AudioNervosa
Self Medicating => General DIY => Topic started by: richidoo on September 19, 2014, 12:42:53 PM
-
Construction underway of the 3 way boxes. These are the woofer cabinets/stands. Outer layer is plywood, inner layer is MDF, joined with 1/16" layer of GreenGlue, staples in corners and one staple in the middle of the panel. Even without the front baffle on place, and before the GG is fully cured, the panel resonance is greatly reduced.
Next up the baffles so I can get them playing while I work on the tops.
First time using Tapatalk, does it work? :D
-
Pics
Internal bracing, acoustic damping (stuffing,) wiring, vibration isolating woofer mount still to come.
-
Now were talking. Good cabinet size there.
charles
-
Aiming for Qtc .5 before bracing and driver volume. Bracing is installed, will post pic before I close it up. Almost done!
-
Yay!!! Can't wait to hear 'em. :)
I'll have to plan something special when they're ready..
-C
-
Some progress made recently... pics:
Photo descriptions
1. Strut bracing connects adjacent walls together, triangulating the corners which are the mechanical "ground" in a box structure. I have found this to work very well, without eating up a lot of volume. These are spaced so that 6 inches is the maximum unbraced span. There are also 3 angled, side to side braces per box. All surfaces are braced. They could have been smaller with 4" spacing, but it would be hard to get the stuffing under them.
2. The braces to the front baffle are installed before the baffle is glued on, with one end coplanar to the coming baffle. While installing the stuffing I elbowed a couple of these off by accident, due to MDF being so weak. But most subs are made of all MDF and a lot more internal stress than these, so it should be strong enough.
3. Woofer box upright as it will stand in use. It can be turned upside down to look better with woofer close to the mid driver, but that risks floor bounce cancellation, so I made it invertable. The speaker wire outlet is a simple hole drilled same size as the hookup wire in the height and width centers of the rear panel.
4. With the mighty Eton 11-580 woofer resting loosely in the hole. These are the same drivers used in the Usher Be-10 and Be-20, which have the best bass I have heard. In early Be-20s the woofer to mid crossover was 640hz! necessary because of the tiny midrange driver, but still, even with 2nd order crossover at 640hz, trumpet reference tracks sounded perfect. Amazing performance in speed and non resonance for a big woofer. they give excellent spatial detail in low freq reverb. Wicked scary slam on transients with strong amps, and they can even do organ music with low Fs of 23Hz although not too loud. I like these woofers. Found a demo pair on closeout recently, never used. :thumb:
5. Closeup of the corner roundovers. The box has a 1.5" radius roundover on the front side edges to reduce diffraction from sharp corner. Actually this speaker will never play high enough for 1.5" radius to come into effect, these would need 12" radius to have an effect. So it's just for looks, to match the radius that will be on the top boxes for the mid/tweets. The other edges are 1/2" radius made with a router bit. The 1.5" radius roundovers were made by cutting off extra wood with table saw, making a 45 degree bevel edge, then further paring it down with a couple 22.5 degree cuts on either side. Then rounding off the four 22.5 edges with a 80 grit sanding disk in a drill. Feeling the roundness by hand, it wasn't too hard to make a pretty good smooth roundover. A 1.5" roundover bit is expensive and I don't own a router table, so it's not really safe to run such a big bit by hand. Moving on...
6. Closeup of the edge of the Eton hexacone material. This is very stiff and light, with mild resonance compared to metal. I don't like the cone cry and 3rd harmonic distortion that paper drivers produce as they flex under high air pressure at high spl. I want to play loud symphonies without the woofer distorting. I also like shallow crossover filters which makes using metal woofers difficult as they breakup too harshly for a shallow simple crossover. This solves both problems. Stiff, light, and relatively mild breakup. Low weight of the cone allows 91dB sensitivity.
7-8. Top and bottom of Eton.
9. The internals of the box, with pink FG stuffing on 3 walls.
The combination of plywood and MDF laminated with green glue, the corner bracing and FG stuffing makes the boxes very quiet. Knock test is very dead. More than adequate for this speaker to play no higher than 500Hz, probably more like 250-300. I think baffle step for the woofer box will be around 300, so I will try to put the passive crossover near there. But I will be doing active through JRiver filters first.
Looking at House of Kolor Urethane candy paints. It's an expensive, complicated process that requires considerable skill so I'm not yet commited to that, but looking for similar alternatives. I did find a Behr wall paint color that I like. :thumb: This means, YES, these speakers will actually be painted!
-
I split the thread for the 3-way project, called "Trairos," after the original 2way Kairos designed by Jeff Bagby.
-
This week I worked on getting the speakers working with active crossovers. SUCCESS! :yay2:
I used JRiver20 as the DSP engine. It has a module called Parametric EQ which allows creating all the necessary filters for a 3 way crossover. JRiver DSP processing is 64 bit so there is no loss of SQ. I also use the volume control in JRiver.
