Yes I am, sort of. I glue 2 pieces together and stack 2 of those and clamp. I've done the nut spinning by hand a couple of layers (2 or 3) and that took about 20 minutes for 4 nuts up or down. With the drill it's a 2 minutes job. And it gives me the clamping force where I want it. It is manageable, compared to the routing and rest of the stuff I had to go trough
.
I've been wanting to build some IDS25's also, and I have the reprint of the AudioXpress article from July 2006 that Roger sent me. I have a few comments that may be helpful to some.
These Peerless/Vifa drivers are great (best I know of, all things considered), but difficult to mount since the flange is thin and the ceramic magnets and basket shape is such that rear side venting of acoustic energy is likely to be significantly impeded if the drivers are front mounted even on 1/2 inch baffle material. I would seriously consider having 1/4 inch thick aluminum fabricated for use as the front panel. That may cost too much - I haven't gotten a quote yet (my shop is either the front yard or the bedroom floor so there's just so much that is worth doing myself).
What makes this speaker system rise above most others is the way it works with room acoustics. As tempting as it is to modify the design somehow; adding a ribbon tweeter or a woofer, you would be throwing away it's main feature to a significant extent. An added woofer or tweeter right away has the integration problem, where its acoustic power drops off at closer to 6dB per doubling of distance, rather than 3dB. If the tweeter only operates above 7kHZ, it could put more air into the sound at a certain minimum listening distance and be relatively inconsequential further away, but breaking up the continuity of the array from floor to ceiling comes with a significant price. I've decided not to mess with that.
One of the main reasons woofers are often boomy sounding when you turn up the bass, is because of how a virtual point source interacts with typical listening room acoustics. Two line arrays going floor to ceiling stimulate the room in so many different places, that comb filter effects (cancellations) are largely filled in, for about any location in the room I believe, since the cancellation frequencies will be different for each driver, due to the differing distances to room boundaries and the listener. Not a lot of people understand this. It's a little difficult to put into words that don't get misinterpreted.
Like everybody else, I wanted to second guess Roger Russel; add a tweeter or woofer, maybe use different drivers, but I'm convinced that he's done quite a bit of research and experimenting, and has a very nice system as it is.
Better rear airflow for the drivers, and the rounded cabinet shape that one of the above contributors is doing (or has done) would be even better I would think, but I can't think of anything else I'd want to change.
The black paper cone version of this driver (Peerless TC9FD) sounds real good too. I've built some speakers for my computer with those. My Stereo Holographic soundbar project uses the TG9FD10-08 Peerless drivers that appear to be virtually exactly the same one's Roger Russel is using. They also sound great to my ear. The glass fiber cones are probably better than the black paper, but the difference may be small and I'm still struggling with idea of have 25 light grey cones showing through the grill assembly... I definitely prefer back cones, but those glass fiber cones are VERY nice sounding. The (Polk style) holographic imaging works very well.
I noticed that the 8 ohm version of the TG9 driver has a slightly better top end of the frequency response than the 4 ohm version, according to the published graphs, but the difference is pretty small and can be pumped up with active EQ.
Roger didn't share the schematic for the active equalizer, but does show it's frequency response in the Audio Express article that I have here. 200HZ - 500HZ is the lowest point on the graph (dB scale very difficult to read) but he did say he had about a 15dB peak in the response at 30HZ, apparently relative to 250HZ, with a fast rolloff below 30HZ (very important). There is a gradual rise in the EQ that goes from about 500HZ to 20kHZ. If he's up 15dB at 30HZ, then it looks like he's up about 7dB at 20kHZ. The slope for the curve from 500 to 20kHZ looks like it's about 3dB/octave.
By the way, Gary Pimm pointed out to us that the Behringer analog circuits in his DCX2496 unit are very poorly designed. I have a DEQ2496 that I totally re-designed and rebuilt the analog section of. Way too difficult for most techs, and very time consuming. I'd look at DBX brand digital EQ, or others, or scale some existing analog circuits in a SPICE type program to get the desired EQ response, then add good tone controls to the system if you don't already have them. I hope this is helpful.
