[ST86]

Hi,

I power my Supertowers with a Yamaha DSP A1 home theater amp.  It has EQ for each channel, and while experimenting with the settings I noticed adding 6 dB of boost at around 1 kHz makes things sound much better.  Sort of as if a blanket was removed from the towers. When I mentioned this to Brian he told me there is a dip in the Supertower response at 1 kHz and recommended an upgrade to the 12” midbass Misco driver, a midrange capacitor change from 50 uF to 54 uF, and a woofer inductor change from 1 mH to 2.8 mH.  This was to change the woofer to mid crossover from the original 650 Hz to  450 Hz and eliminate the response dip.  The upgrade seems to have helped but not completely, I still need to add 3 db of boost at 1 kHz. 

Checking into this further I ran some numbers on some online crossover calculators using first order Butterworth.  Using 8 ohms for the midrange driver impedance (2X four ohm drivers in series) works out to very close to 54 uF.  However I am having trouble verifying the 2.8 mH inductor value for the woofers (one 8 ohm 15” and one 8 ohm 12” wired in parallel making 4 ohms).  When I use 4 ohms and 650 Hz I get 0.98 mH. When I use 4 ohms and 450 Hz I end up with 1.4 mH.  The 2.8 mH value works when I use 8 ohms.  Seems Brian may have calculated the new woofer inductor using 8 ohms for the woofers instead of the 4 ohms.  By the way, using 4 ohms and 2.8 mH I get a crossover point of 225 Hz.

Question: Please check my math. Do I have the wrong inductor in my cabinets?  For a 450 Hz crossover point using one inductor with two 8 ohm woofers in parallel, what should be the value?  I know it depends on the driver characteristics, but should my 2.8 mH really be 1.4 mH?  Would a Supertower owner be nice enough to check what he/she currently has?  Extra points if using the Misco 12” driver.

Also, the woofer inductor is connected in series between the negative terminals of the drivers and ground. On the wiring schematic there is a wire from the midrange driver side of the 54uF midrange crossover capacitor to the junction of the negative woofer terminals/2.8 mH inductor.  I don’t understand the purpose of this wire. 

Question:  If any Supertower owners could check to see if this wire is present either on the schematic if you have it or in the cabinet I would appreciate it.  If anyone can decipher what this wire does that would also be appreciated.  I can post the schematic here if there is interest.

Thank you,

Ed

[Lee_in_Canada]

Hi Ed,

I have a pair of supertowers R's.  Original drivers.  I'd be happy to pull out a driver and photograph wires.  However I really know very little about what I'm looking however will look for the 54uF midrange crossover capacitor and it's extra wire connected to the midwoofer.

I am following the response to your request with interest as the misco drivers need to be ordered in quantities of 1000 and the 5" mids are OEM also.   Leaving me wondering about getting the utmost upgraded midrange performance out of these beautiful old speakers.

Lee

[ST86]

Lee,

Thank you.  If your wiring is setup like mine you would see a brown wire from the midrange capacitor to the woofer inductor.  Take note of the cap and inductor values.  The capacitor and inductor may be glued to one of the terminal cups.  As for the Misco drivers, the min order I could get one of the customer service people down to was 250 units. They did not have any units in stock when I contacted them, and I don't know of any place where they can be purchased.

Cheers,

Ed

[ST86]

I wanted to see if my calculations were correct that for a 450 Hz crossover with the Misco 12” driver 1.4 mH is the “proper” inductor value and would make the midrange sound better than the 2.8 mH provided by Brian.  I bought an L-C meter from Parts Express and unwound the 2.8 mH inductors until they were 1.4 mH.  Been listening for a few days now and I like what I hear.  I barely need to add any EQ around 1 KHz and when I add 3 dB as I did with the 2.8 mH it now sounds like too much.

Being very hesitant to second guess anything that Brian did it seems he calculated the 2.8 mH inductor based on 8 ohm woofer impedance instead of the actual 4 ohms of the 15” and 12” in parallel.  Opinions welcome.

I also would like to repeat my request to Supertower owners.  Would someone be kind enough to unscrew the upper terminal cup and report the values of the crossover cap and inductor?  Please also make note of the flavor of 12” midbass as this will affect the cap and inductor values.  Recall the cap (likely a bundle of them, just add all the values) goes from the positive speaker terminal to the midrange drivers.  The inductor is wired between the negative terminals of the woofers and the negative speaker terminal.

I will post the schematic when I finally take some pictures. 

Ed

[Speedskater]

Remember that a loudspeaker driver's impedance will vary with frequency, so you can't just plug 4 or 8 Ohms into the calculator and hope for the best.  Also a good cross-over design accounts for acoustic roll-off of the woofer as the frequency rises.