I use my trusty old Presonus Firepod firewire ASIO audio interface. I have used it in the past to make some awesome high resolution recordings, but I've never used the multichannel line outputs before. They sound very good considering the pro-audio grade output stage with lots of electrolytic caps for connection safety. Sol tells me it uses the same wicked excellent opamps he uses in the I60 preamp stage. My opamp phobia is cured.
I am still using the original Kairos 2Way boxes on top while the bigger Trairos tops are being built. Their footprint will match the bass boxes. and Vb is much bigger, to yield Qtc ~.5. The passive crossovers were removed and the drivers wired directly to the amps, one channel for each driver, nothing in between. The amps have direct control over the driver, and this is where the magic comes from. A cheapo 20uF Xicon electrolytic cap is used to protect the tweeters from the 15w amp.
Sol's I15 amp is on the tweeters, Sol's I60 on the mids, trusty old AudioSource Amp200 (80w) is on the bass. It is getting blown away by the other amps, but I have big bass, so I'm happy enough for now. Not much punch or precision from the budget amp.
After a week of experimenting with crossover settings I finally came up with something halfway decent tonight. I had been trying to do 3 drivers all at once and it was frustrating me. Today I thought it would be cool to just hear the 2way Kairos with active crossover, so I spent a few hours tweaking that to perfection. Then I just crossed in the bass driver at 200hz, 2nd order and adjusted the bass amp volume. Added some baffle step to the woofer. Shazamm. Sounds really good.
Then I started making a punch list of things needafixin. It is pretty long. Short sample:
- Swap AudioSource amp for more channels of I60 for the bass
- Upgrade the I15 to latest current mirror design to eliminate the thumps and the protection caps.
- Add more stuffing to the bass box to lower the resonance to minimum.
- Measure FR to see how close my ears are.
- Buy JRiver. I've already extended the trial period, so I have 9 more days of freebie. It is really good. Sound, features, ease of use.
But there's time, and it's more than listenable now. Classical, jazz, pop, rock, bluegrass all sounds good so EQ must be close. Audiophile sound quality parameters are all improved by a large degree. It is very fun to listen to music like this.
But I was almost ready to chuck it just a couple days ago. Computer hiccups, bad crossovers and EQ, soggy bass, ick. It's not a trivial project, I've been working at it for several months from the panning and researching stages. But it's starting to payoff.
The Pentium 4 with XP and 1GB of RAM works fine. If I run other programs with it like Remote Desktop, or if virus or update kicks in unexpectedly it will cause some dropouts or even long pauses. Closing the offending app cures it. But when playing music normally it is only using 4% CPU and 1/2 of the physical RAM to run the DSP. There is a slight fan noise between tracks, but I plan to move the PC into the room behind the listening room and run the DAC cable through the wall to the system.
Pics:
-
I can use Remote Desktop on my laptop to make EQ adjustments from my listening seat. It's impossible to make progress fine tuning the crossovers and EQ by walking back and forth from sweet spot to computer.
I found a Windows Phone app that does remote control of JRiver. Transport, Volume, etc. called "nMedium"
http://www.windowsphone.com/en-us/store/app/nmedium-play/b43ec7af-cc82-4dee-af94-9cf44a4dc8d3
-
Hi, Rich.
That nice progress you're making. I'll wager they sound pretty amazing now.
I have a couple questions for you specific to this project of yours, and to my (possible) future project. For several months now I have imagined having a setup not too dissimilar to yours. It seems to me like a lot of time and effort must be put in to create a three-way speaker such as this. It is a challenge for you, but one that you are overcoming.
Unfortunately for me, this is entirely too difficult as i am still a beginner. What I hope to do is pair two subwoofers with my Adelphos speakers rather than create a three-way speaker.
So...how would your Trairos sound different from my using subwoofers with my Adelphos?
If I were to dial in the subs, couldn't I achieve nearly the same sound quality?
BTW I finished my Adelphos speakers today. They have been playing now for about seven hours. I know they'll need another few days before they are really broken in, but i am quite pleased with their sound.
Is it just the mid-woofer that is breaking in, or does the tweeter break in as well?
Michael
-
Congrats! It's an advanced level project. You're not a beginner anymore! The "hardest" part is the box building, routing driver rabbets, etc. The rest of it is adjustments or mental chores, for which you can ask for help. The fun is learning, so we always strive to take on the hardest projects possible. Some things I'm considering for next year are way beyond anything i thought myself capable when I started. Still not sure how I'm gonna pull it off, but it will be fun trying. the only thing that changes over time is the percentage of failures. If you use plans which are popular with other builders, and come with support, then the chance of failure is very low.
