Well if you know of a wood product that is...
I usually try to only cut the stuff on a windy day outside. This tends to blow the dust away from me. But with this project I got a burst of energy after a bout with a cold/flu and did it anyway on a still day. The dust was probably hanging around the area the entire time and I didn't even realize it. Lesson learned!
- cheap
- readily available
- cuts cleanly
- does not warp
- is available in large, thin (e.g. 1/4" thick) sheets
- does not contain glues or resin binders
I usually try to only cut the stuff on a windy day outside. This tends to blow the dust away from me. But with this project I got a burst of energy after a bout with a cold/flu and did it anyway on a still day. The dust was probably hanging around the area the entire time and I didn't even realize it. Lesson learned!
Well I found a schematic, but I had just labeled it M-12 crossover.Not really sure this is correct. John should remember.
EDIT: I found the BOM from 10 years ago, and this would be it. This matches the parts used with the 27TDFC.
Fantastic! Thanks Pano. Going to head of to Solen this week and pick up what I'm missing. Going to be driving these with an Audio Research d100b. I think it should be a pretty good match. Also going to hook up a Ripole sub made with 2 12in Sls drivers crossed over actively which should fill out the bottom octave.
Hello all.....
As we are coming up on the 10th anniversary of the Manzinita tread, and at the encouragement of Pano, I need to revisit the design and publish the latest 2 DIY versions.
Both will use the Vifa T9 up top and either the original Peerless 12" or the PE GRS 15" woofer down low. The Peerless is available world wide, but usually at 3 times the price of the GRS, at least here in the states. Unfortunately, outside of the states, the GRS availability is quite limited.
Each woofer has it's own performance advantage, but the GRS - Vifa TC9FD18-08 combination best reflect the original concept of simplicity and bang for the buck.
On the low end performance side the Peerless offers a bit more displacement and raw efficiency, while the GRS has a larger radiating area and much higher QT. The on baffle efficiency of the two is very close. But the GRS will go lower due to the higher QT.
Each version will have its own unique crossover. That is the revisit part, I have to cobble the two versions up and optimize the crossovers using the fewest parts possible.
Unfortunately, this will take some time... shooting for a week or two.
John
As we are coming up on the 10th anniversary of the Manzinita tread, and at the encouragement of Pano, I need to revisit the design and publish the latest 2 DIY versions.
Both will use the Vifa T9 up top and either the original Peerless 12" or the PE GRS 15" woofer down low. The Peerless is available world wide, but usually at 3 times the price of the GRS, at least here in the states. Unfortunately, outside of the states, the GRS availability is quite limited.
Each woofer has it's own performance advantage, but the GRS - Vifa TC9FD18-08 combination best reflect the original concept of simplicity and bang for the buck.
On the low end performance side the Peerless offers a bit more displacement and raw efficiency, while the GRS has a larger radiating area and much higher QT. The on baffle efficiency of the two is very close. But the GRS will go lower due to the higher QT.
Each version will have its own unique crossover. That is the revisit part, I have to cobble the two versions up and optimize the crossovers using the fewest parts possible.
Unfortunately, this will take some time... shooting for a week or two.
John
Since my travel plans abruptly changed this morning I was able to set up and perform some measurements on the active Charlita.
I took measurements on axis and at 10,20,30,40, and 50 degrees off axis using ARTA and with the same voltage input to each driver. The SPL levels are arbitrary as no level calibration was performed but the relative SPLs of tweeter and woofer are correct. At each position I measured the tweeter, woofer, and both tweeter and woofer together. This allows me to extract the differential pathlength at each position, from which I can extract the acoustic delay. The responses and the delay (the z-offset in some parlances) is then a model of the loudspeaker which I use in my Active Crossover Designer program.
The only other measurement I SHOULD have taken but just realized I did not was a woofer nearfield. My measurements were performed at about 1.2m and I only have 120Hz resolution. The nearfield would help me better guesstimate the ultimate woofer rolloff, but it's not really critical since I don't plan for the speaker to operate below 100-150Hz.