[ST86]

Yes, thank you. I am aware of that.  I do not have the impedance vs frequency plot for the 15" woofer but for the Misco 12" it is around 9 ohms at 450 Hz.  The 4 ohms that I used is an aproximation but as of now it appears to be pretty close.

Ed

[James Romeyn]

Every ST/R extant (of which I'm aware, I worked for Brian starting not long after this model arrived through to its demise in the ST/R SE symmetrical array) employs strange, weird, and exotic QSO (quasi-second order) crossover between bass and mids.  QSO is hybrid series/parallel. 

Every non-QSO formula is completely and utterly useless and can not be modified to translate to QSO specs, a complete waste of time. 

Because of that, I didn't read much of the OP.

What you might consider is inputting test signals and checking for V resonance at the outputs.  I can't remember the details, but believe this is the easiest and most direct method to find system xo points (V resonates at each pole IIRC).  Post for details at diyaudio or here in the lab. 

[ST86]

James,

Thanks for the reply.  Does this mean the 1.4 mH inductor that I calculated for a crossover point of 450 Hz is the wrong value?  Would it help if I posted the schematic here?   

I'm not sure you mean resonance, do you mean half-power point?

One thing I considered doing to verify the 3 DB crossover points is use a function generator and AC voltmeter.  Input a test signal from the generator through the amp in the passband so there is maybe 100 mV on both the amp side and driver side of crossover components, then change the frequency until (I hope I have this right) there's 100 mV on amp side of crossover and 70 mV on driver side.  Generator frequency should then correspond to -3 DB crossover point.  I have a sinewave generator and can borrow a Keithly meter that measures frequency and AC amplitude at the same time.  Now I just need to find the time to run the experiment.

Cheers,

Ed

[James Romeyn]

Acoustic xo pole = the electrical pole + the driver's interaction with it. In the of QSO both the bass and mid interact with each other's electrical xo (the bass has more effect on the mid xo than vice versa).  IIRC, in Brian's QSO on the ST/R the bass is 1st order but the mid is 2nd order (both electrical, no idea acoustic).   

IMO one can presume nothing by referencing any single xo component.  This rule applies more so with QSO than non-QSO xo. 

The rest is above my pay grade.  Wish I knew more.  I can draw a generic QSO xo but I know little beyond that.  Guys other than Brian who used QSO are/were Rick Reimer, Fritzpeakers, and Bud Fried.           

[James Romeyn]

Hi,

I power my Supertowers with a Yamaha DSP A1 home theater amp.  It has EQ for each channel, and while experimenting with the settings I noticed adding 6 dB of boost at around 1 kHz makes things sound much better.  Sort of as if a blanket was removed from the towers.

6dB is a huge bump.  What is octave magnitude?  As someone who owned ST/R, built countless, and heard them hundreds of times, I suspect something is amiss. 

When I mentioned this to Brian he told me there is a dip in the Supertower response at 1 kHz and recommended an upgrade to the 12” midbass Misco driver, a midrange capacitor change from 50 uF to 54 uF, and a woofer inductor change from 1 mH to 2.8 mH.  This was to change the woofer to mid crossover from the original 650 Hz to  450 Hz and eliminate the response dip.  The upgrade seems to have helped but not completely, I still need to add 3 db of boost at 1 kHz. 

Brian's original 2710 12" mid bass had way too much ringing mid energy for which polypropylene of that era is well known...me and at least one other person mentioned this several times to Brian.  If everything so far is accurate, I can see a way that the 2710 could overlap with the dual 5" mids @ 1kHz, the result being floor bounce and a suckout at 1kHz caused by disparate pathlengths driver to floor to listener. 

Checking into this further I ran some numbers on some online crossover calculators

With Brian's ST/R xo, the bass absolutely affects the mid xo capacitor.  I'm less sure if the mids affect the bass xo.  Is there online QSO crossover calculator? 

using first order Butterworth.  Using 8 ohms for the midrange driver impedance (2X four ohm drivers in series) works out to very close to 54 uF.  However I am having trouble verifying the 2.8 mH inductor value for the woofers (one 8 ohm 15” and one 8 ohm 12” wired in parallel making 4 ohms).

On every 2710 equipped ST/R: 8 Ohm 2710 has resistor (about 12 Ohms) wired in series for about 20 Ohms, parallel with 8 Ohm 15" = overall bass R about 6 Ohms.   

Yes or no: is the Misco 12" wired the same or different?  Resistor value?   