As for break in, woofer, tweeter, and crossover parts all need break in time before they stabilize, and the speaker cables too if they are new. Don't worry about break in. You can't speed it up, you can't avoid it, so ignore it. Just enjoy listening to them knowing they will get even better, but it's already good now so that doesn't matter. One day you will suddenly notice how open and smooth they sound and how low they play, then you know it's done. Don't create anxiety for yourself by counting days and hours, because that will dilute your listening enjoyment. Just forget about it. Focus on the music. The speakers are just an instrument to connect you to the music. The music experience is always the goal. Put on some music you really love and your attention will be drawn away from the sound and back to the music.
You can set up your subs with similar crossover freq and slope as my active speakers, (12dB/oct@150-200Hz) and that is a good place to start. But you can experiment with steeper slope to match your mains, and different frequencies from 50-250. Expect to spend some time for fine tuning process. I think the Rythmik amplifier has a high pass line out to the main speakers' amp? That should make it pretty easy to tweak.
Active amplification increases damping (control) of the drivers dramatically, so the clarity and punch increase accordingly.
Yours won't sound exactly the same as mine since you have passive crossover with steeper slope on the Adelphos. Kairos' slanted baffle allows a shallower slope crossover. Your subs are active, with servo feedback, so that should sound better than my bass setup with the soggy amplifier. I'm sure you will be very happy with the result!
-
Oops, forgot to reverse polarities on my LR12 crossovers. :duh That explains why I couldn't get the mid and bass crossovers to gel.
I'm already tired of dealing with the computer for music listening. Sonos' ease of use has spoiled me rotten. PC makes changes very easy and fast from the listening seat, so it's good for designing crossovers but for general use it repels me from listening in general. ALso, JRiver can't input an analog signal without clooging around and upgrading my PC/OS. Minimum phase filters are good enough for me if they are active line level, so I have no real requirement for DSP. Everything can be done in analog and should be easier to use, cheaper to acquire and better SQ than my current PC/audio interface setup. I would need DSP to do linear phase crossovers, but no budget for that now, and it's even more computer complications.
I'll use LR12 crossovers based on Rod Elliot's P81. I did LR24 active analog crossovers before, sounded great, even with lesser PS and opamps than I would use now. It should sound better than the firepod without the electrolytic output caps and SE/BAL conversion. I can incorporate analog line level baffle step correction with the crossovers, and will build it as a preamp with input select and remote volume controls. It'll take a while, but it makes sense. I'll buy JRiver and use it in the meantime. Missing my TT and FM will spur me on to complete it quickly.
I have PCBs, output transistors, huge 1000VA power transformers, and big heat sinks on-hand now to build 2 high power version monoblocks of Elliot P101 MOSFET amplifiers for the woofers.
4 or 6 mono channels of Sol's 60 watt amplifier will require making PCBs. It would be great to make that amp available to the world, so we have PCBs on the to-do list. This project will be the impetus for progress on that front.
-
Very nice. And i notice you are using the Satori sb29 tweeter. I am using the regular SB29 and it is very good, that can only be better.
I have to check out the Eton drivers. How do they compare with ScanSpeak?
-
The SB29 is a great tweeter, and for the price it is amazing. JansZen Loudspeakers uses it rear firing for their great speaker. The dimple dome really reduces distortion a lot by fixing the center of the silk dome to a fixed post in the center, so only the donut around the post moves. Normal domes have a lot of slop at the center. Scanspeak puts a felt pad close behind the revelator air circ silk dome to physically dampen the vibrations at the dome center. I don't know how the Satori improves on it in SQ, but there are a lot of upgrade features.
As for woofers Scan makes some different lines for different budgets. I have heard the aluminum 7" Illuminator, it is wicked awesome paired with Raal 140mm ribbon and TACT linear phase crossover in Selah Audio Mejor. Hearing flugelhorn playing through it made me feel like I was playing it myself, hearing all that spitty lippy brassy detail up close to the horn. But I don't know what their large woofers sound like, they are not often used in the DIY world due to the high price. The Etons are very expensive too, but I got these on closeout when a custom builder sold off the parts because the buyer never came to pick them up. :)
I had these same Etons (11-508) in my Usher Be20s way back. I loved the bass they made in ported box, so I went for these again. Their sensitivity should work well in a passive crossover with the SB mids, allowing extra sensitivity for baffle step correction without having to attenuate the mid much. They have very low Fs, low mass, honeycomb constrained layer cone construction, and low inductance, so they can play cleanly very high. The ushers crossed them at 640Hz with 2nd order! I listened to my mono trumpet midrange reference recordings and couldn't hear the crossover to the little 4" midrange at all.
I crossing them at 150Hz now, but once I get things sorted I might take it up higher for bigger scale, more ballsy sound like the Ushers had. I'm hearing a little bass resonance, the boxes need more stuffing. Hardly any in there now and very small resonance. That's why I made such big boxes with Qtc .5, so there would be excellent natural damping from the box. We're listening to solo piano which is torture test for bass accuracy.
