So, had a bit more time to look into biox design etc. and seem to now have the handle on using BoxSim.
So results so far are looking interesting.
So, found a studio monitor that had the mid and tweeter side by side in a 3 way box (Focal Trio 6) and a small front facing slot port, so thought I'd try and copy in Box Sim using the inbuilt Visaton specs.
Chosen Speakers - W170S-8 6.5" woofer, FR10-8 4" full range and G20SC 0.8" soft dome tweeter.
Box design -
Woofer - L slot woofer with a 20mm slot equal chamber to rear and up to top. According to L-Slot Vent Bass Reflex Enclosure Calculator - DB DYNAMIX AUDIO the port is tunes to 43Hz, box volume is approx 20l
Mid - Sealed box design, approx 5l
Tweeter - own sealed box but that is irrelevant.
Tried to get the smallest box possible to reduce baffle effects which turned out to have overall dimensions of H435mm x W280mm x D350mm modeled with 20mm sides and internals and 30mm front baffle.
I then put this through BoxSim with the speaker parameters, used active 24db/octave filters with low xover at 300hz and high at 2000hz this seemed to give a pretty flat response with a small low and high boost and a bit of break up at around the 200hz cross over point. This was the result -
Now, this is not the best balanced system being a £35 woofer, £15 mid and £50 tweeter but the frequency response profiles are fairly well behaved and are similar to many other speakers plus BoxSim has the full data spec built in so no tedious filling in of the forms and finding the FRD/ZMA files.
This has been a steep earning curve but I think I am starting to understand what variable do what to the overall frequency response. Obviously taking out the xover design by using a simple active xover helps.
So, I think I may try to increase the size slightly and run the process for an 8" woofer to give 10-20 more Hz at the bottom end and then look to try and model some other speakers that are more suitable. Looking at the Monacore SPH220HQ, SPH8M or DA RS220P for the 8", the SB W4-1337 or Monacor SPH115HQ or DA RS100P for the Mids and Vifa XT25 or Seas 26TBFC/G fro the tweeter.
All comments welcome as to whether there are any issues I have missed as this seems a workable solution so far.
So results so far are looking interesting.
So, found a studio monitor that had the mid and tweeter side by side in a 3 way box (Focal Trio 6) and a small front facing slot port, so thought I'd try and copy in Box Sim using the inbuilt Visaton specs.
Chosen Speakers - W170S-8 6.5" woofer, FR10-8 4" full range and G20SC 0.8" soft dome tweeter.
Box design -
Woofer - L slot woofer with a 20mm slot equal chamber to rear and up to top. According to L-Slot Vent Bass Reflex Enclosure Calculator - DB DYNAMIX AUDIO the port is tunes to 43Hz, box volume is approx 20l
Mid - Sealed box design, approx 5l
Tweeter - own sealed box but that is irrelevant.
Tried to get the smallest box possible to reduce baffle effects which turned out to have overall dimensions of H435mm x W280mm x D350mm modeled with 20mm sides and internals and 30mm front baffle.
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.
I then put this through BoxSim with the speaker parameters, used active 24db/octave filters with low xover at 300hz and high at 2000hz this seemed to give a pretty flat response with a small low and high boost and a bit of break up at around the 200hz cross over point. This was the result -
An externally hosted image should be here but it was not working when we last tested it.
Now, this is not the best balanced system being a £35 woofer, £15 mid and £50 tweeter but the frequency response profiles are fairly well behaved and are similar to many other speakers plus BoxSim has the full data spec built in so no tedious filling in of the forms and finding the FRD/ZMA files.
This has been a steep earning curve but I think I am starting to understand what variable do what to the overall frequency response. Obviously taking out the xover design by using a simple active xover helps.
So, I think I may try to increase the size slightly and run the process for an 8" woofer to give 10-20 more Hz at the bottom end and then look to try and model some other speakers that are more suitable. Looking at the Monacore SPH220HQ, SPH8M or DA RS220P for the 8", the SB W4-1337 or Monacor SPH115HQ or DA RS100P for the Mids and Vifa XT25 or Seas 26TBFC/G fro the tweeter.
All comments welcome as to whether there are any issues I have missed as this seems a workable solution so far.
So, been following the 3way 8+1 thread recently and had a play surround with a few more options in BoxSim but nothing I tried seemed to hit the mark until I stumbled across this -
First here are the results -
Looks pretty good to my eyes, almost flat from 30Hz to 20Khz but it is a bit of a mishmash of ideas. So what is it-
Started off with the Jim Holtz Statement Monitor and went with the TB W4-1337SDF mid range driver, looks good from 100Hz to 5Khz and in the Statement Monitor seems to get good results. I then added the SB 21RDC small ring tweeter, looks pretty good and flat from about 1khz to 20khz, slightly better response than the bigger 29RDC but obviously does not go as low down. The for a woofer, I have tried many different one and came across the TB 6W-1139SIF 6.5" sub woofer - fast, loads of bass extension and reasonably priced but will need a low cross over and quite a bit of power.
