Hi Folks,
I wasn't quite sure where to put this, but as my project is a 2 way open baffle, and so is a Magnepan SMGa, I though this forum was a fair choice.
I've long enjoyed my SMGa speakers. On seeing and reading about OB designs, an obvious thought came to mind - How much of what I like in the Magnepan comes from the panel drivers, and how much is due to the lack of a box?
Time to experiment, on a reasonable budget. Fortunately there is decent plywood left over from a shelving project. A circle cutting jig was constructed. For drivers, I chose the common starting point of the Eminence Alpha 15A for bass. For mid/treble, I was a bit stumped. Wanting a simpler 2 way design, there isn't much to cover the lower mid-treble with one unit.
Then I recalled that I'd saved the two 6" full range drivers from some old Frazier speakers (cabinets were severerly damaged, but drivers OK). This made for a starting point to explore
Baffles were cut, 18" wide x 36 tall, and holes cut for the drivers. THey tilt backThe 6" units are offset, to keep the distance from the top and two sides all different. The crossover is a guess, aiming for around 1 kHz, first order.
The early impressions are encouraging, and there is progress to be made.
On the negative side:
The "full range" drivers have a somewhat peaky midrange response which would be good to tame. Their high treble is also perhaps not the clearest (roll-off and or distortion). They are also more efficient than the Eminence units. This is noticeable more on some recordings than others.
On the plus side:
There is a very favourable comparison to the character of the Magnepans, if the negatives are put aside. Transient response and bass presentation is very musical. Imaging is great, and there is less beaming of the sound than with the panel speakers.
Next steps:
Attenuate the full-range units, at least near the crossover point to bring the levels in line with the woofer. I might add compensation, to counteract rolled off treble. Need to work on a measurement setup for this to quantify response and changes.
Also I'm looking at other alternatives for the full-range drivers, and may decide to experiment there, so thoughts are welcome.
Anyway, it is a fun and musical journey. My Wife tells me she likes the OB's more than the SMGa for both sound and looks. The latter is surprising, as they are just a board with speakers in at the moment!
Cheers
Ed
I wasn't quite sure where to put this, but as my project is a 2 way open baffle, and so is a Magnepan SMGa, I though this forum was a fair choice.
I've long enjoyed my SMGa speakers. On seeing and reading about OB designs, an obvious thought came to mind - How much of what I like in the Magnepan comes from the panel drivers, and how much is due to the lack of a box?
Time to experiment, on a reasonable budget. Fortunately there is decent plywood left over from a shelving project. A circle cutting jig was constructed. For drivers, I chose the common starting point of the Eminence Alpha 15A for bass. For mid/treble, I was a bit stumped. Wanting a simpler 2 way design, there isn't much to cover the lower mid-treble with one unit.
Then I recalled that I'd saved the two 6" full range drivers from some old Frazier speakers (cabinets were severerly damaged, but drivers OK). This made for a starting point to explore
Baffles were cut, 18" wide x 36 tall, and holes cut for the drivers. THey tilt backThe 6" units are offset, to keep the distance from the top and two sides all different. The crossover is a guess, aiming for around 1 kHz, first order.
The early impressions are encouraging, and there is progress to be made.
On the negative side:
The "full range" drivers have a somewhat peaky midrange response which would be good to tame. Their high treble is also perhaps not the clearest (roll-off and or distortion). They are also more efficient than the Eminence units. This is noticeable more on some recordings than others.
On the plus side:
There is a very favourable comparison to the character of the Magnepans, if the negatives are put aside. Transient response and bass presentation is very musical. Imaging is great, and there is less beaming of the sound than with the panel speakers.
Next steps:
Attenuate the full-range units, at least near the crossover point to bring the levels in line with the woofer. I might add compensation, to counteract rolled off treble. Need to work on a measurement setup for this to quantify response and changes.
Also I'm looking at other alternatives for the full-range drivers, and may decide to experiment there, so thoughts are welcome.
Anyway, it is a fun and musical journey. My Wife tells me she likes the OB's more than the SMGa for both sound and looks. The latter is surprising, as they are just a board with speakers in at the moment!
Cheers
Ed
I'd suggest a foray into the Full Range section. There are many choices for your mid / top and I suspect you will like a better quality unit and a lower crossover to the woofer. On the very cheap side there is the Goldwood 8 inch full range, higher up the food chain are the Mark Audio Alpair 7.3, 10.3, 11ms, and more. Or if you like the No Replacement for Displacement approach and have the budget the Commonsense Audio Super Cast 12 and 15 drivers are pretty nice too.
Thanks for the notes everyone.
I've been looking around all the speaker forums for ideas, and notes on crossovers etc. It's the first time ever assembling any kind of speaker project, so all the learning is fun.
It will be interesting to see what is possible with the Frazier drivers. They are well made 6 inchers with a large magnet, a half-roll rubber surround and a whizzer cone. Using them exclusively as a full range indicates that the artefacts heard in the 2 way design is largely from these units. Relieving them of bass duty helps to show their potential and taming their output to flatten the overall response will be interesting.
As to other FR drivers, it seems to make sense to aim for sensitivity comparable with the Eminence woofers ~94 dB/W if the spec is to be believed. That narrows the field somewhat, but there might be something worth trying. As they're cheap, the visaton BG17 or one of the Faital fullranges, for a bit more, seem like candidates.
Cheers
Ed
I've been looking around all the speaker forums for ideas, and notes on crossovers etc. It's the first time ever assembling any kind of speaker project, so all the learning is fun.
It will be interesting to see what is possible with the Frazier drivers. They are well made 6 inchers with a large magnet, a half-roll rubber surround and a whizzer cone. Using them exclusively as a full range indicates that the artefacts heard in the 2 way design is largely from these units. Relieving them of bass duty helps to show their potential and taming their output to flatten the overall response will be interesting.
As to other FR drivers, it seems to make sense to aim for sensitivity comparable with the Eminence woofers ~94 dB/W if the spec is to be believed. That narrows the field somewhat, but there might be something worth trying. As they're cheap, the visaton BG17 or one of the Faital fullranges, for a bit more, seem like candidates.
Cheers
Ed
One quick experiment was tried this weekend - A 2.2 Ohm resistor was added in series with the full range driver. It was possible to switch this in and out of circuit using a clip lead for rapid A/B comparison. ~2 Ohms was chosen as it (very) roughly equates to 3 dB reduction in power.
This did seem to help tame the sound, although the reduction in level was quite subtle. The added resistance did not seem to affect the higher treble as much as I would have expected. Listening was very enjoyable at medium and high levels.
Soon I'll have the means to make a more quantitative assessment of the response, and impact of circuit changes.
Cheers
Ed
This did seem to help tame the sound, although the reduction in level was quite subtle. The added resistance did not seem to affect the higher treble as much as I would have expected. Listening was very enjoyable at medium and high levels.
Soon I'll have the means to make a more quantitative assessment of the response, and impact of circuit changes.
Cheers
Ed
Over the weekend, I was able to start making measurements. Respect for people who do this sort of thing well increased even further.
The stumblings of a beginner established a little of what was the effect of a rear wall, a very sharp dip due to reflection from the (hard) floor, and a remaining imbalance between the full-range and bass driver output levels.
I also made a comparison measurement of the Magnepans, which was rather interesting. More to follow when there is a chance to upload.
One experimental change was some additional resistive padding of the 6" drivers, and this made a further improvement to the tonal balance, and from measurement, suggested a flatter response.
However, it still seems that the "full range" driver is not the flattest transducer, even as used. The setup can sound absolutely great on some recordings, not so much on others. This, I can now spend a bit of time upon, and have some fun evaluating.
Cheers
Ed
The stumblings of a beginner established a little of what was the effect of a rear wall, a very sharp dip due to reflection from the (hard) floor, and a remaining imbalance between the full-range and bass driver output levels.
I also made a comparison measurement of the Magnepans, which was rather interesting. More to follow when there is a chance to upload.
One experimental change was some additional resistive padding of the 6" drivers, and this made a further improvement to the tonal balance, and from measurement, suggested a flatter response.
However, it still seems that the "full range" driver is not the flattest transducer, even as used. The setup can sound absolutely great on some recordings, not so much on others. This, I can now spend a bit of time upon, and have some fun evaluating.
Cheers
Ed
You can just upload directly to diyaudio.
I also use a full range in OB at the moment. It's quite competent, but it took a lot of experimentation to make it sound good. First I used a lost of DSP power, but sounded awful, then I experimented with xo point which is now at 80 hz and only some soft eq'ing. I find hard eq'ing to often sound bad. But I never get really satisfied with this fullrange, before I used a 4 inch full rang crossed at 300 hz to several 10 inch woofers. Much easier to integrate with small full range, but the 10 inch woofers was not good at the very low end. The bottom line of all this, is that two way might seem easy, but it does come with compromises, and in the end might not be easier than three way.
I also use a full range in OB at the moment. It's quite competent, but it took a lot of experimentation to make it sound good. First I used a lost of DSP power, but sounded awful, then I experimented with xo point which is now at 80 hz and only some soft eq'ing. I find hard eq'ing to often sound bad. But I never get really satisfied with this fullrange, before I used a 4 inch full rang crossed at 300 hz to several 10 inch woofers. Much easier to integrate with small full range, but the 10 inch woofers was not good at the very low end. The bottom line of all this, is that two way might seem easy, but it does come with compromises, and in the end might not be easier than three way.
OK, here is the type of thing I am able to capture with Spectrum Lab*
This was a first attempt. Input to the speaker was "white noise" from an FM radio tuned to an empty channel. I was also able to use a signal generator to explore spot frequency levels. This verified that the FM radio idea was acceptable for a first bash.
Exploring room effects, I moved out to the centre of the room. The mic was level with the full range driver, and 2m from the baffle. This was taken before addition of (more) attenuation to the FR
And after adding resistance (total of 4 Ohms):
Yes, I know, the charts are very scratchy looking, and could use more smoothing. I may move from Spectrum Lab to REW, now that I know about the latter software. One motivation for this exercise is to learn some of the practical physics of speaker build/measure, as well as to build a pair of decent budget speakers.
Cheers,
Ed
This was a first attempt. Input to the speaker was "white noise" from an FM radio tuned to an empty channel. I was also able to use a signal generator to explore spot frequency levels. This verified that the FM radio idea was acceptable for a first bash.
Exploring room effects, I moved out to the centre of the room. The mic was level with the full range driver, and 2m from the baffle. This was taken before addition of (more) attenuation to the FR
And after adding resistance (total of 4 Ohms):
Yes, I know, the charts are very scratchy looking, and could use more smoothing. I may move from Spectrum Lab to REW, now that I know about the latter software. One motivation for this exercise is to learn some of the practical physics of speaker build/measure, as well as to build a pair of decent budget speakers.
Cheers,
Ed
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.