Hello Widget
Hmm well that settles that. I don't have that much time to finish cabinets. I will have to find another finish that may not be as spectacular but still looks good.it can take a hundred hours of love, but when you are done it is quite spectacular...
Well it's raining here big time so I was able to get a set of driver measurements mounted into the cabinet. Took them at 1M at tweeter level. I then changed them to text files and inported them into LEAP. I have found that LEAP is a good tool and is reliable for predicting the outcome. I like to use the simulations as a sanity check to make sure I am on the right track.
This curve is from averaged measurements to see if the network, attenuation, and baffle placement look good. So far it looks good. Now I just need a week to get the networks built.
Rob
Just for fun I decided to see what the waterfall plots looked like for the 3 drivers measured over the frequency ranges they are used in the system. They look damn good. The 108H was a real surprise. I would say the aguaplas is certainly doing what it is supposed to be doing. The first 2 are the 108H the second are the 104H and the last is the 050 Gold tweeters.
Rob
Hi Rob,
By what measure are these good plots? I.e., are these time and amplitude
scales (and resolution) appropriate for these type of devices... That there
appear almost no time-lingering resonances in the plots, I can see; I'm
assuming that is the merit being judged. So my question is, would other,
less capable devices appear any different in these plots? If you or someone
else have particular experience with these plots and can just say that these
are good ... well, OK
...or I can not be lazy and make some comparative measurements myself
(don't have CLIO, but Mac-based FuzzMeasure which is also reasonably
capable). I should just mount up some 104H-2's and do this.
Really nice to see this project coming along.
Hello Grumpy
What makes them good is how quickly they decay and how clean the plots look. The time scales are adjusted so you don't loose resolution and waterfalls can go low enough. The lower you go the longer the time window needs to be. The software will automatically limit how low the plot goes depending on the length of the time window. So if you look at the 108H the window is 13msec and they get shorter as you go up in frequency and change driver ranges.
Here is a plot of a 2435 on a PTH1010 and a 2425 on a 2344. You can see that the plot for the 2435 combination looks better than the 2425 as an example. Same time and frequency range.
The $64,000 question is just how audible any differences that come up in the plots really are.
Rob
Thanks Rob, the compression driver/horn plots are closer to what I'm used to seeing.
The direct radiator plots were so clean in comparison, it made me wonder if I was
missing something .
Correlating the plots to audible differences or artifacts is probably the topic of an AES
paper or two. Whether this type of plot is useful for lower frequency, direct radiator
measurements, is what I'll change my question to... Is it useful? or is there just typically
nothing to see?
I've seen cabinet resonances show up, on wide-band waterfall plots, in the lower
frequency range, which seems like a useful enough reason to run them, including
that range.
These graphs are really interesting, thanks for sharing Rob!
Where those diaphramgs aquaplassed?
Another question is how much of these time delays come from the horn itself. The 2344 is said to have ripples between 1k and 2k (can be call them HOMs?). http://www.audioheritage.org/html/pr...bl/4430-35.htm
That would also be interesting to see the same graph for a H9800!
Rob,
The plots you posted look really promising. I was a little surprised by the dip from 7K-10K on the 050Ti, but as far as controlled decay, they look great. I never saw one before on the 108H.. very nice!
You don't happen to have a 095Ti/093Ti or 046Ti from a XPL200 that you can run, do you?
Got the crossover parts laid put just have to figure out how to fit them on to the pcb's. Lot's of caps going CC and the bandpass for the 108H has large values.
One of these days I will measure them and post the plots.You don't happen to have a 095Ti/093Ti or 046Ti from a XPL200 that you can run, do you?
Rob
Rob, have you got them running yet?
Hello Cosmos
Not yet I am slowly moving along. Combined with a CC network in Solens and the low dcr values for the larger coils the networks are quite a bit bulkier and down right heavier than I imagined. Basicaly I need to get more cabinet work finished to see how the networks will load in. It won't fit on 3 double sided PCB's as it is now so the 160uf capacitor bank and one of the 2 largest coils will have to be off the boards mounted to the enclosure.
I am also going to spike and mass load the upper cabinet so it is isolated from the bass cabinet and may be using neoprene on the baffle like in the XPL series. I think I found a local source for the "correct" material. I have to get a sample and see what it looks like.
I have a couple of the boards finished so once I figure out how it will all go together I should be able to get them up fast once it all comes together.
Thanks for your interest
Rob
I managed to get all the network boards built and ended up using 4 seperate PCB's as their size is a real surprise. They have to be measured to see if the voltage drives are correct. Still need to add the DC bias but I can see if they are correct without it. Have to make the new boards for the DX-1 next. Slowly it's finaly starting to come together. At the rate I am going maybe by X-Mas I will have them actually working
Rob
"I could be arguing in my spare time"
When Guido made my XO , he used the "new" style TIBQ. I have a an additional board which sits in the bottom of each speaker w/many Caps on it. Since yours are from scratch I would guess all of your componets are larger than originals , it is large for a XO !
I enjoy following this thread
Mark
Changing to Legacy Audio and started with a Silver Screen HD for my center between the 250TIs
There are currently 1 users browsing this thread. (0 members and 1 guests)