These Peerless/Vifa drivers are great (best I know of, all things considered), but difficult to mount since the flange is thin and the ceramic magnets and basket shape is such that rear side venting of acoustic energy is likely to be significantly impeded if the drivers are front mounted even on 1/2 inch baffle material. I would seriously consider having 1/4 inch thick aluminum fabricated for use as the front panel. That may cost too much - I haven't gotten a quote yet (my shop is either the front yard or the bedroom floor so there's just so much that is worth doing myself).
What makes this speaker system rise above most others is the way it works with room acoustics. As tempting as it is to modify the design somehow; adding a ribbon tweeter or a woofer, you would be throwing away it's main feature to a significant extent. An added woofer or tweeter right away has the integration problem, where its acoustic power drops off at closer to 6dB per doubling of distance, rather than 3dB. If the tweeter only operates above 7kHZ, it could put more air into the sound at a certain minimum listening distance and be relatively inconsequential further away, but breaking up the continuity of the array from floor to ceiling comes with a significant price. I've decided not to mess with that.
One of the main reasons woofers are often boomy sounding when you turn up the bass, is because of how a virtual point source interacts with typical listening room acoustics. Two line arrays going floor to ceiling stimulate the room in so many different places, that comb filter effects (cancellations) are largely filled in, for about any location in the room I believe, since the cancellation frequencies will be different for each driver, due to the differing distances to room boundaries and the listener. Not a lot of people understand this. It's a little difficult to put into words that don't get misinterpreted.
Like everybody else, I wanted to second guess Roger Russel; add a tweeter or woofer, maybe use different drivers, but I'm convinced that he's done quite a bit of research and experimenting, and has a very nice system as it is.
Better rear airflow for the drivers, and the rounded cabinet shape that one of the above contributors is doing (or has done) would be even better I would think, but I can't think of anything else I'd want to change.
The black paper cone version of this driver (Peerless TC9FD) sounds real good too. I've built some speakers for my computer with those. My Stereo Holographic soundbar project uses the TG9FD10-08 Peerless drivers that appear to be virtually exactly the same one's Roger Russel is using. They also sound great to my ear. The glass fiber cones are probably better than the black paper, but the difference may be small and I'm still struggling with idea of have 25 light grey cones showing through the grill assembly... I definitely prefer back cones, but those glass fiber cones are VERY nice sounding. The (Polk style) holographic imaging works very well.
I noticed that the 8 ohm version of the TG9 driver has a slightly better top end of the frequency response than the 4 ohm version, according to the published graphs, but the difference is pretty small and can be pumped up with active EQ.
Roger didn't share the schematic for the active equalizer, but does show it's frequency response in the Audio Express article that I have here. 200HZ - 500HZ is the lowest point on the graph (dB scale very difficult to read) but he did say he had about a 15dB peak in the response at 30HZ, apparently relative to 250HZ, with a fast rolloff below 30HZ (very important). There is a gradual rise in the EQ that goes from about 500HZ to 20kHZ. If he's up 15dB at 30HZ, then it looks like he's up about 7dB at 20kHZ. The slope for the curve from 500 to 20kHZ looks like it's about 3dB/octave.
By the way, Gary Pimm pointed out to us that the Behringer analog circuits in his DCX2496 unit are very poorly designed. I have a DEQ2496 that I totally re-designed and rebuilt the analog section of. Way too difficult for most techs, and very time consuming. I'd look at DBX brand digital EQ, or others, or scale some existing analog circuits in a SPICE type program to get the desired EQ response, then add good tone controls to the system if you don't already have them. I hope this is helpful.
Last edited:
Thanks for the info Bob - very useful stuff. Is there a scanned version of the 2006 Audio Express article available somewhere? Seems like an important must read. I use the black paper cone TC9FD a lot and it sounds great - the high end is like a good soft dome tweeter almost. The bottom end in a TL or BLH can reach 50 Hz. I don't know how you can get 30 Hz with them but I guess having 25 drivers, they don't have to move as much so the 2.5 mm xmax is enough to get the displacement needed for the bass. The drivers are in lightweight glass reinforced polycarbonate plastic baskets, I don't think you need to mount them in thick wood. 1/4 in or 3/8 in thick ply or even masonite could work. If you are worried about durability, put a layer of fiberglass and resin over the Masonite to reinforce it. Another option if you are going down the fiberglass route is to make it like a surfboard and use foam core for the baffle structure.
