After making a dozen crazy test boxes out of pvc foam board and hot glue I more or less settled on a new 3 way design around the Peerless TC9FD. I totally fell in love with this driver. It sounds so clean and detailed. And even with my cobbled together test boxes I get holographic imaging on some tracks. In the track "Bubbles" by Yosi Horikawa I actually hear sound coming from behind me.
Even though the TC9FD is a great driver often referred to as full range I would call it a wide bander. The bass is impressive for it's size but it doesn't handle a lot of power before seriously distortng. I often see it used without a tweeter because it reaches impressively high but it sounds a lot better with a tweeter.
So for the top end I use a Peerless DX20BF20-08 crossed over first order at 6000HZ.
I use 4 TC9FD 8ohm in a WMMMTM where the top midrange is open baffle a la Nola Brio. I found that this greatly enhances imaging, spaciousness, realism, sparkle.....all the good stuff. This one open baffle driver runs without low pass filter. The three other midrange drivers are in a closed transmission line a la XRK's Nautaloss.
The bottom end is crossed over 1st order to a Peerless SDS160F25PR01-08 in a open tapered transmission line that reaches into the low 30s with room gain.
For the cabinets I have been experimenting with constrained layer damping. I noticed in other speakers that even fairly heavy MDF is a complete acoustic sieve.
My biggest headache is the crossover. Even though 1st order is the simplest crossover there is I am struggling with how to handle multiple midrange drivers.
Question 1 - I read two drivers add 3dB sensitivity. In my case that doubles again so my four drivers will be 6dB more sensitive compared to using only one? So my 84dB driver goes to 90dB when using 4. Or does it also depend on how they are wired? 90 dB would be a little high compared to my 87db woofer.
Question 2 - If I wire them series/parallel so i keep it 8ohm is it still possible to roll of three drivers at 6000Hz and keep one running full range?
If I can get some pointers on how to approach the crossover I can play with it in XSim.
Even though the TC9FD is a great driver often referred to as full range I would call it a wide bander. The bass is impressive for it's size but it doesn't handle a lot of power before seriously distortng. I often see it used without a tweeter because it reaches impressively high but it sounds a lot better with a tweeter.
So for the top end I use a Peerless DX20BF20-08 crossed over first order at 6000HZ.
I use 4 TC9FD 8ohm in a WMMMTM where the top midrange is open baffle a la Nola Brio. I found that this greatly enhances imaging, spaciousness, realism, sparkle.....all the good stuff. This one open baffle driver runs without low pass filter. The three other midrange drivers are in a closed transmission line a la XRK's Nautaloss.
The bottom end is crossed over 1st order to a Peerless SDS160F25PR01-08 in a open tapered transmission line that reaches into the low 30s with room gain.
For the cabinets I have been experimenting with constrained layer damping. I noticed in other speakers that even fairly heavy MDF is a complete acoustic sieve.
My biggest headache is the crossover. Even though 1st order is the simplest crossover there is I am struggling with how to handle multiple midrange drivers.
Question 1 - I read two drivers add 3dB sensitivity. In my case that doubles again so my four drivers will be 6dB more sensitive compared to using only one? So my 84dB driver goes to 90dB when using 4. Or does it also depend on how they are wired? 90 dB would be a little high compared to my 87db woofer.
Question 2 - If I wire them series/parallel so i keep it 8ohm is it still possible to roll of three drivers at 6000Hz and keep one running full range?
If I can get some pointers on how to approach the crossover I can play with it in XSim.
Q1 You read wrong. Only two independent identically and separately driven drivers are 3dB louder than a single one. If you wire them in parallel you also halve the impedance and this add also (up to) 3dB, for a total of (up to) 6dB. If you wire them in series you have +3dB from the second driver but also -3dB for doubling the impedance for a net of zero. If you wire 4 divers in series/parallel your net is +6dB
Q2 Doesn't make much sense, what are you trying to achieve? You'll get comb filtering at high frequencies.
