound slightly bright, probably not helped by the sealed box design means they really need a subwoofer but at least a decent reference.
Still keeping with the LR4 crossovers but adjusting the levels so its a lot closer, and getting a nice result. That's at the listening seat position, with the CM2s running straight from a Bryston B60R and level matched.
Still keeping with the LR4 crossovers but adjusting the levels so its a lot closer, and getting a nice result. That's at the listening seat position, with the CM2s running straight from a Bryston B60R and level matched.
It seems to me that this bass reflex is too big. I use the manufacturer's recommended 75mm diameter and 90mm length in 20 liter net volume and I think everything is fine. Yours seems to be at least 300 by 50mm, (I could be wrong of course..) which is almost 2 times more even for both speakers... At least you can try with a smaller one...🙂 I had the same peak at 1khz at the beginning and cant remember exactly when I got rid of it...- maybe with the stronger boxes that I made afterwards - I'm not sure.. Greetings!
The port is 250mm x 40mm and 139mm long as modelled in Unibox, cabinet measurement was 45Hz so not far off the 47Hz predicted. This was based off the T/S parameters measured from an average of all four drivers.
could it be a diffraction issue are the smaller speakers usually set back like that to time align them ? I'm pretty sure you said they were . could some high frequencies be bouncing off the cabinet ? I know you had some foam there but sound doesn't always behave how you think it will they could be reflecting off a different part of the sub box or the foam might not be sufficient . would expect that to show up in the measurements though ?
The SPL data you showed are dominated by the room acoustics, so these data don't tell you much about how your speakers sound (your ears+brain are able to tell the difference between the direct sound from the speaker and the reflected sound from your room). I'd suggest you measure anechoic response data from your speaker so you have a chance to understand the acoustics of your speakers (not the room).
That said, I'd also suggest you measure each single driver without the xover, both on axis and off axis (say at 15°, 30° and 45° -- but more is better). This will show the off-axis dispersion of each driver and will help you to determine the x-over frequencies such that the summed response will be smooth both on-axis and off-axis. This will greatly help to get your speakers to sound balanced.
mbrennwa what is the bare minimum set up needed to measure each driver individually like this I mean in terms of the room setup I have r.e.w and umik and I'd like to measure some drivers to design a crossover . if I gate out the response can I do this close to the driver in a normal room in the house or does it need to be outside in the garden ideally ? thanks and sorry to hijack
@jccart Put your speaker and microphone in the center of the room such that all walls, floor, ceiling, furniture as far away as possible. The microphone should be about 1m away from the speaker. In a typical home environment, you'll be able to get anechoic data down to about 300-400 Hz.
If you want to know more about this stuff, take a look at the manual of your software, search the internet, ask around here (I am sure there are a bunch of threads already), or read a good book (Testing Loudspeakers by Joe d'Appolito is a good one).
If you want to know more about this stuff, take a look at the manual of your software, search the internet, ask around here (I am sure there are a bunch of threads already), or read a good book (Testing Loudspeakers by Joe d'Appolito is a good one).
OK thank you for that that's what I have done so far just need to brush up a bit more on r.e.w all Mt life pretty much I have dreamed of making my own speakers . I think about it constantly different designs in my head . getting closer slowly to making it a reality 🙂
Well, bit of an update on this project...
Various changes happened and I sold off a few bits but the long and the short is I now have a pair of Volt BM2500.4 drivers, current spec 8Ohm to restart the project around.
These are much newer versions that closer match the current spec rather than the ~90s ones I had before.
One thing I really liked with the previous quad of BM2500.4 drivers was the punch and dynamics in the bass (in a home enviroment) which was quite addictive, I've lost some headroom by only having the singles now but the previous drivers were sold to a friend last year. So ayway, keen to keep the Volt driver as my main Bass-Midrange ideally!
I have a number of drivers 'in stock' from previous projects including midrange/fullrange drivers:
The BM2500.4 only go up to around 1.7kHz before dropping like a stone so seem difficult to integrate into a 2-way without using a compression driver, I do have the BMS 4524 and Seos12 which could work, I know BMS say 1.9kHz crossover but this is strictly domestic use so could try a little lower.