When I use 4 ohms and 650 Hz I get 0.98 mH. When I use 4 ohms and 450 Hz I end up with 1.4 mH.  The 2.8 mH value works when I use 8 ohms.  Seems Brian may have calculated the new woofer inductor using 8 ohms for the woofers instead of the 4 ohms.  By the way, using 4 ohms and 2.8 mH I get a crossover point of 225 Hz.

Question: Please check my math. Do I have the wrong inductor in my cabinets?  For a 450 Hz crossover point using one inductor with two 8 ohm woofers in parallel, what should be the value?  I know it depends on the driver characteristics, but should my 2.8 mH really be 1.4 mH?  Would a Supertower owner be nice enough to check what he/she currently has?  Extra points if using the Misco 12” driver.

Also, the woofer inductor is connected in series between the negative terminals of the drivers and ground. On the wiring schematic there is a wire from the midrange driver side of the 54uF midrange crossover capacitor to the junction of the negative woofer terminals/2.8 mH inductor.  I don’t understand the purpose of this wire. 

Question:  If any Supertower owners could check to see if this wire is present either on the schematic if you have it or in the cabinet I would appreciate it.  If anyone can decipher what this wire does that would also be appreciated.  I can post the schematic here if there is interest.

Thank you,

Ed

Ed,
QSO you describe is rather complicated.  I would stop trying to double guess Brian.  Also, re. xo values, the only speakers you can compare to yours are the exact same original vintage with exact same upgrades.  Brian had a little hand written notebook in which he scribbled upgrades and changes over time.  That notebook being gone, I'd presume nothing regarding values for similar model with any slightly different evolution.

[ST86]

My apologies if I double post.  Was almost done typing and hit some keys on the keyboard that seemed to obliterate the whole thing.
------------------

6dB is a huge bump.  What is octave magnitude?  As someone who owned ST/R, built countless, and heard them hundreds of times, I suspect something is amiss.

Any suggestions on what to check?  The wiring and driver polarity checks out (as far as I can tell) per schematic and build instructions.

Brian's original 2710 12" mid bass had way too much ringing mid energy for which polypropylene of that era is well known...me and at least one other person mentioned this several times to Brian.  If everything so far is accurate, I can see a way that the 2710 could overlap with the dual 5" mids @ 1kHz, the result being floor bounce and a suckout at 1kHz caused by disparate pathlengths driver to floor to listener

There certainly is a lot of midrange energy present in the 12" mid bass.  I suppose lowering the crossover from 650 Hz to 450 Hz would actually increase it but the change made things sound better.

With Brian's ST/R xo, the bass absolutely affects the mid xo capacitor.  I'm less sure if the mids affect the bass xo.  Is there online QSO crossover calculator?

Been looking, so far can't find one.

On every 2710 equipped ST/R: 8 Ohm 2710 has resistor (about 12 Ohms) wired in series for about 20 Ohms, parallel with 8 Ohm 15" = overall bass R about 6 Ohms.   

Yes or no: is the Misco 12" wired the same or different?  Resistor value?

The series resistor was connected to the positive terminal on the original poly 12" driver.  25 ohms.  When I upgraded the 5" mids Brian told me the resistor was there to balance the acoustic level of the 12" to the mids.  Because I upgraded the mids he said I no longer needed the resistor, so I removed it.  The Misco driver that I now use does not have the resistor.

Ed,
QSO you describe is rather complicated.  I would stop trying to double guess Brian.  Also, re. xo values, the only speakers you can compare to yours are the exact same original vintage with exact same upgrades.  Brian had a little hand written notebook in which he scribbled upgrades and changes over time.  That notebook being gone, I'd presume nothing regarding values for similar model with any slightly different evolution.

As stated in another post in this thread I am very hesitant to second guess anything Brian did.  Just trying to understand how the speaker functions.

I would like to post my Supertower schematic.  Anyone have any concerns with me putting the speaker schematic on a public forum?

Thanks,

Ed

[James Romeyn]

Lowering the bass LP pole from 650 to 450 Hz minimizes midrange energy from the bass system.

It's fine to post the xo, nothing proprietary, no license concerns at all, Brian made it public himself for each kit customer who received a copy.

Bass/mids: series wired QSO, IIRC 1st order LP bass, 2nd order HP mids...no LP pole as per every single VMPS mid extant till finally, just short of physical threats of violence (OK, I exaggerate) Brian finally reneged, adding LP pole to all Neo planar panels (circa 6kHz).

Biwire/passive biamp split is between mids and domes circa 4-5kHz, absolutely mandating only exact same amps for any biamp application.  IMO this negates most of the biamp raison d'être being super high current/high voltage for the bass and super fine resolution for the mid/treble, ideally split in the 80-100 Hz range.   