I have a copy of that article. PM me and I'll get it to you. I also have the first article of his prototype using bigger 4" cheap drivers. I was in contact with Roger Russel trough email some years back and he was very helpful in the idea phase of my project.
This year I finally had the time to be able to make my version of it. But I wouldn't recommend it to others if they want to keep their sanity.
I have a DEQ2496 and I have been looking at some mods for it, but I'm not competend enough in that field to even try. Do you have some free time Bob? That's one of the reasons I decided to use my computer with a good sound card to be able to use that as the source and digital EQ to supply the rest.
Good points on the IDS concepts though! I wanted to change a few things, round shape to improve diffraction control. Round inner chamber helping with back wave control. Irregular shaped walls to avoid resonances, thin aluminium baffle for better breathing of the back of the speaker. Making the baffle a sandwich of aluminium and MLV to make it more dead and make up for the thin baffle. Very small wave guide (hole with radius in front of the speaker) to help in the 11 Khz area, this last idea has to be proven yet but I got the idea from here: http://www.diy-hifi-forum.eu/forum/showthread.php?t=2157
Last but not least: improve the looks. (a WAF thing)
I have thought long and hard about both Vifa drivers. In the end I went with the paper cone (even had a conversation about them with Roger) because I seem to have a "thing" for paper coned speakers. The differences will be there, and I think the glass fibre coned speaker might be able to give more detail. Most comparisons I have read between the two describe the paper cones as more laid back/pleasing. That's what I went for in the end. The EQ will take away even more differences but you've got to pick the speaker you like because that is what you get back on steroids with the line array.
The most common error with line arrays is people buy some buy out cheap cones and build the arrays with that. You've got to like the sound of the speaker you're going to use. And I do like the TC9.
This year I finally had the time to be able to make my version of it. But I wouldn't recommend it to others if they want to keep their sanity
.I have a DEQ2496 and I have been looking at some mods for it, but I'm not competend enough in that field to even try. Do you have some free time Bob?
That's one of the reasons I decided to use my computer with a good sound card to be able to use that as the source and digital EQ to supply the rest. Good points on the IDS concepts though! I wanted to change a few things, round shape to improve diffraction control. Round inner chamber helping with back wave control. Irregular shaped walls to avoid resonances, thin aluminium baffle for better breathing of the back of the speaker. Making the baffle a sandwich of aluminium and MLV to make it more dead and make up for the thin baffle. Very small wave guide (hole with radius in front of the speaker) to help in the 11 Khz area, this last idea has to be proven yet but I got the idea from here: http://www.diy-hifi-forum.eu/forum/showthread.php?t=2157
Last but not least: improve the looks. (a WAF thing)
I have thought long and hard about both Vifa drivers. In the end I went with the paper cone (even had a conversation about them with Roger) because I seem to have a "thing" for paper coned speakers. The differences will be there, and I think the glass fibre coned speaker might be able to give more detail. Most comparisons I have read between the two describe the paper cones as more laid back/pleasing. That's what I went for in the end. The EQ will take away even more differences but you've got to pick the speaker you like because that is what you get back on steroids with the line array.
The most common error with line arrays is people buy some buy out cheap cones and build the arrays with that. You've got to like the sound of the speaker you're going to use. And I do like the TC9.
Last edited:
I have a copy of that article. PM me and I'll get it to you. I also have the first article of his prototype using bigger 4" cheap drivers. I was in contact with Roger Russel trough email some years back and he was very helpful in the idea phase of my project.
This year I finally had the time to be able to make my version of it. But I wouldn't recommend it to others if they want to keep their sanity
I have a DEQ2496 and I have been looking at some mods for it, but I'm not competend enough in that field to even try. Do you have some free time Bob?
Good points on the IDS concepts though! I wanted to change a few things, round shape to improve diffraction control. Round inner chamber helping with back wave control. Irregular shaped walls to avoid resonances, thin aluminium baffle for better breathing of the back of the speaker. Making the baffle a sandwich of aluminium and MLV to make it more dead and make up for the thin baffle. Very small wave guide (hole with radius in front of the speaker) to help in the 11 Khz area, this last idea has to be proven yet but I got the idea from here: ViFAST - für 50Euronen (SDS 5 1/4 + Vifa 9 BN 119/8) **fertig! Weiche S.3** - DIY-HIFI-Forum
Last but not least: improve the looks. (a WAF thing)
I have thought long and hard about both Vifa drivers. In the end I went with the paper cone (even had a conversation about them with Roger) because I seem to have a "thing" for paper coned speakers. The differences will be there, and I think the glass fibre coned speaker might be able to give more detail. Most comparisons I have read between the two describe the paper cones as more laid back/pleasing. That's what I went for in the end. The EQ will take away even more differences but you've got to pick the speaker you like because that is what you get back on steroids with the line array.