As a teaser I have attached a screenshot that overlays the on-axis frequency responses of the woofer (in blue) and tweeter (in green). Smoothing is 1/12th octave. My first attempt might be a crossover point around 600-700Hz, which is around the frequency of the woofer's dipole peak.
I took measurements on axis and at 10,20,30,40, and 50 degrees off axis using ARTA and with the same voltage input to each driver. The SPL levels are arbitrary as no level calibration was performed but the relative SPLs of tweeter and woofer are correct. At each position I measured the tweeter, woofer, and both tweeter and woofer together. This allows me to extract the differential pathlength at each position, from which I can extract the acoustic delay. The responses and the delay (the z-offset in some parlances) is then a model of the loudspeaker which I use in my Active Crossover Designer program.
The only other measurement I SHOULD have taken but just realized I did not was a woofer nearfield. My measurements were performed at about 1.2m and I only have 120Hz resolution. The nearfield would help me better guesstimate the ultimate woofer rolloff, but it's not really critical since I don't plan for the speaker to operate below 100-150Hz.
As a teaser I have attached a screenshot that overlays the on-axis frequency responses of the woofer (in blue) and tweeter (in green). Smoothing is 1/12th octave. My first attempt might be a crossover point around 600-700Hz, which is around the frequency of the woofer's dipole peak.
Attachments
Last edited:
Since I had the data in hand, and some time, I came up with a first stab at a crossover. See attached response plot for drivers and system. The effective crossover point is at about 675Hz.
The crossover consists of:
WOOFER:
+10dB input gain
1st order low pass at 100Hz
2nd order LP: 730Hz, Q=0.55
EQ: -2.5dB@580Hz, Q=1.8
TWEETER:
input gain -0.5dB
second order HP: 700Hz, Q=0.6
EQ1: +3.5dB@800Hz, Q=1.8
EQ2: -1dB@1200Hz, Q=2
This can be implemented in software DSP using my LADSPA ACDf plugin.
I also collected off axis data, but that will take more time to process. This is a good first try using only the on-axis data set.
The crossover consists of:
WOOFER:
+10dB input gain
1st order low pass at 100Hz
2nd order LP: 730Hz, Q=0.55
EQ: -2.5dB@580Hz, Q=1.8
TWEETER:
input gain -0.5dB
second order HP: 700Hz, Q=0.6
EQ1: +3.5dB@800Hz, Q=1.8
EQ2: -1dB@1200Hz, Q=2
This can be implemented in software DSP using my LADSPA ACDf plugin.
I also collected off axis data, but that will take more time to process. This is a good first try using only the on-axis data set.
Attachments
Hey Charlie, thanks for the pix, measurements and info on your baffle. I'm really surprised you've gotten as much low end as you have with that "next to nothing" baffle. It's mighty small.
Looks like you low pass is doing the trick, you picked a good points and slopes for it.
The first order LP is doing the same thing as the big inductor in the passive crossover. It works in this particular system because the woofer has a higher voltage sensitivity than the tweeter, so there is some of that to sacrifice. Also, the dipole peak is at +6dB WRT the nominal level. All in all I think the circuit tops out at 10dB of boost to the woofer around 20Hz, but at 100Hz it's not yet reached 5dB. The woofer's resonance rolloff is kicking in around 100-150Hz and below that point it's a losing prospect to try and correct the drooping response so that more or less determines the system's low end rolloff behavior. This is why low Q woofer in an open baffle are a challenge - the start pulling the response down faster than the 6dB dipole rolloff sooner than a higher Q driver would for the same Fs.
I like using very small baffles in OB projects, or even using the drivers suspended in mid air (the no baffle open baffle approach). According to modeling and measurements that I have made on various systems, the more baffle that the acoustic radiation sees the more irregular the off-axis response pattern.
The woofer frame in the Charlita is almost as wide as the baffle and the frequencies are lower, so the baffle size is essentially insignificant (as seen by the acoustic radiation produced by the woofer). For the tweeter, I located it very near the edge and off-center. This helps to reduce the dipole peaks and dips significantly, however, with that placement the dipole peak will be low. For a tweeter you don't need any "low end" so this placement works well and keeps the response pattern even. You don't see much of the usual "dipole pattern" in the tweeter response because of the near-edge position, so it becomes very easy to with with in the crossover.