Fixing the polarity tonite really made everything come together. Just flipped the midrange SCs. I prefer woofer and tweeter positive and midrange neg. I can do it in JRiver DSP also (room correction module) but that's just another bite out of the CPU apple.
Earlier tonight after I switched the polarity and pulled out 1dB of baffle step on the woofer my wife said from the other room, "wow those sound really good now." Then just now sitting next to me on the music couch she said again "these really sound good," with a look of amazement on her face like congratulations you finally made good speakers! Atta boy! Now take out the trash.
Last night I was grumbling about how bad they sound. Sitting in the sweet spot there was too much muddy bass, while in the kitchen 15 feet further back there was no bass at all.
The settings are still fine tuned for the wrong mid polarity, so I'll have another aural session with Aimee Mann again tomorrow when all the haters are gone. Her voice is perfect for voicing speakers. Besides I like to look at the CD cover.
-
Speakers are all tuned up. The EQ is adjusted after correcting the midrange polarity. Nothing changed except the baffle step correction. Instead of separate baffle step corrections for mid and woofer, I put them together into one filter, and placed it before the crossover filters as it would be in an analog filter. Sounding very good.
I'd like to tighten up the bass further. Some of that is the amp, but I will add some more bass stuffing to lower the resonance to the minimum, that should tighten up the bass some more too.
I still have the cheap Xicon 20uF electro caps on the speaker wires as protection for the tweeters because the amp I'm using has a turnon thump. But the cap adds phase error to the crossover, so I need to get it out of there. The amp has a revision coming which will cure the thump. The Xicon also does the typical nasty gravelly, screechy electrolytic things to the treble so I'll be glad to remove it.
The good news of the day is that TedB on AC gave me a clue about inputting analog sources into JRiver. So I poked around a bit and found the setting to input ASIO audio directly into JRiver. So now I have vynil and radeo playing. And it is pretty easy to switch sources.
I'm continuing to design the all-analog crossover on a protoboard, but with the analog sources working well in JRiver, I may stick with JR for a while. I paid for it today, so might as well get to know it. There are so many features it will take a while to use it all. Sounds good!
-
An Update, for anybody following along...
The electro coupling caps on the sound card's outputs have broken in, so it sounds much better than it did when I first started using it. I've owned this card for 8 years but I'd never used the line out jacks. The protection electros on the tweeters also sound a little smoother. I don't notice the gravel like before.
The cabinets are underdamped due to not enough acoustic damping inside. Voices are better now than they were with passive crossover on the mid box alone, but solo piano (the acid test for clarity) still sounds a bit loose on some notes. I will add pink 2" stuffing glued to all interior surfaces, then add more until the Qtc bottoms out at the designed .5. Active amps can make up for the lost SPL. I stuffed them lightly because I intended to make passive crossovers for them sometime in the future. But now I see the big benefits of active amps so I don't care about passive. Passive can still be done at the cost of a few dB of sensitivity.
I downloaded National Semiconductor's free version of MultiSim Blue and will be using that to create PCBs to build the six I-60 monoblock amps for power. That will be a big project for the winter.
The old XP computer probably needs to go. It drops out too often from background processes, network issues (buffering songs,) etc. I'm deciding whether to go with analog crossovers and use a small linux computer for transport, or build a Win8 i5 PC and store songs locally, run linear phase crossovers and JPlay with plenty of headroom to avoid dropouts. The price difference is about 4 to 1 in favor of the analog/cubox method, but what about value?
I think the Firepod DACs are good enough for high end system when used with active crossovers and amps. Going active adds so much that it more than makes up for any refinement lost in the pro audio DAC. I think I can improve the Firepod by upgrading or removing the output caps. It already uses my fav opamps that I would use in an analog XO. But greed tells me that it would be nice to have both the active XOs AND an ultra refined DAC like my Buffalo. Decisions... I'll probably have to try the analog crossover to see which is better. It need Belleson +/-15v regulators and a good volume control to adjust phono level, it's more of a preamp/xo combo so it won't be as cheap as I have been imagining. I need to add it up and see how it compares to building a computer.
This morning I purchased "nMedium Play" JRiver remote control app for my Windows Phone after the trial had expired. It's not bad. I can browse and search the library pretty easily, and use transport/volume controls. That makes up for the lost Sonos remote capability.
Instead of playing music for my morning coffee this morning, JRiver had to do an upgrade on itself. I stepped away for a minute and when I returned the upgrade had failed because I did not give Windows permission to install non-MS approved software. And when I cancelled the failed install it went BSOD. Nice. Haven't seen one of those in decades.
Anyhoo, that's where it is today. Next steps are to get more FG into the boxes to make them sound nice and tight. If that succeeds these will be within the top 5% of speakers I've heard. Then i can start listening for smaller flaws.
I'm enjoying the added dynamics and better detail and control that I hear in loud symphonic passages that previously were more like "wall of sound." Now I can hear more inner detail that is calm and matter of fact rather than a clusterf. Despite the added detail, the music is still easy to get lost in.
-
Stuffed the tops and bottoms with polyfill today. I started with 4" thick bonded logic recycled blue jeans batting in 12x12 squares but it is too dense for stuffing a speaker. It did not effect the impedance measurements as much s polyfill did. Polyfill is less dense and more springy so it can fill the whole innards with variable density. I have it packed pretty tight now. 2 pounds of fill in a box size of 3 cu. ft. (84 liters,) target Qtc .5.
Filling the whole insides very dense is too much, exactly as I hoped. It does seem to have helped the resonance thing which was the goal. Before stuffing I hummed into the boxes and the resonance in midbass was very strong. Not sure how that happens in the small box but it did so that's probably the issue. After stuffing to the gills resonant sound inside the box was way down.
Listening to the Kairos as overstuffed was interesting. No resonance, but it did dull off some of the speed and some sensitivity, especially low freq. Mids and uppers were not affected much by the acoustic damping. The Kairos designer Jeff Bagby recommends filling the back half of the box with loosely packed polyfill batting. I think that will be proven to be good advice.
In the bass boxes the stuffing has even more effect because all the signal is low freq. I think the resonance is gone, but I didn't have as much time to listen to it as 3 way, mostly as 2way tonight. But the speed and zap of the woofer was swallowed up by the stuffing.
Tomorrow I'll try to find the happy compromise of no resonance with speed and micro details restored. Or tolerable resonance with tolerable loss of speed. :)
I am wondering how compares the full box of low density polyfill to a denser more absorbent material like 1" cotton batting lining the walls while the middle volume is empty. I saw that they do sell the 1" cotton 12x12 tiles at Homedepot.com
-
I pulled out about half of the woofer box stuffing last week. It was not enough. I pulled about all but a pound from each woofer, and highly fluffed what remained. That was the most I could get in there without messing up the natural sound. But the bass is still not well damped enough. One of my cheap amps acted up after I did the final stuffing adjustment, so I couldn't test it until today. Carl advised me to get the amps sorted out before wasting time tuning the box as the amp could be the cause of the wobbly bass sound.
Carl dropped by to pick up his DAC and had a listen. His comments are always very helpful in getting me back on track, set in the right direction, and motivated. Thanks Carl!
The amp trouble of last week, plus the difficulty of designing PCBs for Sols amps has set me off on a couple new amp projects. I am building a pair of each of these:
Modulus-86 (http://www.neurochrome.com/audio/?page_id=1092)
Sympatico (http://www.twistedpearaudio.com/amplifiers/sympatico.aspx)
They both use variations of TI LM3886 chip amp. I will compare them and buy more of the winner to make 6 channels for these speakers. They are both fully balanced amplifiers which will take advantage of my computer audio interface. Then if the interface is found to be the weakpoint I'll build a new multichannel Buffalo3 DAC. I am in love with the sound of Buffalo DACs with Russ White's Legato discreet output stage and PlacidHD BP power supply. My Presonus Firepod might be good enough with some changes, but if not I know where to go. I may just go there anyway.
The Firepod has electrolytic coupling caps on the output. I think they are there for safety reasons needed in pro-audio situations. If I can eliminate those I think it will make a big difference. Also with the new amps I will be using the whole balanced signal rather than only one half of it as I am doing now. Shorting one half of an opamp sometimes effects the other side.
With Carl's volume adjustments on the amps today the system is sounding really good to me, and I am excited about what's coming up.
-
Sounding good there professor. Having another ear in the room helps.
We should call you Sinatra now. " I did it my way" I can see a NC sleepover Rave. Gentleman start your engines.
charles
-
Thanks Charie! That would be VERY fun! I have enough room, bring it! But it might be forever before I finish these buggers.
I am thinking now that this driver wants to be in vented box. I am not getting the clarity that I wanted. I think my prejudice for large sealed box only works with drivers that are designed for that, with higher Q. I am prejudiced because previous ported DIYs I tried sucked. But I may be forced to try it again with these. I think these boxes should be adaptable to vented.
-
Update. I suspected the woofer was not playing clear. I experimented with stuffing, but it didn't help. I asked for help on diyaudio, and a local DIY speaker expert called to talk about speaker boxes, Q, acoustics etc. I met him once before but I'm glad to meet him again now that I am doing speakers more seriously than before.
My measurements indicated that my speakers are fine. The problem is room acoustics blurring the sound. So today I used my new OmniMic to better position the speakers, fine tune the baffle step compensation to match the room gain on each side individually, then applied 4 different paraEQs to each speaker to get flat, and R/L level compensations. The digital adjustments available in JRiver are very useful.
The speakers are a foot closer together now, and a foot closer to me. While the sound is still not perfect, it is much better. The room effect has been reduced enough (in the sweet spot) that I can now hear that the room is the problem. Before the room mud was everywhere so there were no clean references, so I assumed the speakers were the problem.
I have some EQ notches of up to -15dB and some boosts of 8dB. I'll probably reduce the boosts as I can hear them causing wicked blooms at those ringing freqs. Speakers can't be moved to positions that would stop those rings at the listening position.
Resulting FR looks good on paper, Flat 20-10k +/-5dB, rolling another 5dB top oct. It sounds flattish, much clearer overall than before. Imaging is better. The bass clarity is 50% better so far on first non critical listening. But the blooming fatness and room ringing can still be heard now and then.
Now that levels are equalized I'll try increasing the woofer to mid crossover frequency. The mids can play very low, but they loose dynamics. The woofers can play clean to >1k so I'm safe up to 500Hz, and I can notch the woofer breakup if I need to. My old Ushers crossed 2nd order at 640Hz, with no notch I liked the midrange of those. I have better midrange driver now, and far better center to center spacing. So there's great potential for more tuning.
The OmniMic has a cool feature called "Bass Delay" which shows the room nodes ringing. Now we can see our enemy. We could always see the enemy in the 4 walls and ceiling, but now we see the dirty deed as it happens.
With the bass improved the midrange blooming and peaking is more noticeable now too. Ahh Waiter!?! More acoustic treatments please!
I have some new tweeter DC protection caps coming, 100uF good sounding electrolytics to get rid of the cheap chinese cap harsh on the tweeters, until my new amps are built then I can dispose of the protection cap. The amp I'm using now has a turn on thump so I need the cap for now. The new tweeter amps have opamp servo to remove DC completely from the output.
All the parts to build the Modulus86s have arrived from Mouser, even a new roll of Cardas solder. The Sympaticos came with all the parts already. I keep saying I will start building them tomorrow, but it didn't happen yet! Wintertime I am 1 lazy MF.
Tomorrow will be a fun day of listening to "that new system sound."
-
Well, the above experiment was a bust. I had absolutely no desire to listen to these new tweaks. After a couple weeks I twisted the whole room 90 degrees, a position that I have used a couple times before, it's always the best setup, but it doesn't accommodate a Christmas tree and I'm not always motivated to twist it back after Christmas. So it already sounds good without even placing the speakers. I listened for a few hours last night.
I finished building the chip amps called Modulus-86. They sound great. I converted the 2way speakers into passive crossover again, so I can do some amp comparisons with only 2 channels of each amp. When i hear them all I will know what amp to put where in the active crossover scheme. The Modulus is so good, I think I will be using it on tweeters and mid, and maybe also on the bass, unless the very similar Sympatico amplifier sound just as good. It should since it uses similar parts and concept, but not exactly. I still have to make the Sympatico ready to play. But I am much less of a hurry with the Mod86 in the system.
-
Maybe a bust but a learning experience for you and a fun read for us. Priceless.
charles
-
Update:
System finally starting to sound good now.
I inserted Steve's (stereofool) AR balanced tube preamp into the mix and that brought everything to life, filled out the midrange and mid bass. It also enables listening to my SE sources like radio and phono with the balanced input amplifiers. I retubed it last year but didn't use it because it sounds "too pretty," imo, when playing into SE amps. But with balanced output it is awesome. Thanks Steve!!
The speaker boxes still have audible resonances, even after changing the speaker positions, the room acoustic have no effect on the resonance. So I'll have to try a different bracing scheme. It would be cool to use a vibration sensor to see what's really going on, where and what freq. I will try to modify the existing boxes with BW matrix style internal bracing in lieu of the failed corner bracing scheme I tried. Chopping them open will be challenging. Removing the old corner bracing sticks cleanly is not hard, with a sawzall leaving a thin layer of wood at the joint and a heat gun to melt the glue and scrape off the remainder. I have plenty of volume inside the box to burn with more bracing, since the speakers were designed for Qtc=.5. Since then I've learned that Qtc .57 is potentially even better, since the attack is cleaner at the expense of a little more ringing. So adding 1/2" MDF matrix style bracing will bring the Q up and hopefully silence the box to kingdom come. Does anybody know the spacing of B&W Matric bracing panels? How big are the cells? For my speakers either 4" or 6" would work well.
I'm using a HiFace USB>SPDIF adapter now, which is a big improvement over the Sonos ZP80 digital out and the Presonus Firepod digital out. Carl's identical DAC still sounds better than mine, so either his newer design Legato DAC output stage is better, or the USB>I2S input is better than Hiface/SPDIF, probably both.
I am working on the custom active analog crossovers now. The design is finished, I am ordering parts this week. Two separate 3 way crossovers will go into the 3 way amp boxes near each speaker, fed by a full range balanced signal from preamp. The crossovers will have overall baffle step correction adjustable 0-6dB, and attenuating pots on each output as well as true balanced input and output stages. The filter stages in the middle are single ended. I'm using the THAT Corp. 1200 balanced line receiver, and MC33079 opamps which sound great in unity gain to power the crossover filters and for balanced output.
I'm also still intrigued by using tube amplifiers, so the passive crossovers will follow when I finish the active crossovers. Then I can try Steve's matching AR power amp again too.
Spring has sprung in Carolina, bulbs are up and cherry trees are in full bloom, so I can get out and make sawdust. Gotta pull it all together to have a nice G2G soon.
-
I finally figured out the low midrange blooming problem. It took a year but a local friend finally helped me understand what was going on. Too much baffle step correction for my taste.
I replaced some junk chinacaps in my Firepod with some Sol-recommended parts for a huge improvement in SQ. (Shhhh: Nichi KL) As it continues to break in I am amazed electros can sound this good, but that's what he promised. Now I can really hear how bad the old caps were when I play an analog source into the input of the Firepod which still has the old shit caps. The inputs have many more and larger shitcaps, so the inputs still sound terrible. In days past I recorded some pretty nice sounding sessions through those inputs and the Firepod mic preamps got rave reviews (for the money) when it was introduced. Just goes to show audiophile SQ standard is higher than pro. I will upgrade (or eliminate) the input caps and it might be able to stay in the system afterall.
I'm fine tuning the levels now with much less bass bost overall, then see how it wears over a couple weeks, then maybe a G2G. That is if there's any NCs still here?
Hello?
Hellllloooooooooooooo??
AAAAAAnnnyyyyyybbooooooooodddyyyyyyyyyyyyy ooooooouuuttttt Thhhhhhheeeerrrrrreee?
-
Do I count Rich :rofl: :rofl: :rofl:
-
I'm still here :)
glad you are working it out and have found the issues :) can't wait to hear it!!
-
Hey Guys! :thumb:
You are always the first invite to my G2Gs Bill!
-
Following along Rich. Wow you are making quite a bit of changes. Mine have been stable for a wile except now that I have a new preamp I had to add some more attenuation to the midrange. I am using Janzen Superior Z caps (red ones) I am thinking of by passing with a premium cap. Do you think there will be audible difference?
-
Dis shit's way over my head but I would like to hear those speakers ya got there. (I have been tailing your posts)
Don
-
Cool, glad to know you guys are watching. I'll post more often. When/if I have a g2g you are welcome to drop over Don.
The Jantzen Red cap is pretty detailed, without being overly so. I find it very musical, complements the warmth and detail of my Satori tweeters. I don't feel the need for more resolution at all. A bypass cap would need to be pretty darn resolving to notice a difference, imo. But I'm sure it's possible, maybe with teflon?
A bypass cap does add 90degrees of phase shift to the mix, which is supposed to be harmless, if not inaudible, but it's there so I beleive it's audible, so I prefer to use just one cap that is good enough top to bottom. ymmv.
What bypass did you have in mind Jim?
-
hmm, now you have me thinking better to leave well enough alone. jupiter? I don't see how it would add phase shift since you are bypassing a cap that is already in the circuit and just changing the value.
-
I've heard Jupiters are nice. I think Mike said something like "out of this world..." :D You're probably right about the phase.
-
I reduced the 2way Kairos' passive crossover's baffle step correction to 3dB. Mids move forward in the mix which I like, but there is less bass extension. I think 4dB of BSC might be about right for a stand alone sealed Kairos in a small room with bass reinforcing boundaries.
But since this is supposed to be a 3way, the slightly thin bass will be supplemented with a separate bass driver so it's OK. I made this BSC change to try to reduce the chestiness sound, but it persists, though lessened. The overall sound of the speaker is better so it's a net positive for me. I am tweaking the tweeter resistor now, with interesting results.
I bought some Mills wirewound 5R resistors for the tweeter but it was too hot. I tried 10W 6.2R metal oxide, a little too soft, and it lost a lot of life. Interesting. So I compared a xicon 5W 5R MOX with the Mills 12W 5R wirewound, OMG what a huge difference. Same resistance value, but a gigantic difference in SQ. The wire resistor is much more alive and detailed. Not surprising since with wirewound the signal goes through copper/nichrome wire, while with MOX the signal goes through a film of metal oxide (rust) painted on a ceramic insulator. So I ordered some Mills 5.6R resistors, I'm hoping that's just right, reducing the treble energy, but not killing off the life and detail.
An interesting question is whether a more detailed sounding wirewound resistor needs to be a lower value than duller sounding MOX resistor in order to "musically" balance the treble to the mids?
I put on "Blue Trane" this morning, and to tame Lee Morgan's bright snappy trumpet I resorted to the 6.2ohm MOX. The life is much less, but the brash is under control. I might end up going up to the 6.2R Mills. What's another $10 of shipping for 1/2 oz merchandise? Sonicraft is the only seller I could find with a large selection of Mills in stock.
