Once again, Dave, you sound like you are way ahead of me. I am just grappling with importing those files for a predictable result. And it's not easy, because manufacturers are very inconsistent and the simulators are often hard work too.
I thought useful tweeter files were always done on infinite baffle, or at least IEE in anechoic chamber. Takes all the diffraction and bafflestep issues away. Way beyond my capabilities or inclination. Morel used to publish FRD and ZMA in a format they called CLIO measurements. Look up the defunct Morel MDT20 at parts express if you are interested. It was considered a drop in replacement for the SEAS H087 in the Dynaco A25.
I can usually get an idea how a speaker will behave from the published curves. I know SEAS use a standard box, but I assume they do tweeters on IEE (about 2 metres square) baffle.
The 8" U22 might surprise me, like the Volt driver. It looks like it might bend the frequency response at 800Hz from the impedance bump, as woofers often do. I know how to fix that with an LCR, but it really needs accurate modelling.
H1659-08 U22REX/P-SL
The waveguided DXT tweeter might be what you need with that 10" two way. You take your beating on dispersion. Joachim Gerhard rates it highly on simple filter, but I really don't know. Looks like it might be ferrofluid, which usually helps at low frequency crossover, but I'm not sure. Again it needs modelling, because it is nowhere near flat.
H1499-06 27TBCD/GB-DXT
I thought useful tweeter files were always done on infinite baffle, or at least IEE in anechoic chamber. Takes all the diffraction and bafflestep issues away. Way beyond my capabilities or inclination. Morel used to publish FRD and ZMA in a format they called CLIO measurements. Look up the defunct Morel MDT20 at parts express if you are interested. It was considered a drop in replacement for the SEAS H087 in the Dynaco A25.
I can usually get an idea how a speaker will behave from the published curves. I know SEAS use a standard box, but I assume they do tweeters on IEE (about 2 metres square) baffle.
The 8" U22 might surprise me, like the Volt driver. It looks like it might bend the frequency response at 800Hz from the impedance bump, as woofers often do. I know how to fix that with an LCR, but it really needs accurate modelling.
H1659-08 U22REX/P-SL
The waveguided DXT tweeter might be what you need with that 10" two way. You take your beating on dispersion. Joachim Gerhard rates it highly on simple filter, but I really don't know. Looks like it might be ferrofluid, which usually helps at low frequency crossover, but I'm not sure. Again it needs modelling, because it is nowhere near flat.
H1499-06 27TBCD/GB-DXT
I use an ancient piece of software called speaker workshop. But I find it quite good and as long as you don't put garbage in your simulation and measured results can be very close indeed
As for the maths. I have no understanding at all. I'm really just a trial and error hack but have fun. I'd rate myself as a novice. Although I almost like the 3 ways I'm building. Post to come
As for the maths. I have no understanding at all. I'm really just a trial and error hack but have fun. I'd rate myself as a novice. Although I almost like the 3 ways I'm building. Post to come
The 27tbfcg is awful on some recordings. It's a pity as I love certain music and the 27tbfcg makes it unlistenable. It's sister the 27tdfcg with a soft dome is better but not tolerable. I have a pair of 27tdfc (without grille and acoustic lens) to try to serif that helps. It doesn't really matter what crossover slope or point or volume level. They shout on certain recordings. Will try a sloppier tweeter too.
I'm surprised you are having problems with such a highly regarded tweeter. Metal tweeters have a harder sound than soft domes, but I like what they do. Very good on electric guitar, for instance. In fact, Harbeth use that tweeter, and they know what they are doing.
It's often hard to trace distortion. It can be breakup from the bass, or too much overlap in the drivers producing a peak. Put your hand over the tweeter, and listen to the bass on a troublesome passage of music.
A full or half zobel (15R + 0.33uF) is essential to tame the top end IMO. They then don't sound metallic at all. And everything works well with a 4th order filter. I really wouldn't go back to second order these days.
Use 6 ohms and 3500Hz and BW4 or LR4 in this for an estimate of good values to use.
2-Way Crossover Calculator / Designer
Online calculators do a good job at this. You'll get good slopes and impedance. It's how it works.
It's often hard to trace distortion. It can be breakup from the bass, or too much overlap in the drivers producing a peak. Put your hand over the tweeter, and listen to the bass on a troublesome passage of music.