I threw in a couple of EQ bands that help to smooth the response. I need to check these against the off axis data to make sure that the peak is not shifting with angle. I will do that later, in a second pass at crossover development in ACD.
Anyway, that's how I turned this...
...into this:
Last edited:
An "old" design as you say, but it does get many things right!
I had general listening use in mind - I will add a dipole sub for that purpose.
It probably could work well as a nearfield desktop monitor, too, since the bass will get some boost from the proximity. There might be a bit of a hole between 50Hz and 100Hz depending on the listening distance.
Your thoughts?
Near field listening
I have used the Manzanita for near field use at the computer and with the original inductor induced correction it was very enjoyable. I was surprised how well integrated the Vifa T9 was... but then that is what it is supposed to do well given its intended use in the TV and similar markets. Still nothing out there for anywhere near the money and quality that can beat it.
I have used the Manzanita for near field use at the computer and with the original inductor induced correction it was very enjoyable. I was surprised how well integrated the Vifa T9 was... but then that is what it is supposed to do well given its intended use in the TV and similar markets. Still nothing out there for anywhere near the money and quality that can beat it.
Really curious to see what John cooks up for new updates to the Manzanitas. There is a small Manzi and an Ultra Manzi in the house and they both have a lot of time on them.
The original Manzi's have been in my basement workshop for a couple of years now. Keep saying they are going to get updated to the new tweeter and crossover but never have gotten to it.
They've been hooked up to a Squeezebox and a 70's Harmon Kardon receiver and hadn't been getting a lot of play lately. So the paradesia dac that used to be in the system was put back into the mix and that change made the magic return. Spent a few hours with them yesterday and had a great time. It turned into a celebration of Walter Becker and Steely Dan.
The original Manzanitas are laid back and the tubed paradesia is also laid back so you'd expect it not to work. Yet the dac fixes the somewhat coarse treble from the Squeezebox and the old receiver. In spite of the speaker being laid back in the midrange you could really hear the roughness of the treble. Now it's a laid back, sweet and a very musical little combo.
I'll do at least one of the Manzi updates, maybe both, depending on the complexity. Really looking forward to it.
Thanks John!
The original Manzi's have been in my basement workshop for a couple of years now. Keep saying they are going to get updated to the new tweeter and crossover but never have gotten to it.
They've been hooked up to a Squeezebox and a 70's Harmon Kardon receiver and hadn't been getting a lot of play lately. So the paradesia dac that used to be in the system was put back into the mix and that change made the magic return. Spent a few hours with them yesterday and had a great time. It turned into a celebration of Walter Becker and Steely Dan.
The original Manzanitas are laid back and the tubed paradesia is also laid back so you'd expect it not to work. Yet the dac fixes the somewhat coarse treble from the Squeezebox and the old receiver. In spite of the speaker being laid back in the midrange you could really hear the roughness of the treble. Now it's a laid back, sweet and a very musical little combo.
I'll do at least one of the Manzi updates, maybe both, depending on the complexity. Really looking forward to it.
Thanks John!
Last edited:
Hi Jørgen,
What a question - if I were not so polite, I'd say that it is an impertinent one, but nevertheless I'll answer as truthfully as I can. Part of the reason why I answer is also that you poked at a question I've asked myself: Why don't I use the Orions more than I do?
Disclaimer(s)
But first the usual disclaimer: So much of it is subjective - and personal preferences do play a major role, thus it may not be God's honest truth as such - but my truth! Another important aspect in this context is my hearing. Over the years I've found that I can't stand very loud music. It seems that the changes (deterioration?) of one's hearing goes in one of two directions: Either you need much higher SPLs - or lower to enjoy the music. I have a friend who wants thunderous levels - and I usually withdraw to the street when he's really into it. (Interestingly: Linkwitz notes somewhere that he also prefers high SPLs - whereas his friend Nelson Pass prefers much softer levels.)
