The good news: I've made my first few measurements....
The bad news: during the fitting of the outer front baffle, one of the wires of the full range driver set came loose. So it's not firing on all cilinders after all...
.
I had noticed the left side playing (a lot) softer with the old filters, but did not make much of it at that time. That is, until I made my measurements! All kinds of theories crossed my mind...
A lot of difference, especially the top end was lacking! After doing all simple tests I could think of, like swapping amplifier sides etc. I figured I'd start up Vituixcad, sure
enough... the difference between flat and center group missing is what looked most likely:
An impedance check of both arrays should be able to confirm that...
Yep! One of these is showing 4 parallel groups of 5 drivers in series! Last time I checked both, they were still identical, but I didn't have the outer baffle on yet.
In hindsight I remember moving the inner baffle on one array 'carefully' (with a hammer!) to position it... a few well meant hits with that hammer...
As I'm off work today, one would figure I get to repair it right away. But it's my son's birthday, so that's a no go. I am a patient man, I've had a lot of practice.
I could have kept this all a secret, but hey... maybe it will be helpful for one of you guys one day...
one can never make to many measurements to check all is well in your world!
The bad news: during the fitting of the outer front baffle, one of the wires of the full range driver set came loose. So it's not firing on all cilinders after all...
. I had noticed the left side playing (a lot) softer with the old filters, but did not make much of it at that time. That is, until I made my measurements! All kinds of theories crossed my mind...
A lot of difference, especially the top end was lacking! After doing all simple tests I could think of, like swapping amplifier sides etc. I figured I'd start up Vituixcad, sure
enough... the difference between flat and center group missing is what looked most likely:
An impedance check of both arrays should be able to confirm that...
Yep! One of these is showing 4 parallel groups of 5 drivers in series! Last time I checked both, they were still identical, but I didn't have the outer baffle on yet.
In hindsight I remember moving the inner baffle on one array 'carefully' (with a hammer!) to position it... a few well meant hits with that hammer...
As I'm off work today, one would figure I get to repair it right away. But it's my son's birthday, so that's a no go. I am a patient man, I've had a lot of practice.
I could have kept this all a secret, but hey... maybe it will be helpful for one of you guys one day...
one can never make to many measurements to check all is well in your world!
Attachments
Left array is back in the garage. I'll try and fix the problem over the weekend.
In the mean time I've been reading a lot, but I've also played with the proper working right array (obviously). As said, the Vituixcad sims made me rethink a few DSP decisions. For a while now, I have tried to limit/avoid pré-ringing in my DRC-FIR templates.
A while ago I said to be interested in the other side of the impulse, looking at how it arrives at the ears, being able to see if there's anything arriving early etc.
Here's a graph showing the right array, 3 ms Window, Time Range 10 ms, Rise Time 0,1 ms:
Smoothed 1/48, can't turn it off completely. A look at the reversed IR.
I bet I've listened to a lot worse looking graphs than this. This has been an attempt to optimize the first wave front, and let the room have it's play with it afterwards. Only adjusting the (overall) balance when filtered/seen with a Psy filter.
Here's the (non reversed) IR:
I've also used the measurement to check/see what my latest cross talk ideas would look like with real measured results. Can't wait to really try that! The idea seems to pan out and work to fill the ~1850 Hz dip due to cross talk. With an even response on both left and right ear, no funny differences like previous attempts showed.
In a graph it looks like this:
If nothing is done, you get something close to the light green graph at both ears. Aside from head shading at higher frequencies, that is. My proposal has a fixed signal added to the IR, at about a floor bounce interval after the main pulse (close to ~2ms). Left and right differ (one of them is inverted) and that is why they actually sum favorably. The average of a single channel will hardly be different from an IR without the 'synthesized' floor bounce. Red and Blue traces represent the L and R channels, dark green is their average.
The main part that makes it different from a real bounce is that one side is inverted. As far as energy and FR curves this pans out. The bigger question: how does it sound? Combined with earlier mid/side EQ? This tweak could be done to the left + right signal, but I can also split it out to mid/side.
