3way XO help greatly appreciated!

[omni]

Grant, It looks like you have been busy at the research........I checked out the link you posted, and it looks interesting, I will have to study it further........Are you gonna try to implement it?The Speaker Workshop learning curve is certainly daunting, however, seems you are progressing. I saw the rabbitz post about the TDFC and from what I have read Zaph say about it runs along the same lines as far as its' durability.......Part of the reason, I think, is due to its' Ferrofluid, plus it's probably simply a well made tweeter. I chose 12dB for a few reasons,some of which I may be wrong, however, here goes: With its ferrofluid, it is already being "slowed down" in its' transient response a bit. Perhaps a 6dB rolloff may slow it down a bit more in terms of transients, thereby increasing "ringing" and sibilance? Also, at times I do play my music loud, though the tweeter does get a smaller portion of the total amplifier power, 75% goes to the woofer..............With the woofer 12dB is definately the way to go, for me, as I believe it will tame the 92dB SPL a little bit more, helping to bring it closer in line with the CA15RLY which already is lower in SPL, Again this is only supposition.............The CA may benefit from a 6dB rolloff, however, creating a bit of overlap in the midrange...........The FRD, ZMA file concept of modelling is definately a new light, compared to my original way of designing a crossover, and I have been studying it intensely. Bit by bit I am beginning to see its' value and will continue down that path. Bagbys' crossover designer5.1 makes use of these files and the FRD consortium information and tools look simpler to use, the more I study them. I have downloaded the programs and hopefully will proceed accordingly....................For Sreten, I have a question or two which may seem premature to ask at this time, however I am going to inquire anyways........ With the FRD consortium tools, once the FRD and ZMA files are processed, using those tools, then imported into a crossover similator, will adding a Zobel into the mix alter response results? It seems to me the addition of the Zobel would. Or is it inserted AFTER crossover is similated or during the design phase of the crossover?..........How are crossover component values determined in order to insert into the model?....................Grant, I look forward to sharing and comparing results as we move through our designs. That's what this forum is all about. I very much enjoy our communication.........Respectfully........Omni.... ...........Where's TINITUS ........ I miss reading your thoughts...............

[grantnsw]

Hi Omni, many thanks for your comments.

My reason for liking the idea of a 6db/O network for each driver (if at all possible) is the low parts cost. A 'full' LR2 xo with quality air-core inductors would probably cost more than tweeters and mids combined. So I have to make compromises. Use iron-cored where necessary and suffer any distortion. Also, I was not aware of the 'slowing down' factor you mentioned, thanks for that. It would be interesting to find out how significant it is.

I guess I could try the 6dB option initially using limited volume and upgrade the xo later when funds become available. Probably the 6dB option won't be acceptable, but I don't know till I try it.

Re: audiotester. It looks good but after scanning the 'doco' it may be limited to 30 days. So, I don't think I'll pursue it.

I still haven't progressed much with SW and 'consortium' tools ( I'm trying to get my elderly mum into aged care). I'm interested in the frd/zma concept and believe they need to be renamed to .txt?
in some circumstances, but I still haven't got that far yet. Re: enjoying communication - *ditto*
thanks Omni, grant

[sreten]

Hi,

A crossover simulator based on SPICE will use the zma files only.
It will give you the electrical levels at the driver terminals.

A more advanced simulator will also use the frd files, if these are
measured in the final boxes at a a reasonable distance you should
get a fairly accurate response.

for the given 3-ways I'd recommend a series 1st order c/o bass/mid,
this frequency chosen to give accurate baffle step compensation.
(Depending on the box baffle this may not be possible, you may
need to go 2nd order on the bass unit to shape its response
correctly, but this will not be a "textbook" alignment.)
Generally you don't need to Zobel the bass unit for a 1st order series.

The mid / treble is harder to say without playing with it.
Possibly 1st order series also or 2nd order parallel.

An exercise in Basta :
Model the mid unit in a small box with various settings of an L-pad attenuator.
Try the same with various values of series capacitors.
This is the same as a 1st order parallel c/o.

/sreten.

