I've been working with a pair of vintage Acoustic 807 cabs and I've run into an issue. When I bought them, one of the Heppner 4"x10" midrange horns was nonfunctional and efforts to replace its diaphragm/voice coil assembly were unsuccessful. There have been only two models of new-production integrated 4"x10" horn loudspeakers available in the United States (regardless of how they're branded - brands include Peavey and Pyle) and I have bought a pair of each to try as replacements. I wound up using the metal-horn model. In order to get the behavior of the two cabs close to the same (the crossover connections are undocumented and there was no telling which way of connecting them was what Acoustic intended), I bought two Dayton Audio units with crossover points at 500 and 4000Hz and installed one. As a casual test, I was blowing sine waves through the cab and when I had the frequency set to 1000Hz - which the crossover should be sending predominantly to the midrange horn - and I made a disturbing discovery. Above a certain amplitude - still a very low power level - I could hear a set of harmonics start up and I confirmed what I was hearing with a measurement mic and the REW app. The appearance of upper harmonics wasn't gradual; turn up the gain when they aren't there and at a certain threshold they switch on. When I swapped the horn out with the one from the other cab the threshold wasn't there - that is, until I grabbed the back end of the horn and wrenched it around with my bare hands. Then I tried the other pair of horns that I'd bought - same thing. Then I tried the cabs' surviving original Heppner horn - same thing.
I am left to conclude that horns of this sort are hot garbage. Lest you think it was something other than the horns themselves, I tested out all five of the horns I had on hand with the power amp connected directly to them, and I didn't get the harmonics when connecting the power amp to a JBL L56 cabinet or cheap piezo tweeters (albeit at like 5000-8000Hz). If these horns are acting like kazoos with a sinewave signal, they've got to be making absolute hash out of program and I can't abide it.
In looking for combinations of individual compression drivers and horn lenses, I've been unable to find anything that's suitable yet will also fit into the cabs' recessed mounting place, so I've decided to just screw a blank panel over the hole and set a horn/driver combination atop the cab, either bare (Voice-Of-The-Theater-style) or in separate enclosures that will rest atop the 807 cabs. This frees me up to use any of a pretty wide range of horn lenses, and I've identified a driver that seems workable in terms of power handling and frequency range (this one is rated 75W RMS and 400-8000Hz). So here is my question: given that the original horn was 4"x10" and recessed into a slot with panels to either side that sit diagonally in plan, front to back (thereby kinda-sorta extending the horizontal horn profile to nearly the width of the cab), what do you think it would do to use a horn lens the size of, say, 5-1/8" x 15-3/16" instead? The only thing I'm worried about is that if the dispersion is much wider or narrower than that of the splayed piezo tweeters mounted underneath that slot or the two vertically-arranged 12" speakers, then the off-axis performance might be mid-high or mid-low. BTW, mostly these cabs will be used for playing keyboards through at significant volumes, widely spaced, and set well back from the audience so getting even dispersion between woofers, mid, and tweeters is meaningful.
I am left to conclude that horns of this sort are hot garbage. Lest you think it was something other than the horns themselves, I tested out all five of the horns I had on hand with the power amp connected directly to them, and I didn't get the harmonics when connecting the power amp to a JBL L56 cabinet or cheap piezo tweeters (albeit at like 5000-8000Hz). If these horns are acting like kazoos with a sinewave signal, they've got to be making absolute hash out of program and I can't abide it.
In looking for combinations of individual compression drivers and horn lenses, I've been unable to find anything that's suitable yet will also fit into the cabs' recessed mounting place, so I've decided to just screw a blank panel over the hole and set a horn/driver combination atop the cab, either bare (Voice-Of-The-Theater-style) or in separate enclosures that will rest atop the 807 cabs. This frees me up to use any of a pretty wide range of horn lenses, and I've identified a driver that seems workable in terms of power handling and frequency range (this one is rated 75W RMS and 400-8000Hz). So here is my question: given that the original horn was 4"x10" and recessed into a slot with panels to either side that sit diagonally in plan, front to back (thereby kinda-sorta extending the horizontal horn profile to nearly the width of the cab), what do you think it would do to use a horn lens the size of, say, 5-1/8" x 15-3/16" instead? The only thing I'm worried about is that if the dispersion is much wider or narrower than that of the splayed piezo tweeters mounted underneath that slot or the two vertically-arranged 12" speakers, then the off-axis performance might be mid-high or mid-low. BTW, mostly these cabs will be used for playing keyboards through at significant volumes, widely spaced, and set well back from the audience so getting even dispersion between woofers, mid, and tweeters is meaningful.