I also need to give some credit to JSadzi for his gorgeous floor stander that also provided inspiration for the W6 driver being used with a slightly larger full range and also the XT25 which is similar to the SB RDF tweeters. - Thurston 3-ways Build Thread XT25 Peerless 5.25" TangBand W6-1139SIF
For the box design, it is basically a slightly truncated Statement monitor, H465mm x W260mm x D300mm. The main box comes out at 17.5l without the Mid enclosure and the mid enclosure at 2.8l. So, plumbing these drivers and box design into BoxSim gave some pretty good results. Playing around with the crossovers and port resonant frequencies I came up with the results above. Run down of the parameters -
Sub Woofer - 17.5l vented box, light stuffing, port frequency 28hz
Mid Driver - 2.87l vented box, light stuffing, port frequency 86hz
Notes - the low port frequency is required to flatten out a bass hump, the use of a vented box for the mids really helped the lower crossover integration and the high crossover can really be anywhere from 1500hz to 3500hz.
For the crossover, I have always intended on using either an active setup or possible minidsp, so these ended up as -
Low - 135hz 24db/octave
High 2500hz, 24db/octave
plus gain adjustments of 12db, 4db and -0.5db for bass, mid, high respectively.
I know the gain adjustments may cause a bit of an issue with amps but as it will be active I can use this to my advantage, tailoring the output to the driver, so as an example - 2xRotel RB850 in bridge mode 150w for the bass, Rotel RB850 in stereo mode 50W for the mids and possibly a class A or valve for the highs at 25W.
So, this looks feasible assuming I can get the low port frequency, in a reasonable size package with lots of bass extension and it is fairly cheap for the drivers (approx £130 per speaker).
All comments welcome.
First here are the results -
An externally hosted image should be here but it was not working when we last tested it.
Looks pretty good to my eyes, almost flat from 30Hz to 20Khz but it is a bit of a mishmash of ideas. So what is it-
Started off with the Jim Holtz Statement Monitor and went with the TB W4-1337SDF mid range driver, looks good from 100Hz to 5Khz and in the Statement Monitor seems to get good results. I then added the SB 21RDC small ring tweeter, looks pretty good and flat from about 1khz to 20khz, slightly better response than the bigger 29RDC but obviously does not go as low down. The for a woofer, I have tried many different one and came across the TB 6W-1139SIF 6.5" sub woofer - fast, loads of bass extension and reasonably priced but will need a low cross over and quite a bit of power.
I also need to give some credit to JSadzi for his gorgeous floor stander that also provided inspiration for the W6 driver being used with a slightly larger full range and also the XT25 which is similar to the SB RDF tweeters. - Thurston 3-ways Build Thread XT25 Peerless 5.25" TangBand W6-1139SIF
An externally hosted image should be here but it was not working when we last tested it.
For the box design, it is basically a slightly truncated Statement monitor, H465mm x W260mm x D300mm. The main box comes out at 17.5l without the Mid enclosure and the mid enclosure at 2.8l. So, plumbing these drivers and box design into BoxSim gave some pretty good results. Playing around with the crossovers and port resonant frequencies I came up with the results above. Run down of the parameters -
Sub Woofer - 17.5l vented box, light stuffing, port frequency 28hz
Mid Driver - 2.87l vented box, light stuffing, port frequency 86hz
Notes - the low port frequency is required to flatten out a bass hump, the use of a vented box for the mids really helped the lower crossover integration and the high crossover can really be anywhere from 1500hz to 3500hz.
For the crossover, I have always intended on using either an active setup or possible minidsp, so these ended up as -
Low - 135hz 24db/octave
High 2500hz, 24db/octave
plus gain adjustments of 12db, 4db and -0.5db for bass, mid, high respectively.
I know the gain adjustments may cause a bit of an issue with amps but as it will be active I can use this to my advantage, tailoring the output to the driver, so as an example - 2xRotel RB850 in bridge mode 150w for the bass, Rotel RB850 in stereo mode 50W for the mids and possibly a class A or valve for the highs at 25W.
So, this looks feasible assuming I can get the low port frequency, in a reasonable size package with lots of bass extension and it is fairly cheap for the drivers (approx £130 per speaker).
All comments welcome.
Last edited:
Study these carefully and make up your mind after that 😉...
I would like to see the results of mounting the driver on the flat side of the hemisphere.
I agree the truncated pyramid looks interesting. Not so sure about the WAF however. 🙂
I thought it was too nice a graph.