In general the sensitivity for a multi-way is the one of the woofer, your mid and tweeter need to be more sensitive than that, and will be attenuated (for example with an L-pad). Keep also in mind that a sensitivity spec for a driver is for an infinite baffle, but under a frequency related to the baffle a woofer radiates in the full space losing gradually 6dB under that frequency and for a woofer that means that you need to compensate for this effect (baffle step), losing 3-6dB of the nominal sensitivity. If not compensated the bass will be thin.
1st order (electrical) crossover has nothing magical, it is only harder to get right than a 2nd or 4th order (acoustical) crossover.
Ralf
Q2 Doesn't make much sense, what are you trying to achieve? You'll get comb filtering at high frequencies.
In general the sensitivity for a multi-way is the one of the woofer, your mid and tweeter need to be more sensitive than that, and will be attenuated (for example with an L-pad). Keep also in mind that a sensitivity spec for a driver is for an infinite baffle, but under a frequency related to the baffle a woofer radiates in the full space losing gradually 6dB under that frequency and for a woofer that means that you need to compensate for this effect (baffle step), losing 3-6dB of the nominal sensitivity. If not compensated the bass will be thin.
1st order (electrical) crossover has nothing magical, it is only harder to get right than a 2nd or 4th order (acoustical) crossover.
Ralf
I just said I am using Xsim. But what good does it do me to just drag some components onto the screen and cobble them together like a 6 year old without having any idea why...... I thought the function of a forum like this was to learn from each other and elevate our knowledge beyond kindergarten level. Regarding crossovers I am clearly at kindergarten level!
I read many explanations but never really got it. Your explanation is great and I get it now. Thanks!
If it was about making sense I would buy a speaker and be done. I like the alchemy of making my own. In my experiments I found that a speaker in a box gains balls/muscle/control but loses something too. Then I stumbled on the Nola Brio speaker using one driver in a BR with the same driver on top OB. When I just put one driver naked on top of my test boxes a certain magic is added. But here is where my problems start.
Most of this magic happens in the upper octaves so I cannot cross the extra OB driver over to the tweeter together with the other 3. I realize that comb filtering might be an issue but I doubt that if these drivers are mounted close that is an issue. The tweeter faceplate is very small and it has mild horn loading so it is naturally time aligned with the mid. Have you ever seen the very well reviewed Tekton speaker? They use 7! tweeters. 6 in a circle with one in the middle. I am sure many would laugh at the idea of 7 tweeters but they made it and clearly many people like the result. I own a pair of very well engineered (by well engineered i mean it was made by well educated engineers with proper equipment, testing facilities and a good budget) and well reviewed JBL 305P studio monitors and they suck. They just murder music, suck the life right out of it. So doing things right is not always the best. That is why audio is so interesting.
While writing this I just realized I never tried the simplest solution. Just treat that OB driver as a second tweeter and cross them over together....I will play with that.
Last edited:
Erik,
Sure a first order filter can be simple. But it's very hard to get a real 1st. order to behave properly. Check out a Thiel crossover. That doesn't look to simple or easy. I'd try a second order LKR filter as a good basic starting point. Remember you also have to account for the natural roll off frequencies of the individual drivers.
Thiel first order crossover network - Google Search
Sure a first order filter can be simple. But it's very hard to get a real 1st. order to behave properly. Check out a Thiel crossover. That doesn't look to simple or easy. I'd try a second order LKR filter as a good basic starting point. Remember you also have to account for the natural roll off frequencies of the individual drivers.
Thiel first order crossover network - Google Search
I realize 1st order crossover has some downsides but everything in engineering is a compromise. In my case i think it is a good compromise.
The TC9FD is a special driver because it is almost full range. This gives a lot of margin in the crossover design. If a razor flat frequency was my only goal it would probably not be a good idea but hey, we listen in rooms not anechoic chambers. The TC9FD can run without a tweeter but from about 6000/8000Hz the driver becomes less clean. Nothing majorly offensive but not ideal. With large midrange drivers there would be beaming at high frequencies but since the TC9FD is only 3.5" this is not a problem. So there is an incredibly wide marging for a 1st order crossover. I see no need to complicate it unnecessarily.
At the low end there is also a good overlap. I would cross it over as low as possible, just enough to reduce midrange excursion to get decent power handling.
I am not stuck on the idea of 1st order but I don't see major downsides with my driver combination. And since I am in Thailand where quality components are sometimes hard to find a first order has extra upside.
I made several boxes in different sizes, differently stuffed, some closed, some closed TL some open TL.
Closed boxes are very well behaved but sound uninspiring to me. Open TL adds some air and space around the music....it sounds less constrained. But I found the combination of closed TL + open baffle to be the most realistic sounding. It adds the "I am actually there" feeling. This was the most noticeable on live recordings in large venues. You can sense the space. On some recordings it is subtle but sometimes the difference is day and night. I listened to a recording of Formula 1 race car sounds with and without the extra OB speaker. The sealed + OB combo just oozes space/realism/excitement. By comparison the closed box and to a lesser extent the open TL sounds lifeless and dull....you are not really there anymore.
So this fascination with the added OB speaker created a XO headache for me because just a few weeks ago I hardly knew anything about crossovers. So any help/education here is welcome.
OK, It seems I have not been applying enough logic and misinterpreting what my ears were hearing. The extra OB speaker raises the high frequency response since it has less bass response than the TL mid. So a lot of that spaciousness was just a simple boost in the high frequency response.
I will do some more tests but it seems I can scrap the unnecessary OB idea.
If this is the case I can build the boxes, get measurements and import them to Xsim to play around with crossover variants.
I will do some more tests but it seems I can scrap the unnecessary OB idea.
If this is the case I can build the boxes, get measurements and import them to Xsim to play around with crossover variants.
Malibran - Unison Research - Official Website
some writings from the author about the presence of rear firing tweeters and the specialized crossover technique
il CLD (Cross-Link-Dipole) e le sorgenti ausiliarie
Mario Bon - Filtro Progressivo
some writings from the author about the presence of rear firing tweeters and the specialized crossover technique
il CLD (Cross-Link-Dipole) e le sorgenti ausiliarie
Mario Bon - Filtro Progressivo
Last edited:
After reading countless conflicting articles and posts I was still completely confused. So I decided the only way was experimenting and using my own ears. I built test boxes with 2 TC9FD in closed stuffed transmission lines and a tweeter and another tweeter out of phase on the back. I soon concluded i didn't care for the sound at all. It was dull and tiring.
My open transmission line sounds a lot better. Then I decided to try TC9FD on an open baffle. They had a little less bass but the sound is more airy, relaxed and realistic. Also the lack of treble seemed to have vanished. To make sure it is not an illusion I connected both with a 2.2uf cap highpass and A/B tested the sound with different tracks. A huge and striking difference. The open baffle sounds so much more realistic and articulate. Then I A/B/C tested closed TL, open baffle and a baffle mounted tweeter (peerless dx20bf) with 1.5uf caps and to my surprise the OB TC9FD had as much treble as the tweeter and the closed TL was quite lacking in treble. Furthermore the tweeter had a tinny quality that was totally absent in the OB driver.
So all this experimenting gave me a new direction. It seems I really like the open baffle sound and it also seems like the TC9FD does not need a tweeter when used in an open baffle.
So my new plan is a 2way open baffle with 4x TC9FD in a mini line array combined with a 15" or 2x10" high Q bass driver. Finding suitable drivers will probably be quite a headache. The Eminence 15a is not available in Thailand. The only driver I found so far is the GRS 12PR-8 12" Poly Cone.
I would prefer a paper cone.... Does anybody have experience with this GRS driver in an open Baffle? It is a very cheap driver. Will it perform on par with Eminence or sound clearly inferior?
My open transmission line sounds a lot better. Then I decided to try TC9FD on an open baffle. They had a little less bass but the sound is more airy, relaxed and realistic. Also the lack of treble seemed to have vanished. To make sure it is not an illusion I connected both with a 2.2uf cap highpass and A/B tested the sound with different tracks. A huge and striking difference. The open baffle sounds so much more realistic and articulate. Then I A/B/C tested closed TL, open baffle and a baffle mounted tweeter (peerless dx20bf) with 1.5uf caps and to my surprise the OB TC9FD had as much treble as the tweeter and the closed TL was quite lacking in treble. Furthermore the tweeter had a tinny quality that was totally absent in the OB driver.
