Hi there,
I have just built this:
This is an attempt at the Seas Idunn DIY kit, using the woven polypropylene - U18RNX/P woofer, and a 27 mm aluminium/magnesium dome with a 'DXT' lens - 27TBCD/GB-DXT.
The cabinet is a 12 litre back vented cab, built in mdf with double front/back baffles, and internal bracing. It is based on a commercial 6.5" design. I am using the standard crossover suggested by Seas.
So far I have been pleased with the performance, but the sound seems a bit bright, and I am not sure if what I am hearing really is the small cabinet or a lack of mids. Is it possible that the new cab requires a different tuning for the woofer
I know this is not the first time someone has raised this issue about the Idunn, though. Also, the Idunn kit is very similiar to Mark K's ER18DXT:
The Seas ER18DXT ported two way
Could I take something from this crossover, and possibly tweak it to suit my Idunns?
I wouldn't know, but it seems to me that this particular crossover renders a smoother response, but perhaps more importantly;
it is tuned to a cabinet that more closely resembles mine. Any suggestions welcome
I have just built this:
An externally hosted image should be here but it was not working when we last tested it.
This is an attempt at the Seas Idunn DIY kit, using the woven polypropylene - U18RNX/P woofer, and a 27 mm aluminium/magnesium dome with a 'DXT' lens - 27TBCD/GB-DXT.
The cabinet is a 12 litre back vented cab, built in mdf with double front/back baffles, and internal bracing. It is based on a commercial 6.5" design. I am using the standard crossover suggested by Seas.
So far I have been pleased with the performance, but the sound seems a bit bright, and I am not sure if what I am hearing really is the small cabinet or a lack of mids. Is it possible that the new cab requires a different tuning for the woofer
I know this is not the first time someone has raised this issue about the Idunn, though. Also, the Idunn kit is very similiar to Mark K's ER18DXT:
The Seas ER18DXT ported two way
Could I take something from this crossover, and possibly tweak it to suit my Idunns?
An externally hosted image should be here but it was not working when we last tested it.
I wouldn't know, but it seems to me that this particular crossover renders a smoother response, but perhaps more importantly;
it is tuned to a cabinet that more closely resembles mine. Any suggestions welcome
Wasn`t the Idunn kit suggested for a 20l vented cab? If so, the 12l you have there may be a reason to lift your F6 frequency. What port tunning have you used and how much stuffing does the enclosure have on the inside?
The crossover looks like designed for a 2.4-2.6Khz ( my guess ) crossover point. Do they sound thin or overly bright? You can add a 1-1.5ohm resistor between the amp and the crossover if they`re bright.
The crossover looks like designed for a 2.4-2.6Khz ( my guess ) crossover point. Do they sound thin or overly bright? You can add a 1-1.5ohm resistor between the amp and the crossover if they`re bright.
I am glad you brought that up. I didn't answer you at once as I took the time to do some testing. I thought I had tuned the bass okay, but shortening it definitely brought more energy to the low end.
According to my (rough) calculations the port is now tuned to 65 Hz. 4.5 dia./10 cm L.
The enclosure is stuffed with acoustic foam on all sides.
XO-frequency is 2.2 kHz 2nd./3.rd order. What would the resistor do, change impedance?
According to my (rough) calculations the port is now tuned to 65 Hz. 4.5 dia./10 cm L.
The enclosure is stuffed with acoustic foam on all sides.
XO-frequency is 2.2 kHz 2nd./3.rd order. What would the resistor do, change impedance?
Hi Alhe, as Mario posted the Idunn's x/o is designed for a 20l cabinet, therefore I'd presume the bass and mids to be are not right.
I'm nearly finished building 2-ways similar to Idunn's, same cab size different tweeter (27TBFCG) and am currently redesigning the x/o. My friend has a design similar to Gorin's from Audioexcite.com which uses the Seas 5" PP cone with the Seas DXT in a 12l cab and he has just finished Gorin's x/o design with Jantzen Audio conponents and is very impressed. He has had problems with brightness with previous x/o design and notived the speaker cables to the dxt tweeter was sensistive to different ones used, eventually using Chord cable which calmed the tweeters brighness down till he got the new x/o sorted.
