Understanding the crossover filtration on the RTi12

Starting to mod my old rti12's , one had popped a tweeter, not sure if from a amp clip or just flat out over powered them . Regardless I'm going to try a few different new tweets in them from dayton and peerless/typhany . While into the speaker i figured i would go ahead and update the crossover with better caps, inductors, and resisters.
I'm far from a electrical engineer but have a decent understanding of crossover componants and what not. Ive built a few diy speaker projects and used bassbox pro 6 and xover pro 3 software to design and simulate with excellent results. Unfortantly the pc that had the software on it died and i cant find the cd's .( Moved twice since i got the software). Anyway i managed to get a schematic of the rti12 xo and there are a few things in it that i dont understand such as its topography especially on the mids has me confused on exactly why polk used certain values and that the filtering doesnt seem to follow linkwitz/riley , bessel, or butterworth or even cheybyscheff architecture . I would just like to understand the why and how of the architecture better . The tweeter is padded down clearly but uses a near cheybyscheff type 12db xo which should be causing a slight db gain just above the xo point .

Maybe some of the more tech savy guys here can help me understand some of this midrange xo , maybe some of it is impedence stablization or something. The 220uf cap followed by a 1.3mh coil in series is really confusing with a Q way out there!

Below is the xo schematic and what info i could find on the drivers minus the tweet. Anyone got software that will model this ? Or understand what polk did here? And thanks and sorry for the long first post lol

x432f4vwco0y.jpg

nckeld4f1lmo.jpg

r56yyfguica4.jpg

Comments

  • You guys got nothing ? Lol ok ill take a jab at some of it myself , the tweeter is fairly ez to figure out . 12db slope at around 4300hz with a Q somewhere between .8 to 1.1 depending on exact impendence of the tweet(polk customer service stated its roughly 3.5 ohm) indicating a cheybychev architecture . Odd as those are hardly seen or used in xo's that ive seen . Generally they cause a slight rise in the response curve just above the xo point (e.i. not a flat response curve like a bessel or butterworth arrangement) then its level is padded down some with 2.5 ohm resistor dropping its output a db or so across its range.

    The woofer low pass also seems to follow a cheybychev type 12 db sloped filter crossed at around 137hz with a Q of around .9 to 1.1 depending on the actual impendence . The drivers tested around 2 ohm average from 50hz to 200hz. 3 drivers in series so 6 to 7 ohm .

    Now the midrange has me confused. The Re i found for them ( the rd5537-2 driver) was from another forum and may not be accurate but if so 2.33 is pretty low and that 220uf cap followed by the 1.3mh coil would create a high pass of around 200 to 250hz but at a Q of around 1.10 or another +1 db gain slope just above xo point. Any higher resistance beyond 2.7 ohm for that filter puts it way out of range of a workable Q or puts it into a +2db or greater cheybychev arrangement, not likely.
    The two mids are wired strangly series/parrelled with a 2.7 ohm resistor giving the total circuit around a 5ohm base load. But i cant follow or dont understand the 10uf and 68uf caps , clearly one must be an addition 6db cascaded hipass , 68uf ? and the other the low pass, 10 uf ? Wouldnt the lp or rather the top end of a bandpass need to be coil ? The only coil is the 1.3mh in series creating the lower portion of the bandpass aka midrange highpass filter in unison with the 220uf cap.

    So am i on track or way off base somewhere here?
    No midrange band pass ? All signals above 200hz or so are fed thru these mids including tweeter ranges?

    Hoping someone chimes in with some logic lol
  • Tony M
    Tony M Posts: 11,009
    edited February 2018
    You might be on to something! These tweeter resistors seem to burn out a lot! Why does this happen so often?

    I need to fix my RTi12 cross-overs to get the tweeters to play again.

    I bought a set of RTi8's and their tweeter resistors were blown also. I didn't see any damage on one cross-over so I bought a new set for each speaker. The tweeters came alive again, but those white ceramic resistors are the same on the replacement ones. I didn't expect them to be changed, so no problem. I just know in the future those things need upgrading. Another forum member mentioned to up the resistance from 5 to 10 I think. I'd definitely research and get advice again when the time comes and I have them in my site and hands. ;)
    Most people just listen to music and watch movies. I EXPERIENCE them.
  • Tony M wrote: »
    You might be on to something! These tweeter resistors seem to burn out a lot! Why does this happen so often?

    I need to fix my RTi12 cross-overs to get the tweeters to play again.