So all this experimenting gave me a new direction. It seems I really like the open baffle sound and it also seems like the TC9FD does not need a tweeter when used in an open baffle.
So my new plan is a 2way open baffle with 4x TC9FD in a mini line array combined with a 15" or 2x10" high Q bass driver. Finding suitable drivers will probably be quite a headache. The Eminence 15a is not available in Thailand. The only driver I found so far is the GRS 12PR-8 12" Poly Cone.
I would prefer a paper cone.... Does anybody have experience with this GRS driver in an open Baffle? It is a very cheap driver. Will it perform on par with Eminence or sound clearly inferior?
I built MTM with TC9FD, great sounding little speakers. Once put on top of the woofers, ideally biamplified, can fully satisfy.
MTM with TC9fd
MTM with TC9fd
Thanks Adason. I didn't notice the severe beaming of the TC9FD. They have plenty of treble on axis but even 20 deg off axis it virtually disappears. Can't believe I needed your post to figure this out!
So I guess I do need a tweeter. I am playing with the idea of a line array with 4x TC9FD and 8x TC5FB (or 4xTC5FC). These are 1.5" mini versions of the TC9FD. they measure quite good off axis and should blend well.....
So I guess I do need a tweeter. I am playing with the idea of a line array with 4x TC9FD and 8x TC5FB (or 4xTC5FC). These are 1.5" mini versions of the TC9FD. they measure quite good off axis and should blend well.....
But we do not listen only on axis...even if we are sitting on axis!
You hear direct arriving signal first. Then few milliseconds later you hear floor bounce, then few tens of milliseconds later you hear side walls reflection, back wall reflection and so on...
If the full range has significantly different fr response on axis and off axis, then there will be significant difference in direct and reflected sound. Such speakers do not sound natural. Best sounding speakers have uniform off axis response. This has been discussed ad nauseam and proven true.
I recently measured and posted such case.
Suitable Enclosure for TB W5 1611 SAF
You hear direct arriving signal first. Then few milliseconds later you hear floor bounce, then few tens of milliseconds later you hear side walls reflection, back wall reflection and so on...
If the full range has significantly different fr response on axis and off axis, then there will be significant difference in direct and reflected sound. Such speakers do not sound natural. Best sounding speakers have uniform off axis response. This has been discussed ad nauseam and proven true.
I recently measured and posted such case.
Suitable Enclosure for TB W5 1611 SAF
I am a little confused by your post. Are you implying using 1.5" fullrange drivers as tweeters is not a good idea or confirming that this is a sensible idea?
An externally hosted image should be here but it was not working when we last tested it.
I plan to cross over the TC9FD's to the TC5FB's around 5000HZ before the TC9FD starts beaming. The dispersion of the TC5FB is comparable to that of a good dome tweeter but has the benefit of being a dipole and having the same sound signature and sensitivity as the midrange drivers.
A mini array of 4 midrangers can have 4 operating from 400 Hz upward. The top and bottom TC9's can roll of early so at the crossover the 2 middle ones hand over to the 4 tweeter array that is exactly as tall and right next to the 2 middle TC9's.
I ordered 2 of these 10 inchers to play around with:
They are only 11 US$ each so no tears if they suck.....
They have reasonably high QTS and and low Fs. I only wish they had a vented pole. But probably the xmax is so low it is not needed. If these do the job I can use 4 or maybe even 6 per side.
They are only 11 US$ each so no tears if they suck.....
They have reasonably high QTS and and low Fs. I only wish they had a vented pole. But probably the xmax is so low it is not needed. If these do the job I can use 4 or maybe even 6 per side.
Last edited:
No, I am not implying anything. I was talking about fullrange, not tweeter. I have no experience with your 1.5" fullrange.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.