It will fill the usual hole, and hopefully (phantom) mimic the sound of a signal coming from straight ahead. Together with the mid/side EQ already applied, tried and tested.
In theory, it will work similarly at (slightly) off axis positions. As the virtually created bounce will never show peaks/dips at the exact same location in the left and right channel.
In the mean time I've been reading a lot, but I've also played with the proper working right array (obviously). As said, the Vituixcad sims made me rethink a few DSP decisions. For a while now, I have tried to limit/avoid pré-ringing in my DRC-FIR templates.
A while ago I said to be interested in the other side of the impulse, looking at how it arrives at the ears, being able to see if there's anything arriving early etc.
Here's a graph showing the right array, 3 ms Window, Time Range 10 ms, Rise Time 0,1 ms:
Smoothed 1/48, can't turn it off completely. A look at the reversed IR.
I bet I've listened to a lot worse looking graphs than this. This has been an attempt to optimize the first wave front, and let the room have it's play with it afterwards. Only adjusting the (overall) balance when filtered/seen with a Psy filter.
Here's the (non reversed) IR:
I've also used the measurement to check/see what my latest cross talk ideas would look like with real measured results. Can't wait to really try that! The idea seems to pan out and work to fill the ~1850 Hz dip due to cross talk. With an even response on both left and right ear, no funny differences like previous attempts showed.
In a graph it looks like this:
If nothing is done, you get something close to the light green graph at both ears. Aside from head shading at higher frequencies, that is. My proposal has a fixed signal added to the IR, at about a floor bounce interval after the main pulse (close to ~2ms). Left and right differ (one of them is inverted) and that is why they actually sum favorably. The average of a single channel will hardly be different from an IR without the 'synthesized' floor bounce. Red and Blue traces represent the L and R channels, dark green is their average.
The main part that makes it different from a real bounce is that one side is inverted. As far as energy and FR curves this pans out. The bigger question: how does it sound? Combined with earlier mid/side EQ? This tweak could be done to the left + right signal, but I can also split it out to mid/side.
It will fill the usual hole, and hopefully (phantom) mimic the sound of a signal coming from straight ahead. Together with the mid/side EQ already applied, tried and tested.
In theory, it will work similarly at (slightly) off axis positions. As the virtually created bounce will never show peaks/dips at the exact same location in the left and right channel.
Attachments
Last edited:
Anyone in the environment of Assen Drenthe,NL has some spare TC9FD18-08?
I've got two broken drivers and only had one spare. Still can't figure out why they died on me. Both measure "open". I guess something must have gone wrong in transporting the array between garage and home (and back).
It started with one busted driver. When I measured in my garage it turned into two broken samples. Pretty frustrating as it is a bit of a mystery.
During transport and handling the array, it's quite possible the drivers in question were pushed in by some misplaced hands.... I had help both times but can't be sure something like that happened. It remains strange that I've had 2 of such accidents.
Oh well, there's a first time for everything...
I've got two broken drivers and only had one spare. Still can't figure out why they died on me. Both measure "open". I guess something must have gone wrong in transporting the array between garage and home (and back).
It started with one busted driver. When I measured in my garage it turned into two broken samples. Pretty frustrating as it is a bit of a mystery.
During transport and handling the array, it's quite possible the drivers in question were pushed in by some misplaced hands...
. I had help both times but can't be sure something like that happened. It remains strange that I've had 2 of such accidents.Oh well, there's a first time for everything...
Thanks for the kind offer, but that could take quite a while.
I figured if anyone close by had any spares around I could go fetch them and move on quickly. I'm feeling pretty bummed out and can't quite get my head around it. I'm pretty sure I only had one broken driver when I measured them in my room, as the impedance at lower frequency still showed about 8 ohm at bass frequencies. But when I moved the array to the garage, it measured ~10 ohm. I replaced the broken driver in the full range group with a spare and found the other broken one soon after (in the top most group). Sadly my other spare has a dented cone, not going to put that one in there.