So, checked the input FRD files that I was using (can't remember where i got them from) and they look like they have been smoothed quite a bit. I re-modeled them using the manufacturer's data sheets and FPgraphtrace and then checked them by importing them into Excel, creating a graph from the FRD file, copying and pasting the graph and a screen shot of the datasheet into PowerPoint, finally scaling and overlaying. FRD's now match the manufacturers datasheet.
Quick tip - if you import photo/graph picture into PowerPoint you can use the format/colour/set transparency button, clicking on the white background, to just have the graph lines showing and so use as an overlay on other graphs.
Anyway, had a little play with my modified Statement Monitor BoxSim model and it is now quite lumpy but still has good bass reach and is still fairly flat. The W4-1337 does have a peak at about 2k so a lower than usual crossover help this, also had to higher the bass box port frequency 35hz which is line with JShadzi's findings plus I also now get the bass boost hump that he also saw at around 100hz. So I think this is a more realistic model but still a feasible speaker I think. Crossovers now at 270hz and 2000hz.
Next job is to look at a port slot at the front rather than a port tube at the back. Initial look seems to show a 1/2" port slot full width needs a 18" tract so easily doable inside the box.
Results.
So, checked the input FRD files that I was using (can't remember where i got them from) and they look like they have been smoothed quite a bit. I re-modeled them using the manufacturer's data sheets and FPgraphtrace and then checked them by importing them into Excel, creating a graph from the FRD file, copying and pasting the graph and a screen shot of the datasheet into PowerPoint, finally scaling and overlaying. FRD's now match the manufacturers datasheet.
Quick tip - if you import photo/graph picture into PowerPoint you can use the format/colour/set transparency button, clicking on the white background, to just have the graph lines showing and so use as an overlay on other graphs.
Anyway, had a little play with my modified Statement Monitor BoxSim model and it is now quite lumpy but still has good bass reach and is still fairly flat. The W4-1337 does have a peak at about 2k so a lower than usual crossover help this, also had to higher the bass box port frequency 35hz which is line with JShadzi's findings plus I also now get the bass boost hump that he also saw at around 100hz. So I think this is a more realistic model but still a feasible speaker I think. Crossovers now at 270hz and 2000hz.
Next job is to look at a port slot at the front rather than a port tube at the back. Initial look seems to show a 1/2" port slot full width needs a 18" tract so easily doable inside the box.
Results.
An externally hosted image should be here but it was not working when we last tested it.
Last edited:
Had a bit of spare time so have been having another google and think.
Thinking back on the equipment I have previously had I always go back to my Quad 44/405-2 and kef 104/2 setup, wish I had never sold them on. Decent bass thump but the mid/upper mid was to die for, sweet and detailed.
So, when doing a bit of modleeling I could not get an 8+1 monitor to work without the box being bigger than I would want on a desk and the Bass started to tail off from about 60-70Hz. So, I think I will have to go back to the 104/2 and a more modern 3 way floor stander.
So, having got used to BoxSim which is basically a Visaton tool for use with their products but other FRD/ZMA files can be imported and the parameters manually entered. I have searched an found that I like the more modern, thin/tall floor standers (always liked the B&W P4) and came across the following -
Zaph ZDT3.5 - Zaph|Audio - ZDT3.5 - nice look but want a single bass driver and this is a three (2 used) cavity design.
Visaton 200 - CLASSIC 200 GF | Visaton - again like the look but not sure about the mid band driver.
Jensen/Troels Seas - Jenzen CA - looks a bit chunky but seems to work well.
Scanspeak Studio 85 - Scan-Speak - again like the look but not sure about the two larger drivers
Anyway, after a bit of fun modelling a few arrangements I have come up with the following - this pushed my budget a bit but at around £350 per speaker for drivers and minidsp plus cabinet materials this is just about do-able.
Headlines - Tallish (100x26x34cm) three way floorstander, 8+3.3+1 arrangement, 50l bass with a 30hz front slot port plus a 1.5l sealed mid/tweeter cavity. Mid and Tweeter off set to the side a little. Drivers are the Visaton TIW200XS 8" woofer (sub?), B80 3.3" full range and G20SC 1" soft dome tweeter.
The TIW200XS is boarderline 8" sub but has good extension but is pretty frlat up to over 500hz, the B80 seems to be a pretty good, flat mid driver that goes from 300Hz to over 8Khz (could even be used upto 20Khz with care) and the G20SC look good and flat (slightly rising) from 2.5KHz upwards. So, good crossover of frequencies between them.
Following a bit of modelling the layout above seemed to give pretty good results with an active crossover with 24db/octave at 300hz and 2500hz plus a small amount of gain adjustment (+3db for bass, +0.25db for mid and -1.5db for tweeter). This gives a pretty flat response, slight rise at the high end and a slight drop off at the l;over, 85db sensitivity (likely to be around 6Ohm) and an f3 of around 30hz. Slight dips in output at around 120hz and again at 3500hz but these are pretty small and could be sorted out in the minidsp. Not checked the timing but may have a go at guessing where the centre of the drivers are in the fore/aft plane and see if they are anyway near.
Here's the plan, modelling details and results if anyone wants to check these are OK. As always comments welcome as to whether this looks like a reasonable scheme, if the bass driver is more of a sub than a woofer and any comments generally on Visaton drivers as they do not seem to be very prominent in builds on here.
General Layout -
Driver setup -
Slot port calc -
Driver models -
Final frequency plot -
Thinking back on the equipment I have previously had I always go back to my Quad 44/405-2 and kef 104/2 setup, wish I had never sold them on. Decent bass thump but the mid/upper mid was to die for, sweet and detailed.
So, when doing a bit of modleeling I could not get an 8+1 monitor to work without the box being bigger than I would want on a desk and the Bass started to tail off from about 60-70Hz. So, I think I will have to go back to the 104/2 and a more modern 3 way floor stander.
So, having got used to BoxSim which is basically a Visaton tool for use with their products but other FRD/ZMA files can be imported and the parameters manually entered. I have searched an found that I like the more modern, thin/tall floor standers (always liked the B&W P4) and came across the following -
Zaph ZDT3.5 - Zaph|Audio - ZDT3.5 - nice look but want a single bass driver and this is a three (2 used) cavity design.
Visaton 200 - CLASSIC 200 GF | Visaton - again like the look but not sure about the mid band driver.
Jensen/Troels Seas - Jenzen CA - looks a bit chunky but seems to work well.
Scanspeak Studio 85 - Scan-Speak - again like the look but not sure about the two larger drivers
Anyway, after a bit of fun modelling a few arrangements I have come up with the following - this pushed my budget a bit but at around £350 per speaker for drivers and minidsp plus cabinet materials this is just about do-able.
Headlines - Tallish (100x26x34cm) three way floorstander, 8+3.3+1 arrangement, 50l bass with a 30hz front slot port plus a 1.5l sealed mid/tweeter cavity. Mid and Tweeter off set to the side a little. Drivers are the Visaton TIW200XS 8" woofer (sub?), B80 3.3" full range and G20SC 1" soft dome tweeter.
The TIW200XS is boarderline 8" sub but has good extension but is pretty frlat up to over 500hz, the B80 seems to be a pretty good, flat mid driver that goes from 300Hz to over 8Khz (could even be used upto 20Khz with care) and the G20SC look good and flat (slightly rising) from 2.5KHz upwards. So, good crossover of frequencies between them.
Following a bit of modelling the layout above seemed to give pretty good results with an active crossover with 24db/octave at 300hz and 2500hz plus a small amount of gain adjustment (+3db for bass, +0.25db for mid and -1.5db for tweeter). This gives a pretty flat response, slight rise at the high end and a slight drop off at the l;over, 85db sensitivity (likely to be around 6Ohm) and an f3 of around 30hz. Slight dips in output at around 120hz and again at 3500hz but these are pretty small and could be sorted out in the minidsp. Not checked the timing but may have a go at guessing where the centre of the drivers are in the fore/aft plane and see if they are anyway near.
Here's the plan, modelling details and results if anyone wants to check these are OK. As always comments welcome as to whether this looks like a reasonable scheme, if the bass driver is more of a sub than a woofer and any comments generally on Visaton drivers as they do not seem to be very prominent in builds on here.
General Layout -
Driver setup -
Slot port calc -
Driver models -
Final frequency plot -
That would look very much like the mounting on the end of the cylinder.
I would like to point out that it is well known that Olson did his figures by hand drawing them. They weren't based on measurements or calculation, just how he figured things would work. Sometimes he was right and sometimes he was wrong. Just take his figures with a grain of salt.
Just tried another, cheaper 8” bass driver. The GF200 models quite well with no real changes to crossover points or box. Does not dig as deep and needs the port frequency to be nearer 40hz to work. An option to save £50-60 I guess.
Last edited:
Sorry for digging up this 2 year old thread, but is there a study on how the enclosure response is affected by lining and stuffing? Say that dreaded flat cylinder had a thick lining covering the flat portion. What then?
On the outside, you mean? Damping material will tend to make the wave go away there, similar to if the baffle folded away at that point, only frequency dependent like the damping material.
Mmm... no, on the inside, the side facing the back of the driver. I'm assuming the variations in the response curve are because of reflections from the flat back and am wondering whether the lining would cut down on that and help even out the response.
Last edited:
These effects are the result of the outside of the enclosure, and it's assumed the inside is having no effect.