And finally: The Orion is a vastly more ambitious and complex loudspeaker than the Manzinitas (I hope John isn't offended, because that's really not my intention!) - so a direct comparison is very unfair indeed!
The short answer.
My main points of reference are outlined below, so please read the whole post, but the brief answer to your question: The Manzinitas' ability to render the soundstage is very similar to the Orions. I find the Manzinitas a bit more "relaxed" and "quick" than the Orions. Their frequency response is much more uneven, though. If you look at the uncalibrated far field response in my room, you'll see that they are a bit low between 200 and 400Hz and between 600 and 2000Hz. That can be heard as a somewhat leaner sound - and a slightly pronounced treble (the peak around 7-8kHz is (probably) felt more because of the two dips). The lift around 90Hz is presumably to do with the room, but the over-all bass response is characterised by the bass notes being played if the material has them - otherwise they are quiet. A good example is Niels Henning Ørsted Petersen's "The way We Were" where his characteristic "attack" is rendered were precisely together with the deep notes. On the other hand: The Orions' bass is vastly more powerful and more well defined, of course.
The long answer.
The Orions are really and truly magnificent speakers. There should be no doubt about that. When I first listened to them I was almost awestruck by the amount of detail and the sheer power they could produce. Also their neutrality if you compare their rendition with an actual symphony orchestra is outstanding. The sound stage in a concert hall is never Hi-Fi-precise. Of course, it is possible to more or less pin-point the position of say an oboe playing solo in the orchestra, but it is not - in the real world - as needle sharp as some hi-fi set-ups reproduce it. In the same vein: The position of e.g. a French Horn is even less distinct in a concert hall than in some Hi-Fi-set-ups. And I believe that Linkwitz' measuring rod is the live performance - that is what he has tried to achieve - and he certainly has. The Orions also have a completely unique quality: They can play softly without very much loss of soundstage and timbre. The only thing that happens is that the music source seems farther away. So far so good.
However, they didn't pass the long term test: I've worked with and listened to music and Hi-Hi for many, many years, and the test of time is the most reliable one for me: If I lose interest in listening to music with a particular set-up there's something wrong. Probably not with the equipment as such but with the equipment-human-combination (just as the stereo image is created by the ear-brain-combination). And having had the Orions as my main set up for I guess a couple of years or so the equipment-human-combination wasn't so sweet anymore. It is difficult to say precisely what happened, but I felt they were a bit "sluggish", a bit "heavy" unless I really cranked it up. So, what do to? As always, Linkwitz had some sort of an answer. He states somewhere on his website that the sonic signature of his Pluto speakers is about the same as the Orion bar the SPLs and the bass register, so I got myself a Pluto - and they've stayed there since. In my ears they are perfect in the sense that the music is almost ethereal - even more so than the Orions' - there is absolutely no focus on the speakers. The music is in the room and I completely forget about the speakers and the equipment. Of course, they can't give an anywhere near realistic rendition of say Mahler's 8th symphony. The Orions can't either - but they are much closer to the mark! But if I go a bit softly on the volume control they can play big-band music quite well, a genre Orion excels with. (Later I added some small sub-woofers to the Plutos. Here are some observations: http://hjkm.dk/Pluto/Pluto_w_sub.pdf ).
But as the saying goes: "In tranquillo mors, in fluctu vita" - death is in the still water, life in the waves - so being so pleased with the Plutos I decided to have a go at the LXmini - since Linkwitz claims that they are better than the Plutos. They are! - Not much, but noticeably so. - The most remarkable difference between the LXmini and the Pluto is that the former paints the room behind the musicians as completely rectangular - you can "hear" the corners of the room, if, that is, the material is up to the task. One very good example is Naxo's edition of Paganini's "Sonata for Grand Viola and Guitar" from the album "Music for Violin and Guitar II" (8.550759). I'm in the process of migrating to the LXmini in place of Pluto. But I feel that the miniDSP adds some sluggishness to them that the Plutos don't have, thus I plan to try out Nelson Pass' analogue filter. (htttp://linkwitzlab.com/LXmini/ASP.htm ). (It also seems that Linkwitz himself has abandoned digital filtering in favour of analogue processing for his LX521.)
