For now I think the simplest thing to try is to insert resistors in series with the mid and tweeter to bring the impedance up to 8 ohms or whatever will make the crossover components closer to being correct and see how the speaker sounds since the mid and tweeter (tweeter mostly) seem a bit loud compared to the woofer.
I'll do one first and compare the results.
If it's as simple as that, then I've fixed it while not having to go down the rabbit hole of redesigning the crossover.
I'll do one first and compare the results.
If it's as simple as that, then I've fixed it while not having to go down the rabbit hole of redesigning the crossover.
FWIW Ive repaired a few of these speakers (refoamed the woofers) and from listening tests and performing individual driver measurements, the original crossover design is somewhat lacking.
The crossover either needs steeper slopes and/or compensation networks for the mid and tweeter resonance peaks.
Also the mid dome will eventually disintegrate due to ageing.
good luck
The crossover either needs steeper slopes and/or compensation networks for the mid and tweeter resonance peaks.
Also the mid dome will eventually disintegrate due to ageing.
good luck
The crossover indeed needed some rework.
Calculationse I've done have proven that.
I tried a 3.3 ohm resistor in series with the tweeter, a 1 ohm resistor in series with the midrange and a 15uF cap in parallel with the 14uF midrange cap. I can already hear a drastic improvement in one speaker.
However, if all one listens to is jazz, the speakers would be fine as is.
So taking those impedances I crudely calculated and increasing them by the series resistors I get the following component values.
Tweeter 4.11uF cap
Midrange .21mH inductor 29.29uF cap.
Woofer 1.03mH
Tweeter 4uF will be fine.
Midrange .22mH with some removed to make .21mH will work. 30uF will work
Woofer 1mH will work.
That will give a mid to tweeter crossover of 4.62KHz and a mid to woofer crossover of 860Hz.
Now what I may do is take the crossovers home and use my DATS V3 speaker tester to modify the stock inductors to where they will be the proper values.
For now I'll leave the existing 14uF caps in and add the new 14uF caps in parallel until I get some 30uF caps.
Calculationse I've done have proven that.
I tried a 3.3 ohm resistor in series with the tweeter, a 1 ohm resistor in series with the midrange and a 15uF cap in parallel with the 14uF midrange cap. I can already hear a drastic improvement in one speaker.
However, if all one listens to is jazz, the speakers would be fine as is.
So taking those impedances I crudely calculated and increasing them by the series resistors I get the following component values.
Tweeter 4.11uF cap
Midrange .21mH inductor 29.29uF cap.
Woofer 1.03mH
Tweeter 4uF will be fine.
Midrange .22mH with some removed to make .21mH will work. 30uF will work
Woofer 1mH will work.
That will give a mid to tweeter crossover of 4.62KHz and a mid to woofer crossover of 860Hz.
Now what I may do is take the crossovers home and use my DATS V3 speaker tester to modify the stock inductors to where they will be the proper values.
For now I'll leave the existing 14uF caps in and add the new 14uF caps in parallel until I get some 30uF caps.
Last edited:
The tweeters do still sound slightly loud on some songs, however I'm going to wait until I make the inductors right before deciding if I need to do anything else with the tweeters.
If need be I'll use a series and parallel resistor to make an l-pad that won't alter the impedance.
I'll post a schematic of the updated crossover when I finish drawing it up.
If need be I'll use a series and parallel resistor to make an l-pad that won't alter the impedance.
I'll post a schematic of the updated crossover when I finish drawing it up.
Here's the partially assembled crossovers.
Got the inductors modified last night and secured.
The new caps came in today and I got them glued down. After an hour for the epoxy to cure I'll add the 15uF caps, resistors and solder it all together.
Now a 30uF film cap will not fit so I may leave the 15uF caps in parallel with the 14uF caps.
Once all is soldered, I'll glue the resistors down as well. Then I'll add 3"-4" wires to the connections on the crossover and glue them down. The stock wiring to the crossover is quite short unfortunately.
