3way XO help greatly appreciated!

Filter look fine

If values are correct is difficult to say ... you might need adjustments ... over time

If mid is up to it I would try even smaller inductor .... and it may need RC

You should try reversed polarity on mid

Grant, I read your post here, and was quite intrigued because I am working on a 3 way design using the SEAS CA15RLY midbass crossed over in the range of 250-300 Hz low end and 2500-3000 Hz high end. Also I am using the SEAS 27TDFC tweeter, so naturally, your post caught my eye. Let me preface this also by the fact that I am NOT an expert or guru!....... At any rate I am Zobeling both midbass and tweeter and am not gonna Zobel the woofer which is a Peerless SLS 830669. For my Zobels I used the calculation procedure Rod Elliot describes in his manuscript on passive crossover design on his website : sound.westhost.com aka Elliot Sound Products. The reason I am using this procedure is because it uses actual parameters from my drivers a little more in depth to calculate values. Also, I researched on this forum extensively and the overwhelming consensous appears to be that it is the preferred method. It gives values for the capacitors a bit higher than the Lalena calculator, and again, from extensive research, reports are that the C=Le/Rc squared, Rc=1.25xRe formulas provide the minimal values for components, which probably need to be adjusted upwards. However, Zobel networks are reported to be forgiving in their nature, so the design method one employs may simply be based on personal preference.........I am not sure...........At any rate the values for my Zobels for the CA15RLY are: C=22 microfarads and R=7.25 ohm, for which I will probably use an 8 ohm Mills resistor. For the 27TDFC C=2.1 microfarads and R=6.1 ohms for which I will use a 6 ohm Mills..........Without test equipment, where did you get your impedance measurements? I would like to compare, if possible. I am using 12dB per octave throughout, and will be reversing polarity of midbass, I am not certain about the 6dB network polarity, however, TINITUS recommends reversing it, so I probably would follow his suggestions, as I have grown to respect his input on these matters.........I have made many changes to my system based upon his input. Your questions pertaining to component values may be answered by someone more experienced, however, it is my understanding that the impedance of the particular driver, mounted in the cabinet, at crossover frequency will dictate the component values via your formulas. I am measuring mine with an old frequency generator using the voltage divider method, recalibrating the voltages to a reference resistor for each frequency/impedance measurement. Also am comparing the measurements to impedence charts from the manufacturer and the impedance charts available on Zaphs' site. Hopefully will get a respectable outcome. I look forward to reading of your results since we have similar projects going on here..........Respectfully....Omni

Omni, good to hear you have progress

I like your approach with the "zobels" and it looks like reasonable values .... along the way you could try and "finetune" the zobels by ear .... at the moment I cannot live without double zobels/RC

Edit... remember that what seems to be 6db ... on paper....will rarely be excactly 6db ... nor will a 12db be excactly 12db ... sometimes less, other times more

I think of it as a modified smooth and soft 12db rolloff .... some say it is actually more like 10db ... but getting steeper at the end of the slope

grantnsw said:

1) I would prefer to use 'off the shelf' inductors, ie lowest DCR dependent on price! (Solen $75AU!)

2) the combined impedance of the tweeter(@2700) and voltage divider is close to 6.3 Ohms I think.
I would like to use a 0.56mH choke here, so the LC-product for 2700 yields 6.2uF for the Cap.
But I'm not sure if the resulting impedance @ 2700 matches the 6.3 Ohms mentioned above?
Is 6.2uF the correct value (in conjunction with 0.56mH) please?

3) the P25 woofer crosses at 270Hz where Z= almost 8Ohms and 90dB SPL; thereafter 6dB/Oct
attenuation. Do I need to Zobel it then? P25: Le=2.2mH and Re=5.7 (from specs).
So, again from Lalena for Zobel, C=43uF and R=7 Ohms. Are these values correct? Also, the
CA15RLY impedance rises sharply after 2700, do I need to Zobel it too? ---->C=16.7uF,R=7?

4) the bandpass values I have calculated also slightly differ from Lalena?