A full or half zobel (15R + 0.33uF) is essential to tame the top end IMO. They then don't sound metallic at all. And everything works well with a 4th order filter. I really wouldn't go back to second order these days.
Use 6 ohms and 3500Hz and BW4 or LR4 in this for an estimate of good values to use.
2-Way Crossover Calculator / Designer
Online calculators do a good job at this. You'll get good slopes and impedance. It's how it works.
Last edited:
The Volt 8" drivers are ~89dB above baffle step. So if you allow for ~ 4dB of bafflestep and you parallel 2 drivers, you're looking at 89 - 4 + 6 = 91 dB sensitive system.
the 27TBFC/G is about 91 dB sensitive - so you will likely not need to pad the tweeter.
Are you looking at the 8 or 16 ohm volt drivers? If you can tell me which, I can sim a crossover for you and get @system7 to pick it over.
PS: I'm not sure the 16 version will have the same sensitivity... won't it reduce by ~ 3dB?
I'm going to be trying different things this weekend. I don't think it is a slope or xover point problem. I've tried a 2.5Khz > 4th order slope and the sound on certain instruments / recordings is the same forwardness at low or moderate volume levels.
I'm going to try the following:
1. Keep pushing the XO up - although many people have crossed over at 2Khz, 4th order acoustic and claim no issues (depends on their source and recordings I suppose)
2. Try your half-zobel
3. Try good old impedance compensation
4. Notch the offending frequency.
First... I'm going to measure
Been lazy and haven't so who knows whether I'm hitting the sim!As an example of Seas 27TDFC/G or 27TBFC/G use - here is the D version.
The blue tweeter is my ideal from a flat FR and hits a 2KHz 4th order slope (with some attentuation) as per the white goal line
The pink is what I am currently listening to. I've listened to both blue and pink and hear the same harshness
I've also setup the cyan level and the same. this is a nominal 2.5kHz crossover. I recall listening to this and same results.
Attachments
Without seeing the crossover or knowing what the other drivers are, it is impossible to comment really.
Looks to me like you might be too enthralled with an LR4 acoustic target slope and crossing unnecessarily low around 2kHz, with possibly a huge and irrelevant reverse null. The perfect storm for tweeter distortion. And I can't see any point in your exotic tapered cabinet, TBH.
But it could be cone breakup from the mystery mid as well.
There are some very impressive midbasses that can be used in a 3 way which just don't have nasty breakups.
Peerless 830870 and 830860 for instance.
Solitaire - Peerless 830870 HDS 4" Midrange with Morel MDT29 - AmpsLab
Michael Chua is always interesting.
I can't see any real difference between these two SEAS tweeters. Interchangeable to the max IMO.
H1212-06 27TBFC/G
H1189-06 27TDFC
This is a very good tweeter filter. Naturally you can tweak it. There is a case for the 0.22mH being a 0.15mH, and the input resistor is adjustable for level. It's near LR4 acoustic but crucially it's also near LR4 electrical too. Sounds good with soft and metal domes. I'm using a Morel MDT29 right now.
If you avoid phase with the midbass being more than 90 degrees out near crossover, it will sound OK.
Looks to me like you might be too enthralled with an LR4 acoustic target slope and crossing unnecessarily low around 2kHz, with possibly a huge and irrelevant reverse null. The perfect storm for tweeter distortion. And I can't see any point in your exotic tapered cabinet, TBH.
But it could be cone breakup from the mystery mid as well.
There are some very impressive midbasses that can be used in a 3 way which just don't have nasty breakups.
Peerless 830870 and 830860 for instance.
Solitaire - Peerless 830870 HDS 4" Midrange with Morel MDT29 - AmpsLab
Michael Chua is always interesting.
I can't see any real difference between these two SEAS tweeters. Interchangeable to the max IMO.
H1212-06 27TBFC/G
H1189-06 27TDFC
This is a very good tweeter filter. Naturally you can tweak it. There is a case for the 0.22mH being a 0.15mH, and the input resistor is adjustable for level. It's near LR4 acoustic but crucially it's also near LR4 electrical too. Sounds good with soft and metal domes. I'm using a Morel MDT29 right now.