    I bought a set of RTi8's and their tweeter resistors were blown also. I didn't see any damage on one cross-over so I bought a new set for each speaker. The tweeters came alive again, but those white ceramic resistors are the same on the replacement ones. I didn't expect them to be changed, so no problem. I just know in the future those things need upgrading. Another forum member mentioned to up the resistance from 5 to 10 I think. I'd definitely research and get advice again when the time comes and I have them in my site and hands. ;)

    Ya ive popped a few tweeters but never the resistors yet . I was powering my rti12's with a carver m1.5 . Then i ran 2 carver m1.0t bridged . One on each rti. Sounded excellent albeit pretty bright on the tweets. But a easy 600watts rms or so. Pop went a tweeter.
    Ya I'm replacing those cheap 5watt resistors with either 2 parrelled dayton 10 watters or a 20 watt mox . Im going to try a peerless xt25 ringed radiator tweet in my rti's and maybe up the resistor ohmage a bit and see if i cant reduce some of there brightness and increase the tweets power handling some. The tweets are just glued into the funky shaped faceplate/waveguides and peerless and dayton both make a few faceless/flangeless 1" domes that will glue right back into them for 12 to 18$ each rather then polks 70$ each and they pretty much look and spec out to exactly what polk uses for there tweets in these or the monitor series.
  • Tony M
    Tony M Posts: 11,009
    Nice!
    I like how deep you're diving into making these VERY NICE speakers complete again. ;)

    I haven't heard how bright the RTi12's are yet. I thought the highs were were pretty good but a little soft. But all I was hearing were the two midranges and they go up pretty high themselves. Almost enough for me to forget about the tweeters, but I just can't do that.

    The RTi8's were the same way. I thought they were ok. But when I bought the new cross-overs, I heard the upper highs get sizzle and fill the upper spectrum that I can still hear. I bet they're a lot brighter than I'm hearing with my old ears.

    But I do LOVE how these are made, design-wise and quality of cabinetry! <3B)

    I'd also love to have a surround system made out of my sets one day. ;)
    Most people just listen to music and watch movies. I EXPERIENCE them.
  • Tony M wrote: »
    Nice!
    I like how deep you're diving into making these VERY NICE speakers complete again.

    Well I just hope i help them rather then hurt them lol.

    Sadly im running some monitor 70's as rears in my 5.1 system . I'll have to keep my eyes out for some rti10's or rtia7's to put back there .

    Ya im not sure if the brightness is tweeter related or its the upper mids arent filtered off . Part of the reason im asking if someone can make more sense out of that midrange xo design . I do know with the popped tweeter the highs dropped right off the speaker so there must be some sort of bandpass on those mids.

    Wish i had a newer pre/pro or avr with audessy or some other form of speaker and room measuring to help me but im on a budget like most lol.
  • Tony M
    Tony M Posts: 11,009
    You'll help them..I'm pretty sure.

    Looks like you're going about it the right way too and most importantly, you've come here to get advice and knowledge.

    Buy the way,
    WELCOME TO THE FORUM. ;)

    Hope you stick around.
    Most people just listen to music and watch movies. I EXPERIENCE them.
  • Tony M
    Tony M Posts: 11,009
    TTT ?

    I forgot to mention how nice that Carver 600 rms has to be. B)

    It's great that you have a great combo in Polk and Carver going on. I've yet to hook up my Carver TFM35x to my monitor 10's. Soon I hope.
    Most people just listen to music and watch movies. I EXPERIENCE them.
  • tonyb
    tonyb Posts: 32,902
    If I'm not mistaken, those resistors are there to help protect the tweeters. Kinda like the poly switches in the old SDA's. Raising the value might help but uncertain as to what constitutes too much.

    Know what helps more than resistors ? Turning down the volume dial. ;):)
    HT SYSTEM-
    Sony 850c 4k
    Pioneer elite vhx 21
    Sony 4k BRP
    SVS SB-2000
    Polk Sig. 20's
    Polk FX500 surrounds

    Cables-
    Acoustic zen Satori speaker cables
    Acoustic zen Matrix 2 IC's
    Wireworld eclipse 7 ic's
    Audio metallurgy ga-o digital cable

    Kitchen

    Sonos zp90
    Grant Fidelity tube dac
    B&k 1420
    lsi 9's
  • tonyb wrote: »
    If I'm not mistaken, those resistors are there to help protect the tweeters. Kinda like the poly switches in the old SDA's. Raising the value might help but uncertain as to what constitutes too much.