But why not replace LXmini with Manzinita? I am considering it - and sonically I could easily live with the exchange. The advantage would be much less cable spaghetti, the disadvantage a larger footprint. Potentially it is also a problem that I can give the speakers much less space behind them than they get in the room in which they're playing now. This might be remedied by "taming" the rear radiation in the way it's done with Quad ESL57. Their radiation pattern is more cardioid than figure-of-eight which means that they can be placed closer to the back wall. Moreover, I'd have to shine up the Manzinitas. So far they are just a quick mdf mock-up - but they have played to my great satisfaction in my soundlab for years, - and impressed visitors to boot.
Attachments
Klarskov..
Enjoyed reading your post! And I am certainly NOT offended. Thank you for your positive comments.
To be mentioned in the company of Mr. Linkwitz and Pass is an honor. Both, are very nice and generous men.
You make many good points and observations. I have spent some time listening to the Orion and Pluto. Both are great designs and fun to listen to.
And, as you note, both are very different systems that do different things.
One advantage the simple Manzanita has is... how to say this... it does not provide the last bit of resolution, or anything else for that mater. Lousy recordings don't sound as bad as they are and the few great recordings out there do not sound as good as they are. But most of the middle tends to come through with minimum alteration from the crossover. A lot of good stuff seems preserved.
I have created some very complex crossovers for the Manzanita family of speakers. Some had over 30 components. Yes... it is obvious, I am a bit nuts! But the funny thing I noticed generally speaking, was as I cleaned up the spectrum, and obtained improved frequency and power response, I did not like what I heard as much with most music. After a certain level of performance I would be trading one sonic advantage for a different sonic disadvantage. Measurements looked great though.
I think the reality is you bump up against the limitations of the basic design and drivers used. Not saying improvement is not possible, it certainly is. Knowing when to stop and leave the design alone, that's what is difficult.
Working with Pano on a 10th anniversary Manzanita that incorporates the little revisions that move the design in a positive direction with out it becoming too complex and or breaking the bank. It will be interesting to see comments from those who choose to build it. There will be a couple of versions due to world wide driver availability.
Best
John
Enjoyed reading your post! And I am certainly NOT offended. Thank you for your positive comments.
To be mentioned in the company of Mr. Linkwitz and Pass is an honor. Both, are very nice and generous men.
You make many good points and observations. I have spent some time listening to the Orion and Pluto. Both are great designs and fun to listen to.
And, as you note, both are very different systems that do different things.
One advantage the simple Manzanita has is... how to say this... it does not provide the last bit of resolution, or anything else for that mater. Lousy recordings don't sound as bad as they are and the few great recordings out there do not sound as good as they are. But most of the middle tends to come through with minimum alteration from the crossover. A lot of good stuff seems preserved.
I have created some very complex crossovers for the Manzanita family of speakers. Some had over 30 components. Yes... it is obvious, I am a bit nuts! But the funny thing I noticed generally speaking, was as I cleaned up the spectrum, and obtained improved frequency and power response, I did not like what I heard as much with most music. After a certain level of performance I would be trading one sonic advantage for a different sonic disadvantage. Measurements looked great though.
I think the reality is you bump up against the limitations of the basic design and drivers used. Not saying improvement is not possible, it certainly is. Knowing when to stop and leave the design alone, that's what is difficult.
Working with Pano on a 10th anniversary Manzanita that incorporates the little revisions that move the design in a positive direction with out it becoming too complex and or breaking the bank. It will be interesting to see comments from those who choose to build it. There will be a couple of versions due to world wide driver availability.
Best
John
So this happened on the weekend.... Had to go to Solen to pick up a few things... Ended up picking up the Vifa. Constructed of solid 1inch Birch... Should be up and running this weekend... Can't wait.
An externally hosted image should be here but it was not working when we last tested it.