I have a feeling you may need a lower mH inductor on the woofers which may need a new x/o design completely, changing R1 to 1.5 Ohm may help for now.
http://www.audioexcite.com/?page_id=2386
I'm nearly finished building 2-ways similar to Idunn's, same cab size different tweeter (27TBFCG) and am currently redesigning the x/o. My friend has a design similar to Gorin's from Audioexcite.com which uses the Seas 5" PP cone with the Seas DXT in a 12l cab and he has just finished Gorin's x/o design with Jantzen Audio conponents and is very impressed. He has had problems with brightness with previous x/o design and notived the speaker cables to the dxt tweeter was sensistive to different ones used, eventually using Chord cable which calmed the tweeters brighness down till he got the new x/o sorted.
I have a feeling you may need a lower mH inductor on the woofers which may need a new x/o design completely, changing R1 to 1.5 Ohm may help for now.
http://www.audioexcite.com/?page_id=2386
Last edited:
65Hz port tuning is pretty high for a 6.5" woofer. The Idunn is tuned to 38Hz, thats a big difference, no wonder it sounds thin. I suspect the reason for this pretty high port tuning is the required port lenght. A few notes:
- The port air volume must be subtracted from the effective box internal volume. If you`re not doing it, you have less than 12l in there
- In a compact design that requires a longer port, a flared port ( both ends ) with a 90 degree bend could do the trick - will allow you to use the required lenght. Easiest way to do it is by joining two standard one-end flared ports with a 90 degree bend. Joint must be sealed.
- This acoustic foam is pretty dense. It is good for large cabinets but in this one I believe that a softer material like polyester or wool might be better suited, you can try both and leave the one you like the best.
- This woofer is pretty large compared to the enclosure, running a small diameter port would very likely introduce port noise even at moderate levels due to high air compression.
One thing that made me interested in this project is the combination of a polycone to a hard dome. Personally, I wouldn`t do it based on my experience with other such drivers, but I haven`t heard the two Seas units described so this should be taken with a grain of salt.
And finally, changing only R1 would change the L-pad function and the impedance seen by the crossover, thus shifting its function. If you increase R1, you must reduce R2 accordingly, to something like 8R2 so the impedance seen by the crossover/amp remains the same as the original, even if the impedance graph would be slightly different as the lower parallel resistor value would reduce even further the tweeter`s impedance rise at Fs. The better way - don`t touch the L-pad and just add an extra resistor, R3, before the crossover - i.e. between the woofer inductor and the amplifier. Then, adjust the value to your own taste. My guess would be a 1-1R5 as suggested above, 5W would work fine. The Idunn measurement indicates slightly more power in the upper region, if your taste is like mine, you`d find it bright. I like to tune my speakers with +2-3db or so in the bass region and gradually reducing the level where the top end is some -2db compared to the 100Hz-3Khz region, just find them to sound more dynamic and natural this way.
- The port air volume must be subtracted from the effective box internal volume. If you`re not doing it, you have less than 12l in there
- In a compact design that requires a longer port, a flared port ( both ends ) with a 90 degree bend could do the trick - will allow you to use the required lenght. Easiest way to do it is by joining two standard one-end flared ports with a 90 degree bend. Joint must be sealed.
- This acoustic foam is pretty dense. It is good for large cabinets but in this one I believe that a softer material like polyester or wool might be better suited, you can try both and leave the one you like the best.
- This woofer is pretty large compared to the enclosure, running a small diameter port would very likely introduce port noise even at moderate levels due to high air compression.
One thing that made me interested in this project is the combination of a polycone to a hard dome. Personally, I wouldn`t do it based on my experience with other such drivers, but I haven`t heard the two Seas units described so this should be taken with a grain of salt.
And finally, changing only R1 would change the L-pad function and the impedance seen by the crossover, thus shifting its function. If you increase R1, you must reduce R2 accordingly, to something like 8R2 so the impedance seen by the crossover/amp remains the same as the original, even if the impedance graph would be slightly different as the lower parallel resistor value would reduce even further the tweeter`s impedance rise at Fs. The better way - don`t touch the L-pad and just add an extra resistor, R3, before the crossover - i.e. between the woofer inductor and the amplifier. Then, adjust the value to your own taste. My guess would be a 1-1R5 as suggested above, 5W would work fine. The Idunn measurement indicates slightly more power in the upper region, if your taste is like mine, you`d find it bright. I like to tune my speakers with +2-3db or so in the bass region and gradually reducing the level where the top end is some -2db compared to the 100Hz-3Khz region, just find them to sound more dynamic and natural this way.
Last edited:
Everybody grumbles about the SEAS Idunn being thin on bass and bright in the midrange!
I think drastic measures are needed. I'd fit a bigger bafflestep coil and add some resistance to a smaller shunt capacitor. I really don't see much benefit in a RLC shunt here, nor in that complex treble filter.
This seems to work on phase well enough while reducing the midrange energy 2dB at least. You'll probably need to reduce the treble energy too. 3.3 or 4.7R series ought to do it. Impedance remains good.
I think drastic measures are needed. I'd fit a bigger bafflestep coil and add some resistance to a smaller shunt capacitor. I really don't see much benefit in a RLC shunt here, nor in that complex treble filter.
This seems to work on phase well enough while reducing the midrange energy 2dB at least. You'll probably need to reduce the treble energy too. 3.3 or 4.7R series ought to do it. Impedance remains good.
Attachments
Hi!
Yes, I am one of the ones that have raised the same issue as you. Actually, I am using the same configuration as you are with the U18 woofer and the DXT tweeter.
First I tried the orignal Idunn x-over. Could not live with that, so my search for improvements started.
My conclusion was to discard the Idunn XO and use the Mark ER18DXT filter instead: That resulted in a voicing much more in line with my liking!
I have done some minor adjustments, based on trial and error. I have reduced the Lpad damping slightly, reduced the capacitor in series with the tweeter from 5.6 to 4.7 uF and finally I ended up reducing the woofer series coil from 2.5 to 2.2 mH. Note that my changes have been done to satisfy my personal likings, and also it is worth noting that my speakers are mounted in a bafflewall which reduces/eliminate the need for baffle step compensation (BSC). Actually, I plan to reduce the cabinet size from current 22 liters as I would like more level/dynamics in the mid bass. The "sub" bass (<80 Hz) are too strong as it is.
I hope you succeed with your design, I look forward to read and follow your tread!
Best regards,
Bjarne
Looking at the measurements below, there is a slight rise in the response in the 220-800Hz region, looks like a db or so. This should be inaudible. Tweeter however looks some 2db above the woofer level or for over 4db above the bass region that would equalize the perceived brightness. Human hearing is far more sensitive in the treble than it is in the bass region, often many highly successfull commercial speakers do the trick to add a small bump of about 2-3db in the 60-90Hz region and feature a reduced tweeter level for this matter. I would double the advice to increase the BSC by a larger inductor ( watch for inductor DCR increase here and take it into account when you pad the tweeter ) and add an extra resistor before the crossover, that would save you parts and money. The MarkK crossover is the second option.
To save money - get a small bit of steel - a bolt or anything similar that isn`t very large, drop it in the air core of the inductor and see if you like it better. Increasing the BSC would yield you an even lower sensitivity, which is already pretty low at 85db/2.83V.
Idunn
A bit offtopic, but I didn`t really understand what Seas were trying to achieve with the release of a woofer that can`t go very deep, has below average sensitivity and costs same or more than comparable products from SB Acoustics, Tymphany, Wavecor and Monacor.
To save money - get a small bit of steel - a bolt or anything similar that isn`t very large, drop it in the air core of the inductor and see if you like it better. Increasing the BSC would yield you an even lower sensitivity, which is already pretty low at 85db/2.83V.
Idunn
A bit offtopic, but I didn`t really understand what Seas were trying to achieve with the release of a woofer that can`t go very deep, has below average sensitivity and costs same or more than comparable products from SB Acoustics, Tymphany, Wavecor and Monacor.
Last edited:
The problem, or subtlety, with the SEAS U18RNX-P is actually that it is a low inductance woofer, due to the Copper damping rings.
If you want to get the best out of it, you really need a third order bass filter as in this SB Acoustics example: SB Acoustics :: Eka
Someone should try this.
If you want to get the best out of it, you really need a third order bass filter as in this SB Acoustics example: SB Acoustics :: Eka
Someone should try this.
You definitely do not need to go third order with low inductance woofers. You can quite easily end up with a 4th order target using an oversized primary inductor (for BSC) and a zobel/notch. Sometimes this works, sometimes it does not. Sometimes you don't get enough response shaping around the knee of the xover and need to go true second order and sometimes you do need third. More often than not though you can get by without. But like with all things crossover and loudspeaker related, it all depends.
Looking at the Idunn's frequency response though, it isn't a surprise that people don't like it. I don't necessarily see there being a problem with only using 3dB of baffle step compensation as that gives you options. Sure it wont suit being free in a room, preferring to be 30cm or so from a back wall, but with only 3dB you can get a close wall placement without it sounding too over bearing. This is especially so with the woofer having very low distortion in the bass.
The bit that mainly concerns me with the Idunn is the ever so slight tendency towards a rising/tilted up frequency response. This is contrary to what people usually prefer. It has a general 1-2dB rise from the low end, right up to the top. If it was tilted the other way, as in a 1-2dB drop from the low end, to the top, then I don't think we'd be hearing any problems.
Waveguides especially seem to suit more slightly tilted down tonal balances, or rather they are prone to sounding too hot and in your face if they aren't turned down enough. Usually if a standard dome is turned up a little things just sound a bit bright and not unpleasant, whereas with the DXT this isn't the case. Tilt it up and things get unpleasant, whereas tilt it down and you don't lose air like you normally do with a standard dome.
The first thing I would do with the Idunn is to decrease the value of R2 a little (you can go as low as 5 ohms if it's needed). This should lessen the output of the tweeter whilst leaving its overall response shape intact. With my waveguide designs I have always needed to fine tune the tweeter level until things sound right. They require a slightly different approach than normal domes.
Looking at the Idunn's frequency response though, it isn't a surprise that people don't like it. I don't necessarily see there being a problem with only using 3dB of baffle step compensation as that gives you options. Sure it wont suit being free in a room, preferring to be 30cm or so from a back wall, but with only 3dB you can get a close wall placement without it sounding too over bearing. This is especially so with the woofer having very low distortion in the bass.
The bit that mainly concerns me with the Idunn is the ever so slight tendency towards a rising/tilted up frequency response. This is contrary to what people usually prefer. It has a general 1-2dB rise from the low end, right up to the top. If it was tilted the other way, as in a 1-2dB drop from the low end, to the top, then I don't think we'd be hearing any problems.
Waveguides especially seem to suit more slightly tilted down tonal balances, or rather they are prone to sounding too hot and in your face if they aren't turned down enough. Usually if a standard dome is turned up a little things just sound a bit bright and not unpleasant, whereas with the DXT this isn't the case. Tilt it up and things get unpleasant, whereas tilt it down and you don't lose air like you normally do with a standard dome.
The first thing I would do with the Idunn is to decrease the value of R2 a little (you can go as low as 5 ohms if it's needed). This should lessen the output of the tweeter whilst leaving its overall response shape intact. With my waveguide designs I have always needed to fine tune the tweeter level until things sound right. They require a slightly different approach than normal domes.
I have build a few of these (in) in a 16 liter curved chassis and experienced the same issue with the rising frequency response. I have modified the lead (R1 to 1.5 ohm and R2 to 5.6 ohm) and this works a whole lot better. It also greatly depends on the type of cap being used. I have used Audyn MCAP plus and Clarity Cap SA. I personally like the Clarity Cap SA most, but admit i think similar results (or better) can be achieved with the Audyn by further tuning the lpad.
Amazing the original poster alhe raised the issue then said he really hardly had time to look at the thread!
Apart from an undersized 12L box, the rising response or slope of a 6 inch bass is the perennial issue. It shouts around 500-1000Hz. IIRC, Troels Gravesen often applies an LCR around 700Hz to his 6" designs for this reason.
Studio-101
Applying a bigger bafflestep coil just takes the crossover lower without really giving you good rolloff.
You can, of course use a 4.5kHz LCR as the bass shunt shown below. It works a tidge better and phase improves. You can ignore my different tweeter filter. I'd regard tweeter level in general as an easy adjustment.
Actually the SEAS ER18RNX and SEAS U18RNX/P are very similar. So designs should work for either.
H1456-08 ER18RNX
H1571-08 U18RNX/P
Michael Chua did a good job on his Starling ER18RNX design, and you could use that bass filter with a slightly smaller Zobel correction on the U18RNX/P IMO.
"STARLING" (Seas 27TDFC + Seas ER18RNX) Bookshelf by AmpsLab
His bass filter does more to correct that troublesome upward slope. Some of my own modelling has worked well with a 4th order bass filter, but 6" bass is not something I wish to spend a lot of time on right now. It's a speaker that really doesn't work well on a simplistic filter.
Apart from an undersized 12L box, the rising response or slope of a 6 inch bass is the perennial issue. It shouts around 500-1000Hz. IIRC, Troels Gravesen often applies an LCR around 700Hz to his 6" designs for this reason.
Studio-101
Applying a bigger bafflestep coil just takes the crossover lower without really giving you good rolloff.
You can, of course use a 4.5kHz LCR as the bass shunt shown below. It works a tidge better and phase improves. You can ignore my different tweeter filter. I'd regard tweeter level in general as an easy adjustment.
Actually the SEAS ER18RNX and SEAS U18RNX/P are very similar. So designs should work for either.
H1456-08 ER18RNX
H1571-08 U18RNX/P
Michael Chua did a good job on his Starling ER18RNX design, and you could use that bass filter with a slightly smaller Zobel correction on the U18RNX/P IMO.
"STARLING" (Seas 27TDFC + Seas ER18RNX) Bookshelf by AmpsLab
His bass filter does more to correct that troublesome upward slope. Some of my own modelling has worked well with a 4th order bass filter, but 6" bass is not something I wish to spend a lot of time on right now. It's a speaker that really doesn't work well on a simplistic filter.
Attachments
Last edited:
I am currently experimenting more with the idunn l-pad. I currently have 2.2 ohm in series and 4.7 ohm parallel/shunt. This certainly seems to tame the DXT the right way...but...still finetuning...i am doing all of this by ear only right now...so it usually requires a few hours of listening.
Hi,
There is nothing very unusual in Seas parameter set for the
U18RNX/P, it is comparable to many other high quality 6.5"ers.
rgds, sreten.
And...frankly....the U18 sounds really nice in the mid-frequency...i like this woofer a lot. It could use a bit more 'kick' perhaps compared to the L18RNXP, but it does go pretty deep imho....
Hi everybody.
I have build the seas idunn kit, u18rnxp woofer but scanspeak 2905/97000 tweeter.
Project would be a opera callas 2014 clone, with the difference that my woofer is 8 ohm and callas one is 4 ohm.
Cabinet is 20 liters, chamfers on all sides.
Crossover is like operas one, only woofer arm is different and is the same as the iduun).
The sound is quite forward on the 1-2khz side.
Maybe due to the tweeter.
I would ask any help if somebody has realized the same build.
Regards
I have build the seas idunn kit, u18rnxp woofer but scanspeak 2905/97000 tweeter.
Project would be a opera callas 2014 clone, with the difference that my woofer is 8 ohm and callas one is 4 ohm.
Cabinet is 20 liters, chamfers on all sides.
Crossover is like operas one, only woofer arm is different and is the same as the iduun).
The sound is quite forward on the 1-2khz side.
Maybe due to the tweeter.
I would ask any help if somebody has realized the same build.
Regards
Everybody grumbles about the SEAS Idunn being thin on bass and bright in the midrange!
I gave up on the idunn setup, i have a feeling the x-over is a theoretic design only, i can't imagine that Seas actually likes this themselves. It needs serious redesign of the filter and doing measurements starting from scratch. I have built the design from Audioexcite using the dxt and u16 mid woofer. Very simplistic x-over. The sound is exactly what i was looking for. It amazes me everyday, especially the phase-behaviour (integration of tweeter and woofer) is brilliant.
Attachments
Last edited:
Everybody grumbles about the SEAS Idunn being thin on bass and bright in the midrange!
I think drastic measures are needed. I'd fit a bigger bafflestep coil and add some resistance to a smaller shunt capacitor. I really don't see much benefit in a RLC shunt here, nor in that complex treble filter.
This seems to work on phase well enough while reducing the midrange energy 2dB at least. You'll probably need to reduce the treble energy too. 3.3 or 4.7R series ought to do it. Impedance remains good.
Man designs after my own heart.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.