The most common error with line arrays is people buy some buy out cheap cones and build the arrays with that. You've got to like the sound of the speaker you're going to use. And I do like the TC9.
Wesayso, again, great design work there. I also rear mounted my drivers, on 1/2 inch MDF, and did the 1/2 inch radius rountering of the hole on the front side, but I worry about cavity effect, so I'm not entirely thrilled with that.
Someone at Madisound told me that Peerless and Vifa merged about five years ago, and then got bought by Tympany, which also owns Scanspeak, but about two years ago the production facility for these drivers was moved from Denmark (I think) to China (big surprise), so I'm a little concerned about the new plant in China doing the job correctly. The 8 ohm TG9 has apparently been on back order for several months according to Madisound. Parts Express hasn't carried the 8 ohm version.
Anyway, the 8 ohm version of the TG9 has the flattest and most extended response at the high end, of the Peerless/Vifa 3 inch drivers, if the published graphs are accurate. The TC9 has the significant but not horrible peak at 15kHZ on axis, and is actually pretty flat slightly off axis. The TG9 looks a bit like an eyeball, and having to look at 50 eyeballs is bothering me, so I'll probably go with the TC9 myself, if I go ahead with this project (I'll have to sell my open baffle speakers first to make room for these).
Details about modifying the DEQ2496 are on my website. Again, soldering tiny wires to the SMD DAC chips is very difficult, and only the most experienced should try it. Personally I'm going to try the analog approach, like Russel did. I've seen a picture of the inside of his active EQ box, and it looks very analog. I have some circuits that I'm pretty sure I can modify to get what I want there. The MLV sounds interesting. I'll take a closer look at that.
I also have both AudioXpress articles, which Roger was kind enough to send me after several email conversations (he apparently lives in the Orlando Florida area). The earlier article is where he used very cheap drivers to prove the concept, and he claimed he was very thrilled with the result. Then he knew it was worth rebuilding it with the better drivers and other improvements.
If I come up with a good analog active EQ circuit I'll share it here. In the mean time the Behringer mod can be found on my website at:
Bob's Website
Bob's Website
Production facilities moved longer ago than that, and since originally being owned by Tymphany, ScanSpeak has again become independent (bought out by ScanSpeak Management)
dave
Bob,
Thanks for the link to your Behringer mod. Nice work there - you are very good. I don't have qualms about soldering wire wrap wires to SMD pads - I built a dead bug amp on a foam core board using a SMD type TPA3118D2 class D amp chip flipped upside down (literally looks like a dead bug with 32 legs sticking up in the air). That thing had 0.6mm pitch leads and was it a *pain* to solder. I am more concerned about the overall investment of time into building a what is essentially, should be COTS parts for ADC buffer and DAC output filter. I think TI actually has these analog circuits all integrated onto a single (SMD?) chip. I will check. If you ever do come up with an all analog active EQ for the IDS-25 with TC9FD (paper cone driver) combo that would be fantastic. I for one, will be eagerly waiting for that day to come.
Thanks for sharing.
Regards,
X
Thanks for the link to your Behringer mod. Nice work there - you are very good. I don't have qualms about soldering wire wrap wires to SMD pads - I built a dead bug amp on a foam core board using a SMD type TPA3118D2 class D amp chip flipped upside down (literally looks like a dead bug with 32 legs sticking up in the air). That thing had 0.6mm pitch leads and was it a *pain* to solder. I am more concerned about the overall investment of time into building a what is essentially, should be COTS parts for ADC buffer and DAC output filter. I think TI actually has these analog circuits all integrated onto a single (SMD?) chip. I will check. If you ever do come up with an all analog active EQ for the IDS-25 with TC9FD (paper cone driver) combo that would be fantastic. I for one, will be eagerly waiting for that day to come.