I've really fancied trying the Volt VM527* 2" dome for some time, on paper should be an ideal match with the BM2500.4 with a crossover in the 700-800Hz region and then an easier match with either of the Founkek ribbons I have or the Scanspeak D2004/6020.
*I am aware there is a pair for sale in 'Swap Meet' but since Brexit, I'll be charged VAT and Import duty ontop so will end up pretty much the same as a new pair!
I've another long term plan with the Scanspeak 10F and a pair of Scanspeak 26W/4534 in nother system as well as the Supravox 165GMF as wide range drivers alongside some vintage Goodmans Axiom 301/Audiom 81s but again, that's for another day!
I've since moved our 'spare' room around which works much better for audio now, so has become the main hifi space, bit less restricted with what I can get away with. Shown here with some Tannoy MG15 Lancasters I had for a short while, and the push for me to try some of the vintage Goodmans drivers again as a side project!
Library System by Robert Seymour, on Flickr
Various changes happened and I sold off a few bits but the long and the short is I now have a pair of Volt BM2500.4 drivers, current spec 8Ohm to restart the project around.
These are much newer versions that closer match the current spec rather than the ~90s ones I had before.
One thing I really liked with the previous quad of BM2500.4 drivers was the punch and dynamics in the bass (in a home enviroment) which was quite addictive, I've lost some headroom by only having the singles now but the previous drivers were sold to a friend last year. So ayway, keen to keep the Volt driver as my main Bass-Midrange ideally!
I have a number of drivers 'in stock' from previous projects including midrange/fullrange drivers:
- Scanspeak 10F/8414
- Veravox 5S
- Supravox 165GMF (only the single right now)
- Fountek Neo CD2.0
- Fountek JP2.0
- Scanspeak D2004/6020
- Peerless D27TG-35-06
- Sica 6x8HFU
- BMS 4524 with Seos12 Waveguide
The BM2500.4 only go up to around 1.7kHz before dropping like a stone so seem difficult to integrate into a 2-way without using a compression driver, I do have the BMS 4524 and Seos12 which could work, I know BMS say 1.9kHz crossover but this is strictly domestic use so could try a little lower.
I've really fancied trying the Volt VM527* 2" dome for some time, on paper should be an ideal match with the BM2500.4 with a crossover in the 700-800Hz region and then an easier match with either of the Founkek ribbons I have or the Scanspeak D2004/6020.
*I am aware there is a pair for sale in 'Swap Meet' but since Brexit, I'll be charged VAT and Import duty ontop so will end up pretty much the same as a new pair!
I've another long term plan with the Scanspeak 10F and a pair of Scanspeak 26W/4534 in nother system as well as the Supravox 165GMF as wide range drivers alongside some vintage Goodmans Axiom 301/Audiom 81s but again, that's for another day!
I've since moved our 'spare' room around which works much better for audio now, so has become the main hifi space, bit less restricted with what I can get away with. Shown here with some Tannoy MG15 Lancasters I had for a short while, and the push for me to try some of the vintage Goodmans drivers again as a side project!
Library System by Robert Seymour, on Flickr
A bit of progress, I've started with the BM2500.4 and the BMS 4524/Seos12 combination to see where I get. The volt drivers are in a 55L ported enclosure that they came in, although they will end up in a much smaller sealed enclosure if I go 3-way.
I've started just on the 2-way Volt/BMS combination right now as I only have the one MiniDSP 2x4HD, and then once I've sorted a solution for 3-way DSP and/or Amplifiers (maybe another pair of Hypex Fusion as a neat solution) I'll crack on with the bass.
I have this quad of vintage Goodmans Audiom 81 'Bass' drivers I recently aquired, two well matched pairs. I like the look of going sealed again with these, either with a single pair in a 90L cab, or the compound pair in a 120L cab which gives much more scope for headroom and a bit of EQ.