    Know what helps more than resistors ? Turning down the volume dial. ;):)

    Lol im not worried about the resistors , i understand there fuction just fine, yes there wattage ability needs to be raised up certainly in my case tho .
    Im more intrigue and interested in the layout and why of that midrange network .
  • Face
    Face Posts: 14,340
    Textbook crossover schematics are a shot in the dark as the driver's impedance and response need to be factored in when designing a crossover. This is why you can't simply reverse engineer a crossover.

    After measuring a driver's response and impedance without a filter in place, then inserting the files into a program such as this: http://www.speakerdesign.net/WinPCD/index.htm Then a crossover can be designed.

    As far as resistors go, they can affect transfer function as well as pad down a driver.
    "He who fights with monsters should look to it that he himself does not become a monster. And when you gaze long into an abyss the abyss also gazes into you." Friedrich Nietzsche
  • gp4jesus
    gp4jesus Posts: 1,969
    edited February 2018
    I realize the values and frequencies are a bit different but the following MAY provide insight to your mids/tweeter XO relationship.

    Per Polk, CSI A6 CC XO info:

    Twtr, 2.2k hz
    1st mid, 3.8K hz
    2nd mid, 1.5 hz

    I'm sure someone will chime in w/an easy-to-grasp explanation. As for how the mids are, wired, series* or parallel... ...I've pondered that issue & studied the schematic for years. That's said, if that's* true for the RTi12s, then that* MUST apply to the CSi A6 as its wired the same way.

    Schematic for same:

    http://techtalk.parts-express.com/filedata/fetch?id=1191581

    I can't recall if I measured my A6 mids resistance before completing tri-amping it. I'll get back to you.

    I recall seeing somewhere the RTi/RTi A series tweeters are 4 ohm.

    Hope this helps even a little. Tony

    PS 1: if I were you I would
    1. 86 the woofer's* LP and the mids HP filtering
    2. re-wire them* w/12 ga or bigger
    3. re-wire the mids w/ 12 ga
    4. buy an EXO, set for 125hz
    5. upgrade the mid & tweeter passive XO parts

    and revel in SQ improvement, particularly the bass response!

    PS 2: I have 200 watts RMS, on tap, XO'd 1.9Khz just to drive my 'A7 tweets; 160 watts RMS for my 'A6 tweet. Haven't popped one yet.
    Samsung 60" UN60ES6100 LED Outlaw Audio 976 Pre/Pro Samsung BDP, Amazon Firestick, Phillips CD Changer Canare 14 ga - LCR tweeters inside*; Ctr Ch outside BJC 10 ga - LCR mids, inside* & out 8 ga Powerline: LR woofers, inside* & out *soldered LR: Tri-amped RTi A7 w/Rotels. Woofers - 980BX; Tweets & “Plugged*” Mids - 981, connected w/MP Premiere ICs Ctr Ch: Rotel RB981 -> Bi-amped CSi A6 Surrounds: Premiere ICs ->Rotel 981 -> AR 12 ga -> RTi A3. 5 Subs: Sunfire True SW Signature -> LFE & Ctr Ch; 4 Audio Pro Evidence @ the “Corners”. Power Conditioning & Distribution: 4 dedicated 20A feeds; APC H15; 5 Furman Miniport 20s *Xschop's handy work
  • F1nut
    F1nut Posts: 49,708
    Face wrote: »
    Textbook crossover schematics are a shot in the dark as the driver's impedance and response need to be factored in when designing a crossover. This is why you can't simply reverse engineer a crossover.

    After measuring a driver's response and impedance without a filter in place, then inserting the files into a program such as this: http://www.speakerdesign.net/WinPCD/index.htm Then a crossover can be designed.

    As far as resistors go, they can affect transfer function as well as pad down a driver.

    Listen to this guy, he knows exactly what he's talking about.
    Political Correctness'.........defined

    "A doctrine fostered by a delusional, illogical minority and rabidly promoted by an unscrupulous mainstream media, which holds forth the proposition that it is entirely possible to pick up a t-u-r-d by the clean end."


    President of Club Polk

  • F1nut wrote: »
    Face wrote: »
    Textbook crossover schematics are a shot in the dark as the driver's impedance and response need to be factored in when designing a crossover. This is why you can't simply reverse engineer a crossover.

    After measuring a driver's response and impedance without a filter in place, then inserting the files into a program such as this: http://www.speakerdesign.net/WinPCD/index.htm Then a crossover can be designed.

    As far as resistors go, they can affect transfer function as well as pad down a driver.

    Listen to this guy, he knows exactly what he's talking about.

    Lol of course I know this. I never said reengineer I said trying to understand the unusual midrange layout. I actually got some pretty good info over at the diyaudio forum .

    I would need some good measuring tools like a calibrated mic. Arta and a good impedance sweep tester and then create ACCURATE frd and zma files to fully sim the scenerio to make REAL improvements in the xo design .

    F1NUT did you used to be on the carver forum ? I know ive seen your name elsewhere many years ago ?
  • F1nut
    F1nut Posts: 49,708
    Yep, GG and I kept that place going for the first couple of years.
    Political Correctness'.........defined

    "A doctrine fostered by a delusional, illogical minority and rabidly promoted by an unscrupulous mainstream media, which holds forth the proposition that it is entirely possible to pick up a t-u-r-d by the clean end."


    President of Club Polk

  • Lol i thought so . I dont recall my user name there but i was pretty active there for a few years 6 or 7 or 8 years ago . Good to see ya again
  • My carver inventory . 5 or 6 M1.0t's a M1.5t 3 cubes and m200 a TFM35 , a reciever model eludes me atm . And a PT2400 . And a set of original amazings .

    Most of the amps seen time in the hands of todd hartman in Pa.
  • F1nut
    F1nut Posts: 49,708
    Lol i thought so . I dont recall my user name there but i was pretty active there for a few years 6 or 7 or 8 years ago . Good to see ya again

    Well, if you can't remember who you were I hope you don't expect me to ;)

    Anyway, welcome to Club Polk.


    Political Correctness'.........defined

    "A doctrine fostered by a delusional, illogical minority and rabidly promoted by an unscrupulous mainstream media, which holds forth the proposition that it is entirely possible to pick up a t-u-r-d by the clean end."


    President of Club Polk

  • gp4jesus
    gp4jesus Posts: 1,969
    Face wrote: »
    As far as resistors go, they can affect transfer function as well as pad down a driver.
    Sorry. What does this mean?

    Samsung 60" UN60ES6100 LED Outlaw Audio 976 Pre/Pro Samsung BDP, Amazon Firestick, Phillips CD Changer Canare 14 ga - LCR tweeters inside*; Ctr Ch outside BJC 10 ga - LCR mids, inside* & out 8 ga Powerline: LR woofers, inside* & out *soldered LR: Tri-amped RTi A7 w/Rotels. Woofers - 980BX; Tweets & “Plugged*” Mids - 981, connected w/MP Premiere ICs Ctr Ch: Rotel RB981 -> Bi-amped CSi A6 Surrounds: Premiere ICs ->Rotel 981 -> AR 12 ga -> RTi A3. 5 Subs: Sunfire True SW Signature -> LFE & Ctr Ch; 4 Audio Pro Evidence @ the “Corners”. Power Conditioning & Distribution: 4 dedicated 20A feeds; APC H15; 5 Furman Miniport 20s *Xschop's handy work
  • Face
    Face Posts: 14,340
    gp4jesus wrote: »
    Face wrote: »
    As far as resistors go, they can affect transfer function as well as pad down a driver.
    Sorry. What does this mean?
    Slope or shape of response before/after the crossover frequency. Altering it can cause bumps or dips and response and phase issues.
    "He who fights with monsters should look to it that he himself does not become a monster. And when you gaze long into an abyss the abyss also gazes into you." Friedrich Nietzsche
  • gp4jesus
    gp4jesus Posts: 1,969
    edited March 2018
    gp4jesus wrote: »
    As for how the mids are, wired, series* or parallel... ...I've pondered that issue & studied the schematic for years. That's said, if that's* true for the RTi12s, then that* MUST apply to the CSi A6 as its wired the same way.
    The ‘A6 is “nominally” 8 ohms.
    gp4jesus wrote: »
    Schematic for same:

    http://techtalk.parts-express.com/filedata/fetch?id=1191581

    I can't recall if I measured my A6 mids resistance before completing tri-amping it. I'll get back to you.
    Using a CHEAP DMM the miss measure 2.6-3 ohms. Someone smarter than me please chime in. Further, that may explain why, even at very low volume, driving 2 3 ohm & 1 4 ohm load, band limited mind you, my “A6 amp” is always warm to the touch.
    gp4jesus wrote: »
    I recall seeing somewhere the RTi/RTi A series tweeters are 4 ohm.
    I believe the padding “raises” its impedance to 8 ohms.