The Manzanita Project - Ten Years
It's been ten years since this project and this thread were started – and it's been a good ten years. The basic goals and ideas behind the Manzanita open baffle speaker have remained constant; an open baffle speaker that is easy to build, easy to use and brings the music into your room. There is nothing sexy about the Manzanita, no whiz-bang technology or gimmicks to fire the imagination, no overly expensive parts to stroke your ego and empty your wallet. What there is – is a simple build with amazingly good sound. A well-engineered project that sounds much, much better than its cost or simplicity would lead you to believe. Many of these have been built over the years and all over the world with great satisfaction. Some of those builds are documented in this thread. Over the past 10 years the Manzanita has stayed true to its goals, and thru the judicious choice of drivers, the cost is lower than it was 10 years ago!
There are currently three versions of the original Manzanita that you can build, one with the original Peerless 12” woofer and one using the Parts Express GFS 15" woofer. Along the way, a floor standing big brother "Ultra" was added to the thread. We will look at each of these. But first let's look back at the original concept.
1: The original.
The original design is described in the first three posts of this thread. We have an "off the shelf" 12” woofer and dome tweeter on a simple baffle with minimum back wings. The baffle is designed to be placed on a speaker stand, both to get the tweeter near ear height and to avoid some floor reflection problems. It’s unusual to see a stand mounted open baffle speaker, but sonically it works very well.
The original design uses the Peerless SLS 830669 - 12" Woofer in combination with the Seas 27TDFC dome tweeter. Both are high quality raw drivers at reasonable prices, offering a lot of value for money. The crossover is simple, with only 5 elements. The most expensive part of the crossover is the large inductor required by the open baffle design. The large inductor remains, by far, the most expensive and essential part of the crossover.
Why is the original design no longer current? Two reasons. The cost of the original drivers increased drastically, seriously cutting into the affordability goal of the project, and it was felt that the Seas tweeter just didn’t have enough energy in the midrange to close the gap with the woofer. The search was on for a more capable and affordable tweeter-mid. After many trials with very good, but sometimes very expensive drivers, John settled on the 3.5” Vifa (Peerless) TC9FD18-08 full Range for the top end.
Which leads us to version two in the next post.
It's been ten years since this project and this thread were started – and it's been a good ten years. The basic goals and ideas behind the Manzanita open baffle speaker have remained constant; an open baffle speaker that is easy to build, easy to use and brings the music into your room. There is nothing sexy about the Manzanita, no whiz-bang technology or gimmicks to fire the imagination, no overly expensive parts to stroke your ego and empty your wallet. What there is – is a simple build with amazingly good sound. A well-engineered project that sounds much, much better than its cost or simplicity would lead you to believe. Many of these have been built over the years and all over the world with great satisfaction. Some of those builds are documented in this thread. Over the past 10 years the Manzanita has stayed true to its goals, and thru the judicious choice of drivers, the cost is lower than it was 10 years ago!
There are currently three versions of the original Manzanita that you can build, one with the original Peerless 12” woofer and one using the Parts Express GFS 15" woofer. Along the way, a floor standing big brother "Ultra" was added to the thread. We will look at each of these. But first let's look back at the original concept.
1: The original.
The original design is described in the first three posts of this thread. We have an "off the shelf" 12” woofer and dome tweeter on a simple baffle with minimum back wings. The baffle is designed to be placed on a speaker stand, both to get the tweeter near ear height and to avoid some floor reflection problems. It’s unusual to see a stand mounted open baffle speaker, but sonically it works very well.
The original design uses the Peerless SLS 830669 - 12" Woofer in combination with the Seas 27TDFC dome tweeter. Both are high quality raw drivers at reasonable prices, offering a lot of value for money. The crossover is simple, with only 5 elements. The most expensive part of the crossover is the large inductor required by the open baffle design. The large inductor remains, by far, the most expensive and essential part of the crossover.
Why is the original design no longer current? Two reasons. The cost of the original drivers increased drastically, seriously cutting into the affordability goal of the project, and it was felt that the Seas tweeter just didn’t have enough energy in the midrange to close the gap with the woofer. The search was on for a more capable and affordable tweeter-mid. After many trials with very good, but sometimes very expensive drivers, John settled on the 3.5” Vifa (Peerless) TC9FD18-08 full Range for the top end.
Which leads us to version two in the next post.