5) and lastly, should the Seas mid connection be reversed? I guess I could try it!
Thanks very much for reading all this, and very best wishes, Grant
(I am aware that s'ware calculators have big limitations but these drivers are very well known I
think and I have no test equipment). Maybe later I'll replace the D25's with Seas 27TDFC's.

Click to expand...

Omni mentions Elliot Sound in his post above, I think the info published on that site can provide clear answers to most of your questions, in particular with issues regarding impedance behavior at crossover points. Unlike many such sites this one is written in plain English and is accessible to anyone interested in DIY crossover design, not just engineers and techies.

Here's the link:


http://sound.westhost.com/lr-passive.htm

I don't suggest going as far as winding your own inductors though

Hi,

"Why tweeter needs zobel..."?
I cant say it does .... though from my listening test I have found that high frequencies sound more smooth, might be phase related

"Zobel on bass..."
Actually you do have zobel on bass already .... what I suggest is to ad another zobel with smaller C .... BUT, two zobels wont follow "textbook" values
I do have a notch/ tilt curcuit on bass too.... outside its passband

You can steal a bit here and there, but that is not likely to work with the excact same values ..... because different drivers interacts differently with each others ... but ofcourse it gives a clew about where to start

I do like the KISS method, but it is also important to have some "tools" of adjustments

I would say a inductor/R in parallel with mids lowend crosspoint is needed to get it perfect .... but very difficult to get right ... and it might work without

Same thing with tweeter... parallel inductor/R is VERY difficult to get right

Thats why I need to say ... DO NOT just assemble your crossover and mount it inside your speaker .... I am sure you will enjoy making small adjustments from time to time... part of the fun

Grant, Sounds like you got the research well under way! Cool...... So, the reason I am using a Zobel for the tweeter is that I believe, as Tinitus does, that there is a smoother response at higher frequencies associated with flattening the impedance a bit with a Zobel. But the main reason is that a more stable impedance curve for the tweeter will lend a more accurate impedance maintenance for my L-pad circuit, should I need to pad down the tweeter. In other words, the overall impedance of the tweeter circuit should remain as constant as possible, even when an L-pad circuit is inserted, AFTER the crossover is designed. I won't have to refigure the components again, and the crossover, with L-pad will remain accurate. The addition of the L-pad should not adversely effect the crossover values. The Zobel will facilitate this. The Le for tweeters is a parameter that many manufacturers do not provide, however with the SEAS 27TDFC, I was pleasantly suprised to see it listed in their spec sheet. Zaph also reports this on the TDFC, I believe...... Le= .05mH or.00005 H........ I have chosen not to use a Zobel on the woofer for a few reasons. One is that the impedance, according to Peerless impedance curve, is fairly flat, going from 8 ohms at 60Hz, and dipping down a bit to about 7.5 ohms and back up to 8 ohms at 300 Hz, so for the most part, impedance is reasonably flat in the frequency range which I will be using. My impedance measurements reflect that curve. In addition, and there may be debate on this one, Zobels tend to lower driver impedances, and I don't want the woofer impedance to get too low, especially in the lower frequencies, which could cause strain on the amplifier, with potentially hazardous results. I have spoken with a few designers, and there reports that woofer Zobels can cause amplifier problems, and I simply don't want to take that chance yet. I don't need 75% of my amplifier power, which the woofer draws, causing problems, especially at extremely low frequencies. I don't have a high end amplifier, so I won't take the chance, at least now. Maybe this is ignorant or short sighted, but for now, this will be my approach. I am sure that there are a lot of successful designs which use a woofer Zobel, so I don't want to dissuade you from implementing one, this is just my design choice for now. HOWEVER..............I can look at impedance graphs till the cows come home and they may not mean a **** thing, because I will be designing my crossovers based on the impedance measurements I plot with the speakers installed in my cabinets WITH Zobels for midbass and tweeter intact. These measurements are likely to be different from the free air measurements I have taken so far...........One more thing on Zobels................I will be placing them CLOSEST to my drivers on the crossover circuit, as they BECOME part of the driver. David Weems, in his speaker design book recommends Zobels be placed right across the speaker terminals. I think both attitudes will suffice, but it is an often overlooked detail....... Check out Rod Elliots' Manuscript on this.................Since you are replacing the midrange with the SEAS CA15RLY unit, I am curious to know what the internal volume is for your midbass cavity. Your crossover point at 270 will put the CA in a range where it will experience some amount of pistonic motion, which may require a volume for its' enclosure to be tuned to a particular frequency. This has to do with internal standing waves, which will have a direct effect on smoothness of midrange sound. Vance Dickasons' Loudspeaker Design Cookbook does an excellent job of covering this subject, and it was my reference when I designed my enclosures. If you wish to pursue this, I will be delighted to share with you my experience, along with your request that I share with you my impedance measurement experience. We got time, so I am cool with that. My cabinets will be completed, with drivers mounted in the next few weeks, so keep us posted and I wish you great amounts of joy in your project...........Respectfully.............Omni