Thought I'd resurrect my dead post instead of starting a new thread. Will start with a TL;DR so that the OP can been safely skipped.
These are vintage 3-way cabs originally with a pair of 12" drivers in a reflex horn-of-sorts sub-enclosure, a midrange horn with integrated magnet and coil, and a pair of piezo tweeters mounted underneath the horn. I've replaced the 12"s with some that have much more Xmax, I've replaced the piezo tweeters to get four alike because there was one of the four giving low output...and the midrange has been a problem. When I bought these cabs, one of the original mid horns was blown and trying to replace the diaphragm/coil assembly did not work out and all my attempts to replace them with current production units have failed; they easily distort like kazoos. So I decided to take a step that I was keeping in the back of my mind but really didn't want to do: forego the internally-mounted mid horns and instead set separate horn lens and compression driver assemblies on top of the cabs Voice-Of-The-Theater-style. I have replaced the original crossovers with Dayton Audio 3-way units that cross at 500 and 4000Hz and connected their mid outputs to jacks on the back.
I plan to make proper mounting brackets but for the time being, I'm just gravity-mounting the horns atop the cabs like so:
The challenge, though, had been finding compression drivers that can go that low but I succeeded. As a side note, I had a problem in that the two compression drivers did not sound the same with white noise run through them and I confirmed this using REW. I have since ordered a pair of replacements that are revisions to the ones that I've been working with and hopefully they will be more consistent. I can say, though, that even with the mismatch and after having used REW to set the levels on a 1/3-octave equalizer before the power amp, these things sound really good at far-field distances.
Here is my question: I want to build mounting brackets to more gracefully hold the horns atop the cabs which leads me to wonder: how far forward/backward should the horns sit? My idea for determining this is to run a sinewave into the cab at 500Hz (thereby splitting the signal between the horn and the 12"s), sitting the horn some distance away on top of a stool facing the other direction, putting a pair of microphones about three feet in front of each, and then amplifying each mic signal, sending them into an oscilloscope set to XY mode, and moving one mic backward or forward until the trace forms a diagonal line instead of an oval. The difference in mic distances, then, should correspond to how far forward or backward the mouth of the horn should sit relative to the rest of the cabinet. Does this seem sound (so to speak) to you?
Follow-up question: the wavelength of 500Hz in air is 27 inches but at 4000Hz (the high crossover point) it's about 3-3/8 inches. I could perform this experiment between the horn and the piezo tweeters instead of the horn and the 12"s; do you think I should do one versus the other or do both and split the difference if any? I should point out that with the horn moved to align with the front of the cabinet, the piezo diaphragms are more than a 4000-Hz wavelength ahead of the compression driver diaphragm. I should also point out that Acoustic Control Corporation apparently wasn't too bothered about having the diaphragm of the original horn sitting well back from where the tweeters were.
These are vintage 3-way cabs originally with a pair of 12" drivers in a reflex horn-of-sorts sub-enclosure, a midrange horn with integrated magnet and coil, and a pair of piezo tweeters mounted underneath the horn. I've replaced the 12"s with some that have much more Xmax, I've replaced the piezo tweeters to get four alike because there was one of the four giving low output...and the midrange has been a problem. When I bought these cabs, one of the original mid horns was blown and trying to replace the diaphragm/coil assembly did not work out and all my attempts to replace them with current production units have failed; they easily distort like kazoos. So I decided to take a step that I was keeping in the back of my mind but really didn't want to do: forego the internally-mounted mid horns and instead set separate horn lens and compression driver assemblies on top of the cabs Voice-Of-The-Theater-style. I have replaced the original crossovers with Dayton Audio 3-way units that cross at 500 and 4000Hz and connected their mid outputs to jacks on the back.
I plan to make proper mounting brackets but for the time being, I'm just gravity-mounting the horns atop the cabs like so:
The challenge, though, had been finding compression drivers that can go that low but I succeeded. As a side note, I had a problem in that the two compression drivers did not sound the same with white noise run through them and I confirmed this using REW. I have since ordered a pair of replacements that are revisions to the ones that I've been working with and hopefully they will be more consistent. I can say, though, that even with the mismatch and after having used REW to set the levels on a 1/3-octave equalizer before the power amp, these things sound really good at far-field distances.
Here is my question: I want to build mounting brackets to more gracefully hold the horns atop the cabs which leads me to wonder: how far forward/backward should the horns sit? My idea for determining this is to run a sinewave into the cab at 500Hz (thereby splitting the signal between the horn and the 12"s), sitting the horn some distance away on top of a stool facing the other direction, putting a pair of microphones about three feet in front of each, and then amplifying each mic signal, sending them into an oscilloscope set to XY mode, and moving one mic backward or forward until the trace forms a diagonal line instead of an oval. The difference in mic distances, then, should correspond to how far forward or backward the mouth of the horn should sit relative to the rest of the cabinet. Does this seem sound (so to speak) to you?
Follow-up question: the wavelength of 500Hz in air is 27 inches but at 4000Hz (the high crossover point) it's about 3-3/8 inches. I could perform this experiment between the horn and the piezo tweeters instead of the horn and the 12"s; do you think I should do one versus the other or do both and split the difference if any? I should point out that with the horn moved to align with the front of the cabinet, the piezo diaphragms are more than a 4000-Hz wavelength ahead of the compression driver diaphragm. I should also point out that Acoustic Control Corporation apparently wasn't too bothered about having the diaphragm of the original horn sitting well back from where the tweeters were.
What compression driver do you have there now? The vast majority of 1" exit CDs really aren't suitable for use in PA applications at much below 2khz... maybe 1.5khz at most, that is most likely why you were unsuccessful with getting a satisfactory result up to this point The Timpano TPT-D250x is a dedicated midrange driver that would be more suitable here as is the Selenium/JBL D250x.... maybe the Timpano is a copy of this? They both look very similar to the driver in your picture.
In PA applications these days a speaker that crosses to a compression driver at 500hz would most often use a much larger and more expensive 2" exit CD.. something with a 3" to 4" voice coil, and it would be actively processed and discretely powered. I have some of those big drivers and I still don't cross them below 800-1khz. I lived through the era of cabs like the one you have and can report that they often were not very pleasant to stand in front of, the midrange squawk was painful at times. The problem was any or all of the following.. inadequate( non existant) processing, a crossover that was too low and too shallow for the drive levels, a poorly designed or poor quality horn lens, or perhaps a poor driver-horn junction.. this transition should be smooth.
Do you have an L-pad on the CD to knock it's output level down to compliment the woofers? With around 107dB sensitivity you should have 7-10dB of pad on them. I suspect you would also get much better results with the crossover frequencies moved up a bit, that is easy to do with an activer crossover of course so you might consider going down the route if you can't get the passive crossover dialed in, I also had a lot of experience with piezos when they first appeared back in the '80's and this model is very much a super tweeter which works best when it's just adding some topend sparkle, you don't want to overdrive them as the output turns to mush.. we used to call it the frying bacon effect.
Aligning all the voice coils is a good starting point but you may find ideal acoustic summing is a little off of that. REW may be able to help with that if it can measure impulse reponse. You may be able to do it with an oscilloscope too though, a single mic will pickup 2 distinct impulse peaks when the drivers in question are out of time with one another and only 1 when they are aligned.
In PA applications these days a speaker that crosses to a compression driver at 500hz would most often use a much larger and more expensive 2" exit CD.. something with a 3" to 4" voice coil, and it would be actively processed and discretely powered. I have some of those big drivers and I still don't cross them below 800-1khz. I lived through the era of cabs like the one you have and can report that they often were not very pleasant to stand in front of, the midrange squawk was painful at times. The problem was any or all of the following.. inadequate( non existant) processing, a crossover that was too low and too shallow for the drive levels, a poorly designed or poor quality horn lens, or perhaps a poor driver-horn junction.. this transition should be smooth.
Do you have an L-pad on the CD to knock it's output level down to compliment the woofers? With around 107dB sensitivity you should have 7-10dB of pad on them. I suspect you would also get much better results with the crossover frequencies moved up a bit, that is easy to do with an activer crossover of course so you might consider going down the route if you can't get the passive crossover dialed in, I also had a lot of experience with piezos when they first appeared back in the '80's and this model is very much a super tweeter which works best when it's just adding some topend sparkle, you don't want to overdrive them as the output turns to mush.. we used to call it the frying bacon effect.
Aligning all the voice coils is a good starting point but you may find ideal acoustic summing is a little off of that. REW may be able to help with that if it can measure impulse reponse. You may be able to do it with an oscilloscope too though, a single mic will pickup 2 distinct impulse peaks when the drivers in question are out of time with one another and only 1 when they are aligned.
Last edited:
The Timpano is exactly what I'm dealing with. Apparently the same product is sold under a variety of names and model numbers, much as is the case with one-piece horn units. As it happens, I do have a pair of L-pads that I can put on the drivers but I'd rather handle that in the EQ unless I just can't get the sliders down low enough.
An L-pad is the recommended method for level matching with a passive crossover, hugh cuts on a graphic EQ will create a very lumpy response and generate a lot of unnecessary phase shift. And the added benefit of using an L-pad is it can be used to help flatten the highend response of the driver/horn combo by simply adding a certain value capactior across the series resistor.
I appreciate the advice - thank you. I have the option of integrating the L-pads into the brackets I'll build to hold up the horns. Alternatively I could include them in the cabs and have the knob either pointing downwards near the piezos or even out of the blank panels I've fabricated to go where the old horns go but really, I've fooled around inside those cabs as much as I feel like.
I've been resisting (heh, heh) using the L-pads because of an aversion to turning amplifier power into heat any more than necessary but yes, those horns' output is definitely elevated relative to the rest of the cabs'. But what was surprising to me was that even with the replacement one-piece horns installed, the cabs are way more efficient than my two-way JBL L56es. I figured driving them with the 215W/channel Crown XLS1000 amp would be plenty, but I have since picked up a Hafler P125 at 60W/channel that probably would drive the cabs as loudly as I'd ever need. The Crown isn't bad - and I really selected it from several amps at a used shop because it was crazy light - but the Hafler is I think overall the better amp.
I've been resisting (heh, heh) using the L-pads because of an aversion to turning amplifier power into heat any more than necessary but yes, those horns' output is definitely elevated relative to the rest of the cabs'. But what was surprising to me was that even with the replacement one-piece horns installed, the cabs are way more efficient than my two-way JBL L56es. I figured driving them with the 215W/channel Crown XLS1000 amp would be plenty, but I have since picked up a Hafler P125 at 60W/channel that probably would drive the cabs as loudly as I'd ever need. The Crown isn't bad - and I really selected it from several amps at a used shop because it was crazy light - but the Hafler is I think overall the better amp.
Just 2 cents worth ...
Those 4" x 10" horns really shouldn't really be driven below 1.2Khz, and that is with a third-order/18dB per Octave HP filter. ( 500Hz is just no way )
I hope you haven't damaged them 😕
I agree that an L-pad is the best way to reduce their high efficiency output > it also provides a nice resistive load for the crossover that also damps resonance.
Those 4" x 10" horns really shouldn't really be driven below 1.2Khz, and that is with a third-order/18dB per Octave HP filter. ( 500Hz is just no way )
I hope you haven't damaged them 😕
I agree that an L-pad is the best way to reduce their high efficiency output > it also provides a nice resistive load for the crossover that also damps resonance.
To be clear, the new horns screwed onto the new drivers as shown in the photo are 15"x5". Does that seem at least adequate to take me down to a 500Hz crossover point or do you think they need to be bigger?
At 15" x 5" it is really the specifications of the compression driver that is going to matter most.
Do you know the specs. of the compression driver?
PS.
I think most would agree that a second order crossover is the minimum required. ( not just a single capacitor )
Do you know the specs. of the compression driver?
PS.
I think most would agree that a second order crossover is the minimum required. ( not just a single capacitor )
@WattPusher
Could you please post the Make & Model of your compression drivers > I am very curious 🙂
Could you please post the Make & Model of your compression drivers > I am very curious 🙂
Oh you have adjustable L-pads. Here is what I suggest.. you don't actually want to install those in the cab because they tend to become intermittent in PA applications, you want a fixed L-pad made from ceramic resistors added to the crossover. But you can use the adjustable unit to figure out what value resistors are required, just hook one up temporally with wires, dial in the correct balance with the pot, and then disconnect it and measure the resistance across both legs.
I hear you, but power loss is going to be pretty small with the L-pad correctly wired in between the crossover and the driver, that circuit may only see 10-20w at most.
Yeah PA cabs like this have upwards of 100db sensitivity which is 10-20db more than most home audio/studio speakers, it doesn't take much power to make lots of noise.
@Mister Audio - I appreciate your input; thanks. Specs are at https://timpanoaudio.com/product/tpt-d250x/.
Crossover is indeed 2nd-order: https://www.daytonaudio.com/product/598/xo3w-500-4k-3-way-crossover-500-4-000-hz
The two 12"s that I'm using in each cab are GRS 12SW-4 (4-ohm, connected in series) and their specs say they have a "usable frequency range" of 29–800 Hz. I'm sorry I didn't realize that Dayton Audio also made similar crossovers of
625/5000Hz and 700/5600Hz; I might have tried the 625/5000 if I'd known. I'm still waiting on delivery of the revised compression drivers.
Crossover is indeed 2nd-order: https://www.daytonaudio.com/product/598/xo3w-500-4k-3-way-crossover-500-4-000-hz
The two 12"s that I'm using in each cab are GRS 12SW-4 (4-ohm, connected in series) and their specs say they have a "usable frequency range" of 29–800 Hz. I'm sorry I didn't realize that Dayton Audio also made similar crossovers of
625/5000Hz and 700/5600Hz; I might have tried the 625/5000 if I'd known. I'm still waiting on delivery of the revised compression drivers.
You can not use a "universal" crossover, because this is a special speaker and not the unknown Dayton "universal driver combination".
The first rule of speaker construction: Every driver needs a crossover matched to it.
If you use the Dayton "500-4k" with serious level, you will burn the horns voice coil.
The first rule of speaker construction: Every driver needs a crossover matched to it.
If you use the Dayton "500-4k" with serious level, you will burn the horns voice coil.
@conanski - Great idea re using the adjustable pads to figure out values for "hard-resistoring" more permanent ones. Would I need to be using, say, 20W resistors for that?
BTW, with no EQ, the responses I've been getting with no EQ in line are already lumpy. I can compensate for that somewhat with the EQ even though its passbands are considerably wider than their 1/3rd-octave spacing. But that's okay.
@Turbowatch2 - Padding down the compression drivers post-power-amp has already been suggested and accepted especially since I do have a pair of adjustable pads already. I figure the drivers are going to do what they do and same with the crossovers; I've accepted that the matching of drivers and crossover may be imperfect (although I might should have gotten one of the other models with somewhat higher L/M cross now that I know there were such things).
BTW, with no EQ, the responses I've been getting with no EQ in line are already lumpy. I can compensate for that somewhat with the EQ even though its passbands are considerably wider than their 1/3rd-octave spacing. But that's okay.
@Turbowatch2 - Padding down the compression drivers post-power-amp has already been suggested and accepted especially since I do have a pair of adjustable pads already. I figure the drivers are going to do what they do and same with the crossovers; I've accepted that the matching of drivers and crossover may be imperfect (although I might should have gotten one of the other models with somewhat higher L/M cross now that I know there were such things).
Last edited:
Your choice of compression drivers seems pretty good to me, but the 500Hz XO will probably stimulate a small amount of their resonant frequency.
However the resistance of the L-pad / resistors will contribute a beneficial effect. Like you said, I would have preferred a 625/700Hz crossover F.
Also, because you are using Piezo drivers for high frequencies, I would connect a 10 ohm/10 watt resistor across their connection so the crossover
can recognize them as a 'dynamic driver' impedance > this will however make their impedance quite low @ 20Khz, but most amps won't mind this.
However the resistance of the L-pad / resistors will contribute a beneficial effect. Like you said, I would have preferred a 625/700Hz crossover F.
Also, because you are using Piezo drivers for high frequencies, I would connect a 10 ohm/10 watt resistor across their connection so the crossover
can recognize them as a 'dynamic driver' impedance > this will however make their impedance quite low @ 20Khz, but most amps won't mind this.
Yes
Yes that is to be expected.. especially with inexpensive drivers and even more so when they are horn loaded. Parametric EQ is the tool of choice these days because the filters can be custom taylored to shape the respomse exactly(or very nearly) as needed. Graphic EQ is like carving with a jackhammer.. sometimes it does as much damage as it does good.
Is the intent for the cabinet to still be for musical instrument or similar high efficiency use? If so I'm not sure the GRS 12SW woofers chosen are best for the task.
If you aren't set on keeping it an 8ohm cab there are GRS 12PT woofers that might be better suited & still inexpensive. This will make the cabinet nominally 4ohms, most modern amps won't mind. The crossover required for nominal 4ohm lows will be different.
As others have mentioned, a crossover designed for the chosen components is a better option than an off the shelf crossover. Being able to measure each section - lows, mid, & treble will be critical, especially since the woofers are in a unique cabinet & you may have drivers w/o reliable spec sheets.
If you aren't set on keeping it an 8ohm cab there are GRS 12PT woofers that might be better suited & still inexpensive. This will make the cabinet nominally 4ohms, most modern amps won't mind. The crossover required for nominal 4ohm lows will be different.
As others have mentioned, a crossover designed for the chosen components is a better option than an off the shelf crossover. Being able to measure each section - lows, mid, & treble will be critical, especially since the woofers are in a unique cabinet & you may have drivers w/o reliable spec sheets.
@conanski - Well, a 1/3-octave graphic EQ was something I could pick up from Music Go Round for something like $35. I had considered buying a Behringer DEQ2496 Ultracurve Pro (which takes measurements and EQs digitally but when I was putting all this together you couldn't buy one because they weren't available. So I made do.
For a time I had resisted replacing the cabs' original crossovers but as I became concerned about how unequal the two cabs' responses were, I realized that I really did not know if the existing crossovers were wired up legitimately - they were just components nailed to a board and soldered to a terminal strip with nothing labeled. I sketched the schematics from inspection and they resembled no other crossover circuits I'd found:
So yes, I acknowledge I've just thrown together some parts hoping to get a good result...and other than the unequal mids, I actually have. If they don't have audible problems once EQed as close to flat as I can make them and well-produced music sounds good through them, I'll have succeeded.
For a time I had resisted replacing the cabs' original crossovers but as I became concerned about how unequal the two cabs' responses were, I realized that I really did not know if the existing crossovers were wired up legitimately - they were just components nailed to a board and soldered to a terminal strip with nothing labeled. I sketched the schematics from inspection and they resembled no other crossover circuits I'd found:
So yes, I acknowledge I've just thrown together some parts hoping to get a good result...and other than the unequal mids, I actually have. If they don't have audible problems once EQed as close to flat as I can make them and well-produced music sounds good through them, I'll have succeeded.
@noggers - Because the crossover I'm using has the option of selecting either 4 ohms' or 8 ohms' worth of woofer, I could actually make the GRS 12PT-8 speakers work. I think I decided to go with the GRS 12SW-4 in part because of its Xmax of 8.5mm versus the 12PT-8's 6.8mm. I'm comfortable going forward with the 12SW-4s; so far so good sound-wise.
BTW I wasn't able to make a satisfactory nail file test but best I can tell, these cabs give up at about 80-85Hz. Not great, not terrible. But the original 12s were clearly insufficient and replacing them helped considerably. It does seem that one gets a bit of low range extension by putting the two cabs side-by-side. My test of setting flat panels to either side of the bass horn part of the cab showed that the panels did nothing. My primary use for these is going to be to run a keyboard rig through (physical and virtual instruments) but I am also intending to use them for film shoot playback and general PA.
BTW I wasn't able to make a satisfactory nail file test but best I can tell, these cabs give up at about 80-85Hz. Not great, not terrible. But the original 12s were clearly insufficient and replacing them helped considerably. It does seem that one gets a bit of low range extension by putting the two cabs side-by-side. My test of setting flat panels to either side of the bass horn part of the cab showed that the panels did nothing. My primary use for these is going to be to run a keyboard rig through (physical and virtual instruments) but I am also intending to use them for film shoot playback and general PA.