If you avoid phase with the midbass being more than 90 degrees out near crossover, it will sound OK.
Hi Dave, I’m using 8 ohm Volt BM220.8 drivers. They differ from their latest revision which is the BM 228.8. Which I believe Volt modified for the likes of Robson audio et al as Volt are their Oem suppliers.
Without seeing the crossover or knowing what the other drivers are, it is impossible to comment really.
Looks to me like you might be too enthralled with an LR4 acoustic target slope and crossing unnecessarily low around 2kHz, with possibly a huge and irrelevant reverse null. The perfect storm for tweeter distortion. And I can't see any point in your exotic tapered cabinet, TBH.
But it could be cone breakup from the mystery mid as well.
There are some very impressive midbasses that can be used in a 3 way which just don't have nasty breakups.
Peerless 830870 and 830860 for instance.
Solitaire - Peerless 830870 HDS 4" Midrange with Morel MDT29 - AmpsLab
Michael Chua is always interesting.
I can't see any real difference between these two SEAS tweeters. Interchangeable to the max IMO.
H1212-06 27TBFC/G
H1189-06 27TDFC
This is a very good tweeter filter. Naturally you can tweak it. There is a case for the 0.22mH being a 0.15mH, and the input resistor is adjustable for level. It's near LR4 acoustic but crucially it's also near LR4 electrical too. Sounds good with soft and metal domes. I'm using a Morel MDT29 right now.
If you avoid phase with the midbass being more than 90 degrees out near crossover, it will sound OK.
Hi Steve, I’ll come clean. I’m a sucker for those shiny polycones that you see in Harbeths. The volts are as close as I’ve got , and yes, I know they’re not the same and as it turns out they’re problematic but they’re handmade here in England.
We were getting a bit off-topic with Dave Bullet's project. Perhaps we should focus on this 8" Volt thing. It is slightly problematic on sensitivity with the Scan tweeter. Not much wiggle room on level. Which gives you less wiggle room on the filter too.
Harbeth, Spendor and Stirling Broadcast and another company I have forgotten right now, all make traditional BBC style monitors.
Stirling Broadcast - LS3/6
Alan Shaw is interesting:
YouTube
The 2 cubic foot rectangular box with lossy construction seems to work well enough for 8" bass. You cross over at 3kHz on negative polarity, which is why the bass is often recessed. Reflex because it works well with classical music. Everybody seems to use SEAS tweeters, either soft or metal. But modern low breakup bass units, more rigid than the old polycones. Often phase plugs too, since dustcaps sing their own tunes.
From my own experiments, you can use 3rd or 4th order electrical filters on the tweeter. Better dispersion or lower distortion? Choose your poison. The bass shouldn't need anything more than second order to play nicely.
The SEAS U22 looks like the closest off-the-shelf equivalent of these 8" units. But like the Volt, it takes me outside my experience in the bass filter. No-one here is helping, so it's a question of trying things IMO. But obviously there IS a solution that works. So I don't think the bass filter really holds your project up.
Harbeth, Spendor and Stirling Broadcast and another company I have forgotten right now, all make traditional BBC style monitors.
Stirling Broadcast - LS3/6
Alan Shaw is interesting:
YouTube
The 2 cubic foot rectangular box with lossy construction seems to work well enough for 8" bass. You cross over at 3kHz on negative polarity, which is why the bass is often recessed. Reflex because it works well with classical music. Everybody seems to use SEAS tweeters, either soft or metal. But modern low breakup bass units, more rigid than the old polycones. Often phase plugs too, since dustcaps sing their own tunes.
From my own experiments, you can use 3rd or 4th order electrical filters on the tweeter. Better dispersion or lower distortion? Choose your poison. The bass shouldn't need anything more than second order to play nicely.
The SEAS U22 looks like the closest off-the-shelf equivalent of these 8" units. But like the Volt, it takes me outside my experience in the bass filter. No-one here is helping, so it's a question of trying things IMO. But obviously there IS a solution that works. So I don't think the bass filter really holds your project up.
I'll start another post for my all metal 3 way. I hope you're up for the challenge. No more of the poly cone flabby 2 ways. You wouldn't believe the clarity.
I found your intended baffle in previous posts - thanks.
Did you intent to post FRD / ZMA files above? I can't see any attachments.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.