    My best to ya, sorting through the RTi A12 schematic. Tony
    Samsung 60" UN60ES6100 LED Outlaw Audio 976 Pre/Pro Samsung BDP, Amazon Firestick, Phillips CD Changer Canare 14 ga - LCR tweeters inside*; Ctr Ch outside BJC 10 ga - LCR mids, inside* & out 8 ga Powerline: LR woofers, inside* & out *soldered LR: Tri-amped RTi A7 w/Rotels. Woofers - 980BX; Tweets & “Plugged*” Mids - 981, connected w/MP Premiere ICs Ctr Ch: Rotel RB981 -> Bi-amped CSi A6 Surrounds: Premiere ICs ->Rotel 981 -> AR 12 ga -> RTi A3. 5 Subs: Sunfire True SW Signature -> LFE & Ctr Ch; 4 Audio Pro Evidence @ the “Corners”. Power Conditioning & Distribution: 4 dedicated 20A feeds; APC H15; 5 Furman Miniport 20s *Xschop's handy work
  • K_M
    K_M Posts: 1,627
    edited March 2018

    Maybe some of the more tech savy guys here can help me understand some of this midrange xo , maybe some of it is impedence stablization or something. The 220uf cap followed by a 1.3mh coil in series is really confusing with a Q way out there!


    The 220uf cap and 1.3 mH coil, are a bandpass filter.
    The 220 cap filters out the low end from the mid drivers and the coil, filters off the higher frequencies. Leaving a band of "mid-range" with no bass or treble.

    The caps 10uf and 68uf on mid 1 and mid 2 create a steeper roll off of the high frequencies, more like 12 db/octave.
    Caps "Across" a driver will roll off high frequencies.
    Same as the one on the 3 woofer circuit.

    The mid driver with the 68uf cap, will start rolling off at a lower frequency on the high end, than the one with the 10uf cap.

    The filters polk used will not always match up with some textbook type of slope as the individual drivers response and impedance will vary, and I am quite positive they went from from measure response and tailored the crossover to ideally make that response match their goal.
  • gp4jesus
    gp4jesus Posts: 1,969
    edited March 2018
    gp4jesus wrote: »
    Using a CHEAP DMM the miss measure 2.6-3 ohms. Someone smarter than me please chime in. That* [does] explain why, even at very low volume, driving 2 3 ohm & 1 4 ohm load, band limited mind you, my “A6 amp” is always warm to the touch....
    I’ve since obtained 4-5* ohm readings w/a very good DMM. Add to that* some overlap (2.2khz & 3.5khz) between the tweeter & one mid.

    Tony
    Samsung 60" UN60ES6100 LED Outlaw Audio 976 Pre/Pro Samsung BDP, Amazon Firestick, Phillips CD Changer Canare 14 ga - LCR tweeters inside*; Ctr Ch outside BJC 10 ga - LCR mids, inside* & out 8 ga Powerline: LR woofers, inside* & out *soldered LR: Tri-amped RTi A7 w/Rotels. Woofers - 980BX; Tweets & “Plugged*” Mids - 981, connected w/MP Premiere ICs Ctr Ch: Rotel RB981 -> Bi-amped CSi A6 Surrounds: Premiere ICs ->Rotel 981 -> AR 12 ga -> RTi A3. 5 Subs: Sunfire True SW Signature -> LFE & Ctr Ch; 4 Audio Pro Evidence @ the “Corners”. Power Conditioning & Distribution: 4 dedicated 20A feeds; APC H15; 5 Furman Miniport 20s *Xschop's handy work
  • westmassguy
    westmassguy Posts: 6,850
    Not to hijack the thread, but just repaired a pair of RTi12 crossovers. Both Tweeter resistors were cooked. Both had heat damage near the low pass resistors. One was cooked. All were replaced with Mills MRA-12s
    Home Theater/2 Channel:
    Front: SDA-2ATL forum.polkaudio.com/discussion/143984/my-2as-finally-finished-almost/p1
    Center: Custom Built forum.polkaudio.com/discussion/150760/my-center-channel-project/p1
    Surrounds & Rears: Custom Built forum.polkaudio.com/discussion/151647/my-surround-project/p1
    Sonicaps, Mills, RDO-194s-198s, Dynamat, Hurricane Nuts, Blackhole5
    Pioneer Elite VSX-72TXV, Carver PM-600, SVS PB2-Plus Subwoofer

    dhsspeakerservice.com/