[grantnsw]

Hi Omni,
I've just been trying to get the frd file (by editing .txt values) from my woofer's 20year old spec-sheet graph, which is all I have to go by. And the result looks reasonably close for a first attempt, but a little 'peaky', by 1dB maybe.
(I gave up on SPLTrace - too hard!) Its nice to 'play' with this after driving 200+k's every day to see mum.

So, it looks like I'll have all the frd/zma files. Now for some serious SW study, time permitting of course. I just thought you might be interested, so here is the graph (attached)....

Are you at the xo design stage yet? Hope your enclosures look great! ...best wishes, grant

[grantnsw]

Hi Sreten, many thanks. Please excuse my ignorance.....
From what you said, (correct me if I misunderstood) series and parallel can be combined for different drivers? Have I read you correctly here? If yes, this is another revelation for me!

You know that I'm 'out-of-my-depth' on a lot of things, but, hey, I have to learn somehow. And thanks!

I had at one stage read Tony Gee and Argos (Andy Graddon, I think) and others, re series, but I didn't get to comprehending the 'zeta' factor'! May now have to re-visit it.

I have no idea what a series 1st order bass/mid combined with a parallel 2nd order mid/tweet circuit would look like? Way beyond cookbook stuff! Also, unfortunately my Basta has expired. grant

[sreten]

Originally posted by grantnsw
Hi Omni,
I've just been trying to get the frd file (by editing .txt values) from my woofer's 20year old spec-sheet graph, which is all I have to go by. And the result looks reasonably close for a first attempt, but a little 'peaky', by 1dB maybe.
...best wishes, grant

Hi,

The response appears to measured in a box so that graph includes
the baffle step response of the box, accounting for 6dB of the rise
between 100Hz and 1kHz.

/sreten.

[grantnsw]

Sreten, oic!, tyvm, your amazing!

[sreten]

Originally posted by grantnsw

I have no idea what a series 1st order bass/mid combined
with a parallel 2nd order mid/tweet circuit would look like?
grant

Hi,

Imagine a 2-way speaker box with 2nd order parallel crossed
over to a subwoofer that wired as a series 1st order crossover.

/sreten.

[grantnsw]

Sreten, I read your 'bio' that you are an Elect. Engineer.
I also did 4 years of that in the early 70's! but thats ancient stuff now, besides, long forgotten eg: FFT, Kirchoff etc.. also got ham licence then VK2...(still current) Went on to other things....
So, I respect/acknowledge your advice! , grant.. and many thanks!

[omni]

Grant, greetings from the US at 1:45 am ! I just got done making traces of my Peerless Woofer zma and frd in the trace program and don't know if I did it right but I guess I will find out. I have downloaded the CA and 27TDFC zma and frd files and am getting ready to put them into processing...............Sreten, Have you ever used TRACE? If so, here are a few questions: on settings for registering amplitudes and frequency values, how and where do you choose to set these? For the woofer I used the freqencies of 20Hz and 500Hz, but I am not sure this is right. For the amplitudes I am not certain I did it right. I set them just above and below the highest and lowest points of the plotted curve that was on the grid of the charts from manufacturer. Is this right?......................Grant....My cabinets are coming along SLOWLY........I am having trouble getting the baffle painted the way I want it. I am using a paint that reticulates and am having trouble getting the pattern consistent...........Being the graphic artist that I am, I still have not been able to totally "let go" of my perfectionism...............I will be attempting another spray today. Part of the delay has to do with allowing drying time between coats..............it's costing me days, then I will be ready for the clear coats.............HOPEFULLY...........then mount the baffle, finish sealing it up, put my damping material in, and finish installing the backs, then install the drivers ready for the crossovers. In the meantime, I will be playing with Unibox, FRD Baffle diffraction simulator, FRD Combiner to process these files, then off to SpeakerWorkshop AND Bagbys Crossover simulator/Designer programs..............Grant..........if you haven't checked it out yet, check out Bagbys Crossover Designer 5.1. To me, it looks a bit more user friendly than SW. Also, it has graphs which show responses, and allows for modelling by changing components for tweaking. I'll have more input on this in the next few days, after I have played with these tools...........Sreten,, Please stay in touch..........you never know when I may need you to help me land the plane......................Respectfully........... ..Omni