For the three wires running to the - input terminal I'll just install one wire and connect it to the three wires.
I also used the plastic epoxy to secure the spring loaded terminal assembly back to each terminal cup.
Got the inductors modified last night and secured.
The new caps came in today and I got them glued down. After an hour for the epoxy to cure I'll add the 15uF caps, resistors and solder it all together.
Now a 30uF film cap will not fit so I may leave the 15uF caps in parallel with the 14uF caps.
Once all is soldered, I'll glue the resistors down as well. Then I'll add 3"-4" wires to the connections on the crossover and glue them down. The stock wiring to the crossover is quite short unfortunately.
For the three wires running to the - input terminal I'll just install one wire and connect it to the three wires.
I also used the plastic epoxy to secure the spring loaded terminal assembly back to each terminal cup.
I measured the stock caps and they were within a reasonable tolerance and when I did the mods using the stock caps with the resistors and adding the 15uF cap in parallel with the 14uF cap, the speakers sounded much better. Listened to them pretty much all day Friday at work and was very pleased.
I know the film caps may alter the sound some, but I know they will only make the speakers sound better.
I know the film caps may alter the sound some, but I know they will only make the speakers sound better.
Now to me it sounds like the midrange driver is not quite high enough in output, however I need to do a lot more listening to be sure.
Also I need to remove the midrange and woofer from one cabinet and take them home so that I can measure their parameters with the DATS V3 speaker tester in order to add a zobel and notch filter which will make the speakers sound better. That will also determine if my crude method of measuring the driver impedance was close enough. The zobel will keep the impedance of the drivers more constant as the frequency increases which will make the crossover more accurate and might be all I need to do. I did use zobel networks on a pair of speakers I made a few years back and was well pleased with the results.
https://soundcertified.com/zobel-network-series-notch-filter-calculator/
EDIT:
That said, if I had one more Adcom GFA-535II amp I'd go with a DSP and make the speakers fully active.
That said I do have a Sony TA-E2000ESD preamp that has a problem of no audio on the output. When I fix that I may bring it to work and see if it makes the speakers sound any better than the McIntosh C-24 preamp.
It also has an output for a subwoofer so I might not even need the external crossover provided the main speaker outputs have a selectable high pass crossover.
I could then possibly use one channel of the Adcom GFA-545II for the subwoofer with one Adcom GFA-535II used for the main speakers. Given the amps are pretty much the same only different power outputs, the sub and mains will integrate much better, although they seem to integrate just fine now.
Also I need to remove the midrange and woofer from one cabinet and take them home so that I can measure their parameters with the DATS V3 speaker tester in order to add a zobel and notch filter which will make the speakers sound better. That will also determine if my crude method of measuring the driver impedance was close enough. The zobel will keep the impedance of the drivers more constant as the frequency increases which will make the crossover more accurate and might be all I need to do. I did use zobel networks on a pair of speakers I made a few years back and was well pleased with the results.
https://soundcertified.com/zobel-network-series-notch-filter-calculator/
EDIT:
That said, if I had one more Adcom GFA-535II amp I'd go with a DSP and make the speakers fully active.
That said I do have a Sony TA-E2000ESD preamp that has a problem of no audio on the output. When I fix that I may bring it to work and see if it makes the speakers sound any better than the McIntosh C-24 preamp.
It also has an output for a subwoofer so I might not even need the external crossover provided the main speaker outputs have a selectable high pass crossover.
I could then possibly use one channel of the Adcom GFA-545II for the subwoofer with one Adcom GFA-535II used for the main speakers. Given the amps are pretty much the same only different power outputs, the sub and mains will integrate much better, although they seem to integrate just fine now.
Last edited:
So I fixed the Sony which just needed the output jacks resoldered to the board and tried it. I believe the Sony sounds better with the Adcom GFA-535II and the Advent speakers.
I also determined the sub output of the Sony has no way to adjust the level and crossover frequency. I suspect that Dolby pro logic possibly had a specific crossover frequency for the sub output. Also the main outputs do not have a high pass so it was not possible to use the subwoofer output and make the sub integrate properly.
Thought the tweeters seemed a bit loud, but I believe it's just the streaming station I was using this morning.
That said once I add the zobel to the woofer and midrange that may tame things down a bit.
I use a powered sub cabinet that I put a 100 watt plate amp and a Sony red cone XPLOD 10" driver in as the stock components weren't good.
It sounds good, but what I'm thinking of doing is fixing the Adcom GFA-535II using that for the mains then using one channel of the 545II to drive the sub as given both amps are close to the same, they will sound the same and as such the sub will integrate better. I'll most likely install a 1/4" switched jack on the plate amp so that when a 1/4" plug is plugged in it connects the driver to the external amp. That way if need be, the sub can still be used as a stand alone. I will need to add one section of a dual OP-AMP wired for a gain of 2 with the second section wired as an inverting unity gain buffer to the low pass output of the active crossover. I'll then use a switch to select the non-invert or invert output so that I can easily make the sub be in phase with the mains.
I also determined the sub output of the Sony has no way to adjust the level and crossover frequency. I suspect that Dolby pro logic possibly had a specific crossover frequency for the sub output. Also the main outputs do not have a high pass so it was not possible to use the subwoofer output and make the sub integrate properly.
Thought the tweeters seemed a bit loud, but I believe it's just the streaming station I was using this morning.
That said once I add the zobel to the woofer and midrange that may tame things down a bit.
I use a powered sub cabinet that I put a 100 watt plate amp and a Sony red cone XPLOD 10" driver in as the stock components weren't good.
It sounds good, but what I'm thinking of doing is fixing the Adcom GFA-535II using that for the mains then using one channel of the 545II to drive the sub as given both amps are close to the same, they will sound the same and as such the sub will integrate better. I'll most likely install a 1/4" switched jack on the plate amp so that when a 1/4" plug is plugged in it connects the driver to the external amp. That way if need be, the sub can still be used as a stand alone. I will need to add one section of a dual OP-AMP wired for a gain of 2 with the second section wired as an inverting unity gain buffer to the low pass output of the active crossover. I'll then use a switch to select the non-invert or invert output so that I can easily make the sub be in phase with the mains.
Enjoying your Maestro travels. I too have been enjoying my newly acquired Advent Maestro's. I replaced the capacitors along with replacing the midrange iron core coil with an air coil. I also removed that degraded foam Jensen/Advent used on the front baffle (Naptha and a puddy knife). I braced the cabinets and added addition dampening material. In addition, I used Speaker Gasketing Tape on all the speakers and rear speaker/XO panel. I also have found they tend to be a little brighter but the midrange is somewhat forward but not to a degree that will not need adjusting for my taste. I drive them with a older yamaha 100 watts / channel stereo receiver. I use the variable loudness control common on Yamaha receivers which works very well in reducing the midrange when needed. I believe the Maestros are an incredible speaker which can be had for very little money. Below is the frequency graph of my maestros which I would consider as very close to OEM.
Was there an audible improvement replacing the iron core with an air core?
Also someone over on the parts express forum had restored theirs and posted a response graph as well.
If I had to judge mine based on the graph I'd say mine were nowhere close to being anywhere near proper at the crossover frequencies.
I had to reduce the inductor values and increase the cap value on the midrange driver to get them to sound proper.
I also intend to measure the midrange and woofer parameters using a DATS V3 speaker tester so that I can get the component values needed for zobel networks which will further improve the sound.
Also someone over on the parts express forum had restored theirs and posted a response graph as well.
If I had to judge mine based on the graph I'd say mine were nowhere close to being anywhere near proper at the crossover frequencies.
I had to reduce the inductor values and increase the cap value on the midrange driver to get them to sound proper.
I also intend to measure the midrange and woofer parameters using a DATS V3 speaker tester so that I can get the component values needed for zobel networks which will further improve the sound.
As for the air core coil I did not test play or graph the speakers before replacing the crossover parts. I just had to get that decomposing foam off the front baffle. Then I moved on to bracing the cabinet while I waited for the crossover parts. In the end, replacing the iron core with an air core was within my budget so I just did it. As for the crossover wiring, yes the tweeter and mid are reversed wire as you said. which I did not change. I just finished a one hour listening session with the Maestro's and I just love them. The imaging and sound stage are to my liking. My current project is a set of Advent Legacy III's, what a goofy XO and ohm switch setup. The 10" woofer really benefits from the added midrange in the Maestro's three way design. Good luck with the zobel network upgrade, though I really do not have a clue what that does. I will follow your progress!
The zobel keeps the driver impedance more constant as the frequency increases.
Not sure why, but with the crossover stock and wiring stock I simply could not get good sound and forget trying to integrate a sub.
The best way to describe the sound is the tweeter sounded too loud and the midrange was lacking at the lower end with the woofer lacking at the upper end and not loud enough.
Reversing the midrange and tweeter polarity helped.
I then used a 3.3 ohm series resistor with the tweeter to make the crossover frequency close to the published spec of 4.5kHz.
I used a 1 ojm series resistor with the midrange as that would require an easy to get 30uF cap so the midrange would meet its lower listed crossover frequency spec of 900Hz. Now because I had already ordered the caps (10uF and 4uF in parallel) when I discovered the crossover was out of whack I just put a 15uF electrolytic in parallel with the 14uF film cap for now. Plus a 30uF cap would be too large. When I order the zobel components I'll add a couple 1uF film caps to the order and put them in parallel with the 14uF and 15uF caps in parallel with each other to equal 30uF.
The sound improved greatly after those mods.
After making the inductors the value the crossover calculator said, the sound improved even more.
I expect the zobels to improve the sound around the woofer to midrange crossover frequency and the midrange to tweeter crossover frequency.
The jury is still out on whether or not the tweeters are still too loud though. If they are I'll try a fixed l-pad with two resistors before the 3.3 ohm series resistor as to add more series resistance will require a capacitor less than 4uF.
Given I use these with a powered sub I added a 320uF cap in series with the woofer after the 1mH inductor. That way I don't need an active crossover and can use the speaker level inputs of the plate amp.
For a bandpass given the cap and inductor are in series, does it matter if the cap is connected before or after the inductor?
When I first heard these speakers and heard the deficiencies in the sound, I immediately started calculating and found the crossover was way out of whack.
The woofer low pass was one frequency.
The midrange bandpass was one frequency range.
The tweeter high pass was one frequency.
None of the frequencies at the crossover points matched.
I suspect that may be why the tweeter and midrange had to be wired out of phase as it may have made the speaker sound a little better.
That said I am assuming there were no revisions to the crossover during the production life of the speaker.
Also I'm kind of surprised that no one has posted any mods to the crossover. It could be that people didn't think much of this speaker as it's not the typical two way speaker Advent was known for and felt it wasn't worth the time or perhaps many felt that they needed all sorts of measurement equipment to know what changes to make.
All it required was a crossover calculator, time and my ears.
The original owner listened to them late last week and noted the improvement in sound quality so it brings more validity to the changes I made as it isn't just me thinking the sound quality has improved.
I use a Schiit Saga S in passive mode to drive the Adcom GFA-545 amp powering these speakers at work, however management has taken issue with the speakers but won't tell me why so I may take them home tomorrow.
I'd like to figure a way to add a SPDT or SPST switch to each crossover so that I can switch the woofer high pass in or out.
Eventually I may do like the person on the parts express forum did which was mount the crossover components on the cabinet bottom and covered the hole where the terminal cup went with a wood board where better terminals were installed. I then would have enough space to add a switch.
One possible option is to remove the fuse holder and find a push on/off switch to fit the hole.
Using the DATS V3 speaker tester to make whatever inductor values I need really makes it easier to build and modify speakers.
I suspect this speaker was a case of engineer spec'd one thing, but the parts used were gotten for a good price or were already used in other speaker models Jensen made and marketing probably said who needs resistors they just add to the cost.
Not sure why, but with the crossover stock and wiring stock I simply could not get good sound and forget trying to integrate a sub.
The best way to describe the sound is the tweeter sounded too loud and the midrange was lacking at the lower end with the woofer lacking at the upper end and not loud enough.
Reversing the midrange and tweeter polarity helped.
I then used a 3.3 ohm series resistor with the tweeter to make the crossover frequency close to the published spec of 4.5kHz.
I used a 1 ojm series resistor with the midrange as that would require an easy to get 30uF cap so the midrange would meet its lower listed crossover frequency spec of 900Hz. Now because I had already ordered the caps (10uF and 4uF in parallel) when I discovered the crossover was out of whack I just put a 15uF electrolytic in parallel with the 14uF film cap for now. Plus a 30uF cap would be too large. When I order the zobel components I'll add a couple 1uF film caps to the order and put them in parallel with the 14uF and 15uF caps in parallel with each other to equal 30uF.
The sound improved greatly after those mods.
After making the inductors the value the crossover calculator said, the sound improved even more.
I expect the zobels to improve the sound around the woofer to midrange crossover frequency and the midrange to tweeter crossover frequency.
The jury is still out on whether or not the tweeters are still too loud though. If they are I'll try a fixed l-pad with two resistors before the 3.3 ohm series resistor as to add more series resistance will require a capacitor less than 4uF.
Given I use these with a powered sub I added a 320uF cap in series with the woofer after the 1mH inductor. That way I don't need an active crossover and can use the speaker level inputs of the plate amp.
For a bandpass given the cap and inductor are in series, does it matter if the cap is connected before or after the inductor?
When I first heard these speakers and heard the deficiencies in the sound, I immediately started calculating and found the crossover was way out of whack.
The woofer low pass was one frequency.
The midrange bandpass was one frequency range.
The tweeter high pass was one frequency.
None of the frequencies at the crossover points matched.
I suspect that may be why the tweeter and midrange had to be wired out of phase as it may have made the speaker sound a little better.
That said I am assuming there were no revisions to the crossover during the production life of the speaker.
Also I'm kind of surprised that no one has posted any mods to the crossover. It could be that people didn't think much of this speaker as it's not the typical two way speaker Advent was known for and felt it wasn't worth the time or perhaps many felt that they needed all sorts of measurement equipment to know what changes to make.
All it required was a crossover calculator, time and my ears.
The original owner listened to them late last week and noted the improvement in sound quality so it brings more validity to the changes I made as it isn't just me thinking the sound quality has improved.
I use a Schiit Saga S in passive mode to drive the Adcom GFA-545 amp powering these speakers at work, however management has taken issue with the speakers but won't tell me why so I may take them home tomorrow.
I'd like to figure a way to add a SPDT or SPST switch to each crossover so that I can switch the woofer high pass in or out.
Eventually I may do like the person on the parts express forum did which was mount the crossover components on the cabinet bottom and covered the hole where the terminal cup went with a wood board where better terminals were installed. I then would have enough space to add a switch.
One possible option is to remove the fuse holder and find a push on/off switch to fit the hole.
Using the DATS V3 speaker tester to make whatever inductor values I need really makes it easier to build and modify speakers.
I suspect this speaker was a case of engineer spec'd one thing, but the parts used were gotten for a good price or were already used in other speaker models Jensen made and marketing probably said who needs resistors they just add to the cost.
Last edited: