Hi Everybody,
I'm a new member and this is my first post here. I'm a long term audiophile but have never done a DIY speaker until my recent OS build. I was looking for a significant step up from the Advent Minis I was using as desktop computer speakers, with a 30 WPC MCS (made by NEC Japan) stereo amp. Well, the OS were an easy build, look nice and are a very big step up in SQ. Being an audiophile, I'm always looking for better sound and as much as I like the OS I am wondering if there is a DIY option that is about the same size but has more bass and even more transparent mids and highs.
I have spent several hours looking at threads that discuss possibilities and the EL70 is frequently mentioned, but I've seen no comments about it having better bass response than the OS and one poster that felt that it was not as good in the highs. Several other builds were mentioned such as Zaph audio minis, Paul's Speedsters and a couple of Madisound kits, but it seems that while there is considerable consensus that the OS are very good there is no consensus regarding alternatives that are a clear and significant step up in SQ in a cabinet that is very close in size to the OS.
I doubt that there will be agreement across the board, but I would love to hear from members who have compared the OS (in a computer desk configuration) with another similarly sized speaker and feel that the OS were significantly outclassed in bass extension and treble transparency.
I have moderate woodworking skills, a fairly well equipped shop and a budget of $350.
Thanks!
Jay
I'm a new member and this is my first post here. I'm a long term audiophile but have never done a DIY speaker until my recent OS build. I was looking for a significant step up from the Advent Minis I was using as desktop computer speakers, with a 30 WPC MCS (made by NEC Japan) stereo amp. Well, the OS were an easy build, look nice and are a very big step up in SQ. Being an audiophile, I'm always looking for better sound and as much as I like the OS I am wondering if there is a DIY option that is about the same size but has more bass and even more transparent mids and highs.
I have spent several hours looking at threads that discuss possibilities and the EL70 is frequently mentioned, but I've seen no comments about it having better bass response than the OS and one poster that felt that it was not as good in the highs. Several other builds were mentioned such as Zaph audio minis, Paul's Speedsters and a couple of Madisound kits, but it seems that while there is considerable consensus that the OS are very good there is no consensus regarding alternatives that are a clear and significant step up in SQ in a cabinet that is very close in size to the OS.
I doubt that there will be agreement across the board, but I would love to hear from members who have compared the OS (in a computer desk configuration) with another similarly sized speaker and feel that the OS were significantly outclassed in bass extension and treble transparency.
I have moderate woodworking skills, a fairly well equipped shop and a budget of $350.
Thanks!
Jay
Attachments
If you make your own cabinets, this kit will run you around $250 or so. Magical midrange to treble. Also has (for the size!) significant bass, down to 65Hz or so. There's a build thread here as well.
A Speaker Maker's Journey: The LM-1 Bookshelf Version
A Speaker Maker's Journey: The LM-1 Bookshelf Version
Jeff Bagby's Continuum is a popular one here. But the size is somewhat bigger than the OS. The component kit alone will be a few dollars over your budget, not to mention the cabinet and hardware cost. Other than those, it is the perfect answer to your requirement of "more bass and even more transparent mids and highs".
Meniscus Audio Group
Meniscus Audio Group
Paul Carmody's Speedsters also get a lot of positive reviews and within your budget: Speedsters. I do think the Continuums are worth the extra
$$$ if you can swing it.
$$$ if you can swing it.
I have a lot of respect for both Paul Carmody and Jeff Bagby. Note that both the Speedster and the Continuum use 3rd order crossover with some tailoring to compensate for the drivers' responses. The crossover approach gets the neutral midrange and good sound stage that we are looking for.
I agree with you that the Continuum is worth the small extra. It uses a larger woofer for bass extension which allows it to stay with a closed box design. I always prefer the transient response of a closed box alignment which is audible to me.
My vote for a huge step up is GR Research XL-S Encore. The quality of the drivers and crossovernis superb. It beats many expensive speakers in Blind tests.
X-LS Encore kit
X-LS Encore kit
Just since it's being discussed, the LM-1 uses symmetrical second order filters, and no "tailoring" is involved. While it may be a few Hz shy (if sealed) of the Continuum, it's also 5 dB more sensitive. It's also a little cheaper since the $350/pr cost includes Dayton cabinets and mid-range caps while the Continuum kit cost is $350 without either.
I personally have never heard the transient issues others report with ported speakers, and am quite happy with them ported. In this configuration they have about the same -6 dB point (more or less) as the Continuums. If using them with a sub, or close to the wall though I may stuff a sock in the ports. 🙂
Sadly, I have never heard any of J. Bagby's speakers so I have no idea how they compare musically.
Best,
Erik
I personally have never heard the transient issues others report with ported speakers, and am quite happy with them ported. In this configuration they have about the same -6 dB point (more or less) as the Continuums. If using them with a sub, or close to the wall though I may stuff a sock in the ports. 🙂
Sadly, I have never heard any of J. Bagby's speakers so I have no idea how they compare musically.
Best,
Erik
Last edited:
Erik,
you have to explain how you were able to magically transform a 87dB at 2.83V midwoofer into a 88-89dB speaker, the truth is that your speaker is more or less 83dB. And since we are talking about it, you are the only one here continuously praising and suggesting your only design. You can't even offer true measurements, only simulations. If you want consensus, you have to bring your speaker to a DIY fair, something that both J. Bagby and P. Carmody did.
I don't offer a suggestion because the real problem here is that the OP should clearly state how big the new speakers can be, because at 4.5L the OS are really small.
Ralf
you have to explain how you were able to magically transform a 87dB at 2.83V midwoofer into a 88-89dB speaker, the truth is that your speaker is more or less 83dB. And since we are talking about it, you are the only one here continuously praising and suggesting your only design. You can't even offer true measurements, only simulations. If you want consensus, you have to bring your speaker to a DIY fair, something that both J. Bagby and P. Carmody did.
I don't offer a suggestion because the real problem here is that the OP should clearly state how big the new speakers can be, because at 4.5L the OS are really small.
Ralf
Oh, wait! I may have made a mistake. 🙂 Give me a chance to check that. 🙂
Sorry, I've been in and out of consciousness for months due to illness and as a result even when I'm "awake" i'm in a thick mental fog. it's quite possible I accidentally measured wrong, so thanks for checking that.
Best,
Erik
Sorry, I've been in and out of consciousness for months due to illness and as a result even when I'm "awake" i'm in a thick mental fog. it's quite possible I accidentally measured wrong, so thanks for checking that.
Best,
Erik
Thanks for the input so far, it is very informative. With respect to speaker volume, my guess is that the maximum that would be acceptable given the stated 4.5L OS volume would be in the neighborhood of 5.5L. However, please remember that I am a complete newbe to all the technicalities of speaker design.
Based on some of the comments so far, I investigated some more of Jeff B.s designs. If it were more affordable, his ported Piccolo design, which is apparently one of his favorites, would be the one I would strongly be leaning towards.
Jay
Based on some of the comments so far, I investigated some more of Jeff B.s designs. If it were more affordable, his ported Piccolo design, which is apparently one of his favorites, would be the one I would strongly be leaning towards.
Jay
Hi Giralfino,
I think that often talk about other designs though I am not nearly as well versed in DIY kits as you seem to be.
Please remember I don't make any money on any of these kits but do offer up all the data and simulation files as well as a nuamber of hopefully instructional pages on their design for free. My most important goal is for others to learn about the entire process, as opposed to trying to hawk a kit. I don't make any money whether some one builds a thousand of them or not. I'm very happy if people just take the XSim files and learn from them, rework the crossover, or whatever.
Your criticism that the data is incomplete however is not completely without merit. While the driver measurements are part of the simulation files (which are available here at DIYaudio) I never completed the full system measurements. I'm afraid after completing the LM-1C and then the LM-1 I got pretty sick. I had very dirty versions of the system measurements (no near field). That's why if you look at the LM-1C page you'll find only the far-field system measurement.
I hope to get better and have the energy and mental clarity to do a proper system measurement. This could take months so I thought it would be more useful to newbies for me to post what I have than wait. Apparently you were offended by this approach, so I'm sorry.
Of course, none of this means I have not made a mistake, so having others review my data is a good thing, when done nicely.
Now, back to speaker sensitivity. I've gone back and checked my data. I used OmniMic to the best of my abilities which could be mistaken. Before measurements I checked the level at 60 Hz with a Fluke, and got it as close as I could to 2.83 volts, but it could have been a wee bit off.