So I'm working on the Trairos bass cabinets again because with the Kairos' baffle step correction (bass boost) reduced I am lacking midbass fullness and low bass. So the bass drivers need to come back into the system instead of running Kairos 2way alone as I have been to avoid the chestiness of the bass boxes. So I must now find a way to reduce the chestiness. I have tried absolutely everything over the last year, but there is only one thing left to do... Port it.
I chose these bass drivers, Eton 11-581, because of how great they sounded in previous ported commercial speakers I owned. No hint of chestiness at all, crystal clear from 30Hz up to 1200Hz in that speaker.
One way to determine whether a driver should be ported or sealed is by looking at the "EBP" spec, Equivalent Bandwidth Product.
From Eminence website:
"EBP
This measurement is calculated from driver specs by dividing Fs by Qes. The EBP figure is used in many enclosure design formulas to determine if a speaker is more suitable for a closed or vented design. An EBP close to 100 usually indicates a speaker that is best suited for a vented enclosure. On the contrary, an EBP closer to 50 usually indicates a speaker best suited for a closed box design. This is merely a starting point. Many well-designed systems have violated this rule of thumb! Qts should also be considered."
Both the bass driver's and mid driver's EBPs (and their low Qts') suggest ported boxes, even though they can technically still be used in sealed box. Since I know the Etons sound clean in ported box, I will try porting them first to confirm my suspicion. If it clears up the chesties then I will port the Satori also.
The idea is that the reflex port will damp the movement of the cone and reduce cone excursion at lower freqs, which reduces distortion. Hopefully it will be the KIND of distortion I want to reduce. There really aren't any other options, I've tried everything else. If this doesn't work then it's back to commercial speakers... haha
I am a little intimidated about doing more ported boxes after my utter fail with the Econowave speakers. There I used box sim software and used the "optimum" calculation, which made the bass too low and too punchy, like a disco. After a dozen failed attempts to make that box sound good I have the fear that designing a reflex box is a dark art that must be difficult. In that case I was using compromise drivers in way too large a box because the sim said so. I didn't yet understand how a port really works and how to tune it which I think I understand better now. In the case of these Eton and Satori, there are published hifi oriented reflex box designs that should sound good, rather than relying on the sim software's best guess which uses the old reflex port equations. The equations describe how to make the reflex port as efficient as possible and play as low as possible, nothing to do with SQ or musicality. So I'll try the suggested audiophile ported box designs for these drivers and see how they sound. Any port design needs to be tuned a little anyway, which I will attempt to with more understanding this time. The things that can be tuned are the box volume and the port length. The driver specs remain constant as does the port diameter.
If this works I will be happy and relieved. This chestiness problem has been a major sore for a long time and is prompting me to move the stereo to a different room, build different speakers, all kinds of compromises that I can sweep away if it can be fixed. I think these drivers are among the worlds best. Hopefully I am right and they are just lacking optimal acoustic loading.
That's it for now.
Rich
-
I ported the bass bins today. It was hard to cut that first hole in the pretty box. It is tuned low to 20Hz right now, with -9dB at 20Hz, which is supposed to net flat response in room. It sounds nice and not too punchy, but it can punch if needed. I think I will want it a little dryer, so I will try shortening the port down to as low as 16", which I think is 25Hz? I also want to try reducing the box volume from 3 cu ft down to 2 cu ft.
Playing sine waves at about 10W from 20-29hz shook the whole house, rattling everything in the room. 8)
I'm also working on an analog active crossover to avoid the cost in money and time needed to develop a passive speaker level crossover between mid and bass. Low pass low frequency filters for speaker level require a very large inductor, expensive and too big for air core. I am very happy with the passive crossover between mid and tweets, so I would keep that. Using a line level crossover between the mid and bass would use cheaper parts, easier to design, and better performance by eliminating the big coil.
There are three options: passive RC (resistors and caps) line level crossover (PLLXO,) Passive LC (coils and caps) line level XO (like Marchand XM46,) or active Sallen Key filter LR2 XO. I'll probably tinker with them all eventually, but for the G2G I need to pull it together quickly, so I'm thinking of trying the Marchand balanced XM46 as ~250Hz 12dB hipass for the mains, then use an active 250Hz -12dB sallen key filter to the bass drivers. If I do both crossovers with active, then I can make a baffle step correction, but the mains passive already has 3dB BSC, so the active BSC is probably not necessary. I'd like to keep opamps out of the signal path for the mains, but they are harmless enough in the bass.
The other option is to make a full 3 way active crossover with 6 amps and all active 12dB crossovers. Eventually I'll do that since I have all the amps already, but not sure I can get that going and fine tuned before g2g. Maybe?
Edit: The current bass alignment needs all 85 liters to make the 24" port work, so the port hangs out of the box, since the port itself takes away from the box volume. But if I end up with shorter port with higher tuning, especially with a smaller box, then I can fit it all or mostly into the box so it will look better.