I used the calculations here for combining the two pairs of drivers as although there are two quite well matched pairs, the two pairs vary quite a bit.
http://web.archive.org/web/20080610...ndell/xentar/1179/theory/dddllqd/dddllqd.html
Pair one & two
90L Sealed using Pair Two (well matched Lower Fs, higher Qts) and compound of one from each pair in a 120 Sealed. Qtc of ~0.7 on both.
The overall plan is with the large cabinets below with the Audiom 81s, smaller midrange cabinet for the Volt and Seos. The mid/treble may well be replaced with other combinations such as the Supravox 165GMF/Fountek ribbon for example.
The current large cabinets (below) at the bottom are some spare 80L ported used with Axiom 201 or 301 drivers I have already that have been in use as main speakers up until now. Just handy to provide stands for the other cabs while I'm experimenting but also to give me an idea how it will all sit in the room!
I've started just on the 2-way Volt/BMS combination right now as I only have the one MiniDSP 2x4HD, and then once I've sorted a solution for 3-way DSP and/or Amplifiers (maybe another pair of Hypex Fusion as a neat solution) I'll crack on with the bass.
I have this quad of vintage Goodmans Audiom 81 'Bass' drivers I recently aquired, two well matched pairs. I like the look of going sealed again with these, either with a single pair in a 90L cab, or the compound pair in a 120L cab which gives much more scope for headroom and a bit of EQ.
I used the calculations here for combining the two pairs of drivers as although there are two quite well matched pairs, the two pairs vary quite a bit.
http://web.archive.org/web/20080610...ndell/xentar/1179/theory/dddllqd/dddllqd.html
Pair one & two
90L Sealed using Pair Two (well matched Lower Fs, higher Qts) and compound of one from each pair in a 120 Sealed. Qtc of ~0.7 on both.
The overall plan is with the large cabinets below with the Audiom 81s, smaller midrange cabinet for the Volt and Seos. The mid/treble may well be replaced with other combinations such as the Supravox 165GMF/Fountek ribbon for example.
The current large cabinets (below) at the bottom are some spare 80L ported used with Axiom 201 or 301 drivers I have already that have been in use as main speakers up until now. Just handy to provide stands for the other cabs while I'm experimenting but also to give me an idea how it will all sit in the room!
First working development, a 1st order BW on the BM25004 to match 4th order BW on the 4524, both at 1.7kHz seem to give a nice response. The 4524 has a low Q (5kHz/-6dB/0.5Q) low shelf to flatten the horn response and overall has a nice downwards trend in room. Driver measurements taken at baffle distance on axis of both drivers.
I need to take them outside when I have some free time to set all the equipment up to get a clean 1m measurement and fine tuning, subjectively it sounds nice although a little thin in the midrange (on vocals) which does show up with a dip in the 400Hz-800Hz range in the close plot as well as the seated position,
In this room, I have a very large peak at 32Hz, showing here even after a -16dB peak in PEQ to bring it down.
I do have a noise problem from the mains that shows up on the ~109dB 4524 but not on the BM25004 (or Goodmans Axiom ~92-97dB drivers) so I might try reducing the output with an L-Pad directly (around 9dB) to start with and see if that helps. The noise shows up on these tweeters connected to multiple amplifiers with and without the MiniDSP so its not something equipment specific but a ground loop or something else in the house that I've not found as yet!
I need to take them outside when I have some free time to set all the equipment up to get a clean 1m measurement and fine tuning, subjectively it sounds nice although a little thin in the midrange (on vocals) which does show up with a dip in the 400Hz-800Hz range in the close plot as well as the seated position,
In this room, I have a very large peak at 32Hz, showing here even after a -16dB peak in PEQ to bring it down.
I do have a noise problem from the mains that shows up on the ~109dB 4524 but not on the BM25004 (or Goodmans Axiom ~92-97dB drivers) so I might try reducing the output with an L-Pad directly (around 9dB) to start with and see if that helps. The noise shows up on these tweeters connected to multiple amplifiers with and without the MiniDSP so its not something equipment specific but a ground loop or something else in the house that I've not found as yet!
Last edited: