Hi all,
I'm working on a 3-way active Linkwitz-Riley crossover (24dB/oct), based on the ESP P09 topology. The setup is for home hi-fi use at moderate listening levels — I rarely play loud, and my focus is on sound quality, transparency, and phase coherence.
My speaker system consists of:
Tweeter: Electro-Voice ST350B
Midrange: Electro-Voice EVM-10M in a sealed 18L chamber
Woofer: JBL 4227 in a 130L bass reflex cabinet
The crossover is built on a KMTech board. I’m using LME49720HA opamps with proper bypassing and layout attention. I also have a large stock of mil-spec 10k resistors with tight tolerance, and a selection of precision film capacitors including rare KS and military polypropylene types.
What I’d like your input on:
1. Suggestions for optimal crossover point placement for these specific drivers, considering phase alignment and natural acoustic blending.
2. Thoughts on using LME49720HA opamps in this type of filter — are there any caveats or fine points I should be aware of?
3. In general, what’s your preferred strategy for selecting component values: fixing R and calculating C, or the reverse? Any experience-based sweet spots in terms of impedance?
4. Any practical tips from your own active crossover builds with similar gear?
Thanks in advance for sharing your experience — much appreciated!
I'm working on a 3-way active Linkwitz-Riley crossover (24dB/oct), based on the ESP P09 topology. The setup is for home hi-fi use at moderate listening levels — I rarely play loud, and my focus is on sound quality, transparency, and phase coherence.
My speaker system consists of:
Tweeter: Electro-Voice ST350B
Midrange: Electro-Voice EVM-10M in a sealed 18L chamber
Woofer: JBL 4227 in a 130L bass reflex cabinet
The crossover is built on a KMTech board. I’m using LME49720HA opamps with proper bypassing and layout attention. I also have a large stock of mil-spec 10k resistors with tight tolerance, and a selection of precision film capacitors including rare KS and military polypropylene types.
What I’d like your input on:
1. Suggestions for optimal crossover point placement for these specific drivers, considering phase alignment and natural acoustic blending.
2. Thoughts on using LME49720HA opamps in this type of filter — are there any caveats or fine points I should be aware of?
3. In general, what’s your preferred strategy for selecting component values: fixing R and calculating C, or the reverse? Any experience-based sweet spots in terms of impedance?
4. Any practical tips from your own active crossover builds with similar gear?
Thanks in advance for sharing your experience — much appreciated!
Here are some frequencys that i think are optimal for each driver.
JBL 4227: 30 Hz – 400 Hz
EVM-10M: 300 Hz – 4 kHz
EV ST350B: 3.5 kHz – 16kHz
Im aware that tweeter is not a super tweeter, but i have listened to it since 2016 with passive crossover, and i still like it alot, i see no need to go much higher, it rolls of around 17khz.
JBL 4227: 30 Hz – 400 Hz
EVM-10M: 300 Hz – 4 kHz
EV ST350B: 3.5 kHz – 16kHz
Im aware that tweeter is not a super tweeter, but i have listened to it since 2016 with passive crossover, and i still like it alot, i see no need to go much higher, it rolls of around 17khz.
I would add:
If you don't plan to make further equalization circuits the usual strategy is to cross from woofer to midrange at the onset of baffle step.
Reducing midrange level does the baffle step compensation.
But I suggest to first build the enclosure, then take measurements and then simulate and decide on x-over points.
Oh, I see.
You could also measure the electric responses after the existing filter (with drivers connected) and try to replicate them actively.
That is, if you are satisfied with the passive crossover!
Cross the woofer to mid at 400–500Hz and mid to tweeter at 3.2–3.5kHz for best blend and phase alignment. The LME49720HA is great—just ensure proper decoupling and avoid long cables on the output. Stick with your 10kΩ resistors and calculate caps; it keeps things clean and in the ideal range for film types. Add trimpots for level matching, check polarity, and use REW for final tuning. You're on the right path—this setup should sound excellent.
Or better, use 100 ohm series resistors on the outputs so the opamp is stable whatever the capacitive load...
Active filter is comming along, i used lower value resistors all over , 4k32, then 10nf and 92 nf giving crossover over at 397hz and 3680hz.
I opted for lower value resistors because i had them, and lme49720ha easily can drive them, i suspect a tad lower noise , there is footprint for trimmere yes , and the idea of a 100hm output resistor is noted.. and yes i need washers under nuts..
I opted for lower value resistors because i had them, and lme49720ha easily can drive them, i suspect a tad lower noise , there is footprint for trimmere yes , and the idea of a 100hm output resistor is noted.. and yes i need washers under nuts..
Is that a ESP P09 based board? I thought the P09 used more components than that, is all. I very well may be mistaken (I usually am!).
Last edited:
It should be stuey...
I have spend some days now matching capacitors, all 20 10nf are now within 0.5 percent, alle are Russian k71-7 .
I did a few sweeps, on bench supply it is Rock steady into MHz, 5mhz seems to be the wall.. pretty cool..
Phase follows along nicely
I have spend some days now matching capacitors, all 20 10nf are now within 0.5 percent, alle are Russian k71-7 .
I did a few sweeps, on bench supply it is Rock steady into MHz, 5mhz seems to be the wall.. pretty cool..
Phase follows along nicely
You could use a digital crossover to prototype alternative crossovers.
I'm using a miniDSP for this ATM.
I'm using a miniDSP for this ATM.
It definitely looks like a KMTech board (available on the *Bay), not like Rod Elliott’s P09 PCB — I think.
@amplidude What are those cool cases on the LME49720HAs for?
Btw does anyone know of other boards for a similar circuit to Rod's? I need a 6th order crossover, preferably with output volume trimmers which Rod's P09 boards lack.
Btw does anyone know of other boards for a similar circuit to Rod's? I need a 6th order crossover, preferably with output volume trimmers which Rod's P09 boards lack.
Oops, apologies, Rod's P09 boards do have output trimmers; it's his P148 variable-frequency crossovers which don't.