Thanks for sharing.
Regards,
X
xrk971, thanks. I wish I had taken a better picture of the DAC chip and wirewrap wires soldered to it. I had to very carefully lift the DAC inputs and outputs off the PCB, to get rid of the Behringer circuit. And I had to get rid of the entire I/O PCB that was mounted to the inside of the rear panel, thereby losing the calibrated mic input, so that function is gone. The time element makes this project ridiculous. Being one of the very many laid off engineering techs here in Portland, I had time.
I'll try to get to the IDS25 active EQ circuit soon. I just bought a ATEN brand computer switch so I can switch between the new Win 7 PC I recently built, and the old XP PC that has a version of SPICE that works with the M&K woofer circuit that's already modeled. With mods that circuit will give me the perfect low end EQ, with the fast rolloff below 30HZ, but I'll need to dig out a five section pinking filter circuit and integrate it in for the apparently 3dB/octave rise from about 600HZ on up.
Pinking filters create a rolloff, but put in a feedback loop will do the opposite. Maybe a Baxandall section would work better for the slow rise to 20kHZ. I'll have to play with that a bit to get it right. With room acoustics and etc being such variables, I'll also use my four section Baxandall tone circuit as well.
Hopefully this new computer switch will work. Apparently many other switches are full of problems. It switches virtually all of my peripherals (mouse keyboard monitor and USB ports) back and forth between two computers. I actually like XP a lot better than Win7, but I needed a better computer for video editing, so I built that.
I'll try to get to the IDS25 active EQ circuit soon. I just bought a ATEN brand computer switch so I can switch between the new Win 7 PC I recently built, and the old XP PC that has a version of SPICE that works with the M&K woofer circuit that's already modeled. With mods that circuit will give me the perfect low end EQ, with the fast rolloff below 30HZ, but I'll need to dig out a five section pinking filter circuit and integrate it in for the apparently 3dB/octave rise from about 600HZ on up.
Pinking filters create a rolloff, but put in a feedback loop will do the opposite. Maybe a Baxandall section would work better for the slow rise to 20kHZ. I'll have to play with that a bit to get it right. With room acoustics and etc being such variables, I'll also use my four section Baxandall tone circuit as well.
Hopefully this new computer switch will work. Apparently many other switches are full of problems. It switches virtually all of my peripherals (mouse keyboard monitor and USB ports) back and forth between two computers. I actually like XP a lot better than Win7, but I needed a better computer for video editing, so I built that.
In case anybody out there is good at using the SPICE or equivelent circuit modeling program, and has time to modify this circuit to do what you want for the IDS25, here's a starting point. This partial M&K circuit originally had about 20dB of gain peaking at 23HZ. As shown, the lo-pass function has been removed. The rise above 600HZ will have to be figured out separately. It may be a while before I can get to this myself.
Attachments
Do you run Windows 7 Pro? If you have that you could run a virtual XP workstation offered free for download from Microsoft. I've done that for a while since I still had apps that required XP to run. Microsoft calls it Windows XP mode. There are even better free virtualisation applications available that can run the XP Mode from Microsoft on a Windows 7 PC.
I used Virtual Box with VMlite plugin and even ran an old version of Autodesk's Inventor on that! It ran better than the XP mode on Microsoft Virtual PC and it is totally free of charge.
Install XP Mode with VirtualBox Using the VMLite Plugin
Might make things easier for you, you can run it on one PC and Virtual box even lets you run one inside the other...
True, but that would still require running 2 PC's
. Personally I don't trust hard drives older than about 5 years anymore. One of the reasons I run everything on my PC in raid mirror. One PC was ease of use for me. That said, I still run a win 2000 pro PC in the garage as seen in the picture on the previous page.If you need any help Bob, just PM... there are indeed several ways to control the XP PC from your new Win 7 PC.
A very simple way might be logmein.com free edition. That way you can run it from anywhere you have a network connection
.
Last edited:
Whatever you do, don't get Win8 it's just awful. Win7 professional allows you to install a virtual xp machine and I run AkAbak from there. I like XP too. In fact, I was happy with Win2k prof uses a lot less resources and is very snappy on a p4 with 2 GB of ram. Alas, we digress.
Can't wait for the full EQ circuit.
Can't wait for the full EQ circuit.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.