Domes: same as every VMPS dome extant, HP 1st order (circa 4kHz), padded down through the L-pad, no LP pole.

Supertreble: planar parallel with piezo (same as TII with planar upgrade, standard TII employed dual piezo w/ parallel resistor, TII vertical array SE planar only/no top firing piezo, MT and MTII employed single piezo with parallel resistor or upgrade to single planar)...same as every VMPS supertreble circuit extant, HP 2nd order circa 10kHz, .2mH parallel coil, poles selected by different capacitor values.   

That's my story and I'm sticking to it.

Brian's best overall values in radom order:
TII all models with arcing mid-treble array, I dislike all vertical array models
ST/R original, ST/R SE symmetrical array
STIIa/R
STIII pretty much every version
RM2
RM40
RM30
RM50
Wings RM60 nice but kinda rickety mechanical couplings

Avoid:
All MT
RM1 OK but target RM2 if space allows, humungous upgrade

The old 808 12" 3-way kicks butt, but I'd consider adding a small LP pole to the mid.

I want twelve good mechanical condition VMPS "X5109" midrange, 4 Ohm only, steel basket, poly cone, rubber surround.

Avoid FF series (just too much upper mid/low treble energy) except upgrades to full ribbons or planar models, for full BG ribbon watch out for ribbon buzz, and in all cases of FF with passive bass xo bypass the passive and go active, even mono to both channels is likely OK or consider two plate amps for bass. 

[ST86]

Yes, my mistake.  Lowering the crossover point decreases the midrange energy coming through the 12" driver.

Here's the schematic:




James, if this was a recall test you'd get an "A". 

The biggest question I have right now on the wiring is the wire labled "brown" that goes from pin 3 of the midrange L-pad to the junction of the negative terminals of the woofers and the inductor.  Can't figure out what this wire does.

You'll also notice Brian crossed out 49uF for the midrange series cap and wrote in 52.5 which is what I installed when I built these.  The cap bundle changed with the 12" upgrade to the Misco driver and crossover change.  Same with the woofer inductor, started as 1.0 mH then changed to 2.8 mH with the upgrade, then I changed them to 1.4 mH.

All the other crossover values have not changed, just been updated with new components.  The softdomes were replaced with new Morel MDT 29's , they originally came with MDT 29's and I was happy with the performance but wanted to update them.

The piezo is original, as is the planer supertweeter.  The 15" woofer was replaced aprox. ten years ago, the passives are a couple years old.

Ed

[James Romeyn]

Wow, strange to see the top firing piezo in series with the domes.  I would have bet the piezo shared the planar pole.  Interesting.  I wonder if the TII shared this feature. 

Actually, it makes more sense for the "ambiance" tweeter to cover the extra 1+ octave of the dome pole rather than only the top most octave of the planar.

Those two terminal cups (later boards) had wires hanging everywhere like an octopus.  As complex as that xo was, Brian could tell jokes, make fun of people not in the room, talk on the phone, hang up on telemarketers, while making them.  They seldom needed attention later except for the occasional little fix.

I always tried to talk people out of the kits and just buying them assembled, but the occasional buyer really wanted to tackle it.  I can only imagine it took days for them to build a xo Brian assembled from start to finish in about 1.5 hours.

A story I hope to never forget, concerning this same ST/R speaker: A guy calls, said he bought his ST/R kit years prior, and asks for new kit instructions which were lost.  Brian happily sent him instructions.  Then he called back and requested advice on how to evict his cats from the empty cabinets, where they had taken up residence a long time ago.  First I learned of feline squatter's rights. 

Brian actually operated a thriving business for years, a large commercial space filled with speakers, cabinets, and parts.  Generally, housekeeping was borderline chaos.  Brian would tell me to go unbox "X" "over there."  I'd look for a few minutes, tell him I couldn't find it, ask him again where it was, and he'd reply, "Over there by the boxes" in a warehouse space filled with cartons 6' high and more.     

[ST86]

James,

Thanks for the story.  I vaguely remember reading somewhere about Supertower cabinets that were turned into cat houses.

So, just to try to wrap up the schematic questions any idea what that brown wire between the mid crossover cap and the woofer coil does?  Anyone?  As far as I can tell it is part of the mid crossover until the woofer coil impedance increases to where it becomes an open circuit. Seems to me it might syphon off some of the lower midrange but again I don't completely understand how it works.

I will recheck the wiring, again, to see if there is anything I missed that I built wrong that could be causing the midrange dullness.  If anyone has any other suggestions to check or try I will be glad to hear them.

Cheers!

Ed

[baco99]

Hi!  Thanks Ed for forwarding me this thread. 
I am working on some SuerTower/R that sound terrible.

My main question is, the manual states that the 15 and 12" woofers are crossover at different frequencies, but there's only one 1.0 mH inductor used as a choke for both drivers.  If the drivers were different impedances, then I could see, in theory how running them in parallel might get to the prescribed cutoff frequencies of 250 and 600 Hz, respectively.  Running them in parallel would halve the DCR, which I suppose would halve the Impedance as well.  Then a 1mH inductor across both drivers would get to the desired cut offs. 

However, both the 15" and 12" drivers are 8 ohm versions on mine, so the effective cutoff frequency is 600 Hz for BOTH drivers.  This means there's a terrible amount of overlap between the two, which is why the bass is so terrible and muddy. 

Also, with the 12" and 15" drivers sharing the same rear volume, the 15" driver would inflict unwanted modulation on the 12" driver (and vice versa). 

 Is this a completely terrible design that needs an overhaul, or am I interpreting the design wrong?


What I feel needs to be done is the woofers need be driven separately, with the 12" mid-bass coupler in its own sub enclosure (I might suggest a Sonotube stuffed with fiberglass fill).  Then a 2.0mH (600 Hz) choke on the mid-bass coupler (and let the bottom end roll off naturally in the smaller sealed sub-enclosure). Then the 15" driver gets a 6.8 mH choke, probably an iron core for a true 200 Hz cutoff.  Thoughts?


Thanks!

[baco99]

Actually, I may have a simpler question, is that a 15 ohm pad in front of the 12" mid-bass coupler in the schematic?  I don't see that component here.

[ST86]

I think what you are referring to is the 25 ohm series resistor between the positive speaker terminal and the positive terminal of the 12" driver.  Brian told me that resister was there to equalize the acoustic level of the original 12" poly driver with the original 15" and the mids. When I upgraded the 15" and the 12" the resistor was not needed, and the inductor value changed from 1.0mH to 2.8mH.

If you draw out the woofer and mids crossover you'll see it is a series crossover.  I have gone through a fair amount of heartburn trying to understand this series crossover.  I still don't have a clear understanding of its operation because with this configuration you can't change one component without it affecting everything else, so my suggestion is to keep that in mind if you decide to alter the woofer config.

Ed

[baco99]

I have the original "ribbed" poly cone 12" mid-bass driver, so lacking this resistor might be the issue.  I might throw a rheostat in there and see if that changes things.  That might be the magic sauce. 

I still don't understand how a single 1.0mH inductor can provide a cutoff frequency for the 15" at 200 and the 12" at 600.  Does not compute.

I also still don't understand how to avoid unwanted modulation when both drivers share the same rear chamber.

[James Romeyn]

I have the original "ribbed" poly cone 12" mid-bass driver, so lacking this resistor might be the issue.  I might throw a rheostat in there and see if that changes things.  That might be the magic sauce.

There's only one VMPS ribbed poly 12", being the Gefco 2710, Brian's own design.  Every VMPS 2710 equipped speaker included a series power resistor to the 2710.  This being the case, Brian did not design the speaker described above.

I still don't understand how a single 1.0mH inductor can provide a cutoff frequency for the 15" at 200 and the 12" at 600.  Does not compute.

Brian did not design the speaker you describe.  In all Brian's designs w/the 2710, regardless the fact that the two circuits share one 6dB electrical pole, the two circuits have completely different T/S parameters, hence different electrical and acoustic xo poles.  (The most obvious difference is 8 Ohm 15" and 33 Ohm 2710 12" w/series R.) 

Brian's quoted specs are approximations.  With a VMPS design (yours are not): Separately attach a VM to each circuit/driver.  Sweep a tone generator input signal.  The frequency where V resonates indicates the xo pole for that circuit.   

It seems like you're looking for help in the VMPS forum w/ a non-stock non-VMPS designed circuit.  I hope I'm wrong, but I doubt anyone here can help you w/that project. 

Re. the two drivers sharing one enclosure, I suspect: in the original ST/R, the 15" performance would barely change whether the 2710 was installed per the original design or no 2710 and the empty hole plugged (the 15" would probably cut off a little lower).  This is because the two drivers share so little similar range, and the 12" is significantly padded down.  I've heard the ST/R next to Large Subs w/ the same 15" and 12" low bass drivers.  In the bass range the two sound similar, the LS being more powerful and w/lower cutoff.   

Conversely, if the 2710 was installed in a smaller enclosure, it would ring even worse than it already does, and be almost un-listenable.  You can already hear high mid range from the 2710, and this would only worsen in a smaller enclosure, which would almost surely need to be separately reflex loaded.