I'll be sure to order some spares soon (never had that much laying around, I ordered 52 in 2011). Quite disappointing not being able to fix it this weekend.
I figured if anyone close by had any spares around I could go fetch them and move on quickly
. I'm feeling pretty bummed out and can't quite get my head around it. I'm pretty sure I only had one broken driver when I measured them in my room, as the impedance at lower frequency still showed about 8 ohm at bass frequencies. But when I moved the array to the garage, it measured ~10 ohm. I replaced the broken driver in the full range group with a spare and found the other broken one soon after (in the top most group). Sadly my other spare has a dented cone, not going to put that one in there. I'll be sure to order some spares soon (never had that much laying around, I ordered 52 in 2011). Quite disappointing not being able to fix it this weekend.
Last edited:
I literally have no idea. These arrays have seen al lot of duty and maintenance along the way. I've never had this happen before. It made me think about a discussion a long time ago about drivers being in series. But they have not seen any real action yet.
Oh well, I'll replace the broken driver and move on from there. Scanning the web to find the fastest delivery to order a few of these. I'll order more later to have some reserves.
Edit: placed a first quick order for 6 drivers.
Oh well, I'll replace the broken driver and move on from there. Scanning the web to find the fastest delivery to order a few of these. I'll order more later to have some reserves.
Edit: placed a first quick order for 6 drivers.
Last edited:
You could say that!
I'm just hoping the rewards will be worth it...
It would help ease my mind if i could figure out what happened that caused two open circuit drivers... transport damage seems the most likely, as for both it happened after a physical move. I'll take precautions the next time I move them around.
Playing with one of the broken drivers as I write this. I'll probably dissect it to learn more...
Nothing much to see after opening up the driver.
In other news, I've made a correction for the right side and tested it. Sounds good and promising. Here's the FR with psy filtering pré room curve:
Filtered 1/12 oct:
Now back to the waiting game...
In other news, I've made a correction for the right side and tested it. Sounds good and promising. Here's the FR with psy filtering pré room curve:
Filtered 1/12 oct:
Now back to the waiting game...
Attachments
Had a bit of fun yesterday with a single array with ambience speakers and subwoofers. Kind of a Mono+ setup. Browsing trough my music library was fun and engaging, I can't wait to get the other one fixed.
Anyway, my girlfriend came in and after being silent for a while she said: You only have one of them here, and already it sounds amazing! Next she took over the control of the library making me think: Yes, this will work out fine.
Anyway, my girlfriend came in and after being silent for a while she said: You only have one of them here, and already it sounds amazing! Next she took over the control of the library making me think: Yes, this will work out fine
.
They weren't powered up, so that last part (smell/burnt) doesn't apply. Believe me I've thought about it.
The only thing I can come up with is if the cone is pushed in half sideways, the connection to the coil and the 2 small flexible wires feeding it could be under stress. Not those wires themselves, but at their extension around the spider that connects to the coil. Visually I can't see anything wrong with them though.
Or... I could simply call it a freak accident and move on. 6 new drivers will arrive tomorrow, 12 more are on their way, sourced from another supplier. After that I may even get more to add to the reserve stash. But for those I'll patiently wait to get them at a nice discount or sharp price. Here, in Europe they still are way more expensive than at the other side of the ocean.
I never had a broken driver before. For everything there must be a first time.
The only thing I can come up with is if the cone is pushed in half sideways, the connection to the coil and the 2 small flexible wires feeding it could be under stress. Not those wires themselves, but at their extension around the spider that connects to the coil. Visually I can't see anything wrong with them though.
Or... I could simply call it a freak accident and move on
. 6 new drivers will arrive tomorrow, 12 more are on their way, sourced from another supplier. After that I may even get more to add to the reserve stash. But for those I'll patiently wait to get them at a nice discount or sharp price. Here, in Europe they still are way more expensive than at the other side of the ocean.I never had a broken driver before. For everything there must be a first time.