The specs for the the 830991 call it an 87 dB woofer. My in-cabinet measurements were that it was closer to 89 dB @ 1kHz. Jeff's measurements put the Continuum around 84 dB at 1kHz. Still, it's quite possible I measured 1-2dB off due to the quality of the microphone or the steps in the preamp. As luck would have it, the LP filter has almost no effect on the 1 kHz amplitude, so whether you believe 87 or 89, that is the actual sensitivity at 1kHz. Also please note that I simulate for far-field, not quasi anechoic, so the downward tilt in my simulations is not an indicator of less sensitivity. .
I had full system measurements, but they are far-field and dirty. My very hazy memory of them is that they correspond very closely to the simulated files. That and getting sick made it less than worthwhile for me to go through the effort. I'm afraid the simulations' still don't have the port contributions properly spliced in either. Hopefully I'll get better soon and be able to contribute quasi-anechoic measurements of the complete design soon.
You do point out that not a lot of people really have appreciated my efforts to spread the knowledge of speaker design through simulation and example, and I should take a hint. Thanks!
Best,
Erik
I think that often talk about other designs though I am not nearly as well versed in DIY kits as you seem to be.
Please remember I don't make any money on any of these kits but do offer up all the data and simulation files as well as a nuamber of hopefully instructional pages on their design for free. My most important goal is for others to learn about the entire process, as opposed to trying to hawk a kit. I don't make any money whether some one builds a thousand of them or not. I'm very happy if people just take the XSim files and learn from them, rework the crossover, or whatever.
Your criticism that the data is incomplete however is not completely without merit. While the driver measurements are part of the simulation files (which are available here at DIYaudio) I never completed the full system measurements. I'm afraid after completing the LM-1C and then the LM-1 I got pretty sick. I had very dirty versions of the system measurements (no near field). That's why if you look at the LM-1C page you'll find only the far-field system measurement.
I hope to get better and have the energy and mental clarity to do a proper system measurement. This could take months so I thought it would be more useful to newbies for me to post what I have than wait. Apparently you were offended by this approach, so I'm sorry.
Of course, none of this means I have not made a mistake, so having others review my data is a good thing, when done nicely.
Now, back to speaker sensitivity. I've gone back and checked my data. I used OmniMic to the best of my abilities which could be mistaken. Before measurements I checked the level at 60 Hz with a Fluke, and got it as close as I could to 2.83 volts, but it could have been a wee bit off.
The specs for the the 830991 call it an 87 dB woofer. My in-cabinet measurements were that it was closer to 89 dB @ 1kHz. Jeff's measurements put the Continuum around 84 dB at 1kHz. Still, it's quite possible I measured 1-2dB off due to the quality of the microphone or the steps in the preamp. As luck would have it, the LP filter has almost no effect on the 1 kHz amplitude, so whether you believe 87 or 89, that is the actual sensitivity at 1kHz. Also please note that I simulate for far-field, not quasi anechoic, so the downward tilt in my simulations is not an indicator of less sensitivity. .
I had full system measurements, but they are far-field and dirty. My very hazy memory of them is that they correspond very closely to the simulated files. That and getting sick made it less than worthwhile for me to go through the effort. I'm afraid the simulations' still don't have the port contributions properly spliced in either. Hopefully I'll get better soon and be able to contribute quasi-anechoic measurements of the complete design soon.
You do point out that not a lot of people really have appreciated my efforts to spread the knowledge of speaker design through simulation and example, and I should take a hint. Thanks!
Best,
Erik
Last edited:
The XLS-encore uses components that add up to more than the purchase price because they are being sold at a loss. Peerless paper cone woofers, Vifa soft dome tweeters, heavy duty air core crossovers with film caps, binding posts, port tubes, wires and all hardware. $170 is tough to beat for a pair of speakers.
Hi,
Yes well take a hint. your design does not include BSC and
is just as wrong as the original design for the same reason.
Conversely that makes it a far better nearfield than farfield.
so it is in fact it is best used as a desktop monitor, not hifi.
rgds, sreten.
Yes well take a hint. your design does not include BSC and
is just as wrong as the original design for the same reason.