I dont really build many speakers... too slow, and before I get started I get new ideas, or have no money.....but I have been working with my present speakers a number of years now .... so you can say the same speaker presents many different versions

To know the sound of a violin is a very usefull "tool" to evaluate the sound of your hifi

I can listen to music fore a long period, thinking it sounds almost perfect... and then suddenly I am in need of something more ....its kind of like an urge or maybe I just get bored, and then I try different adjustments....and the frustration starts .... I admitt, its not easy at all

This very moment I am trying different things ....

Now I have a zobel on bass ......10uf/39R ...and you may say, "whats that gonna do, with 39R" ... but its there

Your filter looks very similar to what I use.... its a good start, very simple, and have many possibilities

BTW, what Omni says about placing corrections as close as possible to driver, might be true .... I was told the same thing years ago, but havent tried it yet ... not yet done with adjusting....

Hmmm..... zobelling tweeters but not bass units.... very strange.

take a look here :

http://www.troelsgravesen.dk/Diy_Loudspeaker_Projects.htm

/sreten.

".....very strange" ?

Zobels does a number of things

A not so obvious reason to use zobel is that it "calms down" the impedance .... its not as straight as it looks on measurements.... it ressonates ..... a zobel helps taming that ..... and high frequencies are more sensible to that .... but a woofers highest frequency reacts to it as well

Grant, I am not certain if you will need to Zobel your woofer or not at this time, as my choice , at this time, not to Zobel mine is based possibly on fear of too low of an impedance causing amplifier problems............HOWEVER, in addition, here are some of my thoughts from a musical standpoint : The lowest frequency of a 4 string bass guitar is 42 Hz, which is the open E string. Perhaps, the musical tones that will be reproduced lower in frequency will come from the kick drum. And I am describing contemporary rock music. I also listen to jazz, as well as classical music, which also has a myriad of frequencies inherent to its' performance. Sometimes I enjoy playing it louder which tends to bring out the nuances and sparkles a bit more, for which I have a great apreciation with their imaging and soundstage, so emphasis on the lower registers is not a huge concern for me, however, understanding these low registers do contribute to the overal tonality of the music I listen to. So a great deal of the essence will be coming from mid and upper frequencies. So I may lose a bit in the lower frequencies, soundwise, but right now I am willing to sacrifice that..............UNLESS.........SRETEN, as he described my philosophy as "Strange" and directed us to troelgravesen website........would be willing a clarify his observation AND direct us to the link on that site which may shed further light. So if you are reading this SRETEN, please shed some light................The midrange cavity is of a primary concern for the CA15RLY, which would take a lot of space to discuss, however, I definitely recommend you read Dickasons' dissertation on that . I will say from what you said of your cavity being 1.5 Litres, I am of the thought that this may not be optimal for the CA, especially at your crossover point of 270 Hz. If you want to get into detail about it, thats cool, but if you can, read Dickasons material in the cookbook to familiarize yourself. It is in his section on midrange enclosures............. I do plan on posting some photos after this weekend, as I am right in the middle of surface finishing the Baffle, Maybe I will post them on this thread as your project proceeds. So, I hope you are finding this speaker endeavor an enjoyable one............I know I am..............Respectfully............Omni