Conversely that makes it a far better nearfield than farfield.
so it is in fact it is best used as a desktop monitor, not hifi.
rgds, sreten.
Hi,
The woofers are GR-research and the same as in the XLS classic kit,
which cost $110, which IMO is the killer value speaker kit. There is
nor much difference in the crossover quality, just dubious bypass
caps to up the ante, and your just spendi extra on more expensive
tweeters. which of course will be better, but IMO the premise is flawed,
the GR research bassmid is not so great that its badly let down by the
Classic tweeters, IMO its flannel and a bit of marketing.
They would have to be poor for it to make sense, and they are not.
YMMV but the $110 Classic kit beats the Encore hands down for VFM IMO.
Not that I'm saying the Encore is a bad buy at all, its still very cheap
for all the parts and the very high quality level of the crossover.
rgds, sreten.
$110 :
An externally hosted image should be here but it was not working when we last tested it.
Hi,
No. It would be an appalling nearfield.
Sadly the original write up has been
heavily truncated, why I don't know.
Well not quite true in terms of my suspicion.
The things that make it a genuine LS3/5A
clone have been expunged to make it seem
like a "normal" speaker, whilst in fact it isn't.
Like proper details of BSC and the bass alignment.
rgds, sreten.
Though not my favourite speakers the LS3/5a's I heard
on open stands away from walls did a remarkably good
job of sounding focussed and a much bigger speaker.
Last edited:
Erik, when dealing with economical drivers (most DIY drivers in the under $100 per unit range falls under this category), you need to purpose design your crossover to compensate for their shortcomings. Both Paul Carmody and Jeff Bayby are masters in this and help many DIYers to do it right.
You did not address that and the 2nd order crossover is adequate cost saving approach that yields acceptable results, but is far from optimal. BTW, it is physically impossible to have a 5 dB sensitivity advantage over the Continuum using the drivers in LM-1.
Yet, if you cannot hear the difference, it does not really matter. Bass reflex is a good choice for you. You can trade quality for quantity.
Until you have the chance to compare your design with other established references and get some 3rd party opinion, you may want to pause promoting the LM-1 for now.
It was your recommendation that I got the classics. While very good, I had some issues with the no name tweeters being sub par. After several failed replacements, GR Research finally hand matched a pair for me and sent to me and that worked. His comment was that the Vifa tweeter is infinitely better in quality and well worth the extra price. I might have to agree - although I haven't heard them.
Keilau,
This thread is going kind of in a ridiculous manner. I was really looking forward to constructive, specific criticism, or alternative designs when I started the LM-1.
If anyone would like to participate, the LM-1 crossover files are available, I look forward to helpful, constructive and specific suggestions, which I have solicited since the very beginning. It disappoints me that this criticism comes with such personal rancor which makes it's hard for me to take any of it seriously. I hope instead that anyone participating can return to being as civil as I am used to seeing here.
I remain open to the idea that something went wrong in my sensitivity measurements, either operator error or tool error but it is not entirely impossible that it is 5 dB. For instance if the Continuum's have enough BSC. Again, I am open to the idea of an error of measurement, and apologize I won't be able to return to that for a while.
I do want to say I hope you can all see how puzzled I may be at this reception, or lack thereof. The very first criticism I received was along the lines of the LM-1 lacking parts. << sigh >>
Best,
Erik
This thread is going kind of in a ridiculous manner. I was really looking forward to constructive, specific criticism, or alternative designs when I started the LM-1.
If anyone would like to participate, the LM-1 crossover files are available, I look forward to helpful, constructive and specific suggestions, which I have solicited since the very beginning. It disappoints me that this criticism comes with such personal rancor which makes it's hard for me to take any of it seriously. I hope instead that anyone participating can return to being as civil as I am used to seeing here.
I remain open to the idea that something went wrong in my sensitivity measurements, either operator error or tool error but it is not entirely impossible that it is 5 dB. For instance if the Continuum's have enough BSC. Again, I am open to the idea of an error of measurement, and apologize I won't be able to return to that for a while.
I do want to say I hope you can all see how puzzled I may be at this reception, or lack thereof. The very first criticism I received was along the lines of the LM-1 lacking parts. << sigh >>
Best,
Erik
Last edited:
- Status
- Not open for further replies.