I had to combine parts to get the exact values so the prototype crossover used more than 13 parts. I am glad I have my XO development kit This is my first 3-way passive crossover. I worked from the woofer and mid on up in order to set the level for padding the higher frequencies
This is a special speaker. It sounds fantastic! The sound is so clean and the bass is clear and articulate. Mids are smooth and highs have sparkle from the planar.
Have a listen:
This is a special speaker. It sounds fantastic! The sound is so clean and the bass is clear and articulate. Mids are smooth and highs have sparkle from the planar.
Have a listen:
Last edited:
I found an incorrect resistor value (1ohm vs 10ohm on the notch filter). Replaced that and it measures very well. Listening to it and voicing the tweeter with 3.3ohm padding resistor seems to give the best balance. Here is measured frequency response of the crossover.
Here is the measured distortion at 2.0Vrms:
Here is the measured phase, this needs work as the sims predict flat phase (quasi transient perfect from 100Hz to 15kHz):
And here is the predicted impedance variation, the last one I posted is incorrect:
Here is the measured distortion at 2.0Vrms:
Here is the measured phase, this needs work as the sims predict flat phase (quasi transient perfect from 100Hz to 15kHz):
And here is the predicted impedance variation, the last one I posted is incorrect:
I assembled the second channel with 8 more of the alternative woofers from Pyle (PDMW6) with higher sensitivity. Because of the compact layout of opposed drivers aligned vertically and 90deg relative to baffle, one does not appreciate the “wall of woofers” that is afforded by 8x 6.5in woofers. If one were to mount all of these flat on the baffle it would look a bit crazy big. This is why the bass has such punch. A lot of air being moved and pumped out of those slots.
I had a hard time getting the baskets to sit inside on one side of the woofer box. So all inside or outside on this one. No big deal as still works the same.
Some quick measurements reveal that this can now be an overall 92dB at 2.83v system. But the bass extension is only 49Hz at -3dB. It’s a tradeoff as usual. I think I like 43Hz better at 90dB. Upper reach is the same at 480Hz before falloff due to 1/4wave chamber depth. As usual, important to follow a schematic when wiring something like this 4x parallel and 2x series to 2x parallel for 4ohms. A battery was used to check overall correct polarity of all 8. They all need to push into the chamber on +ve.
Some quick measurements reveal that this can now be an overall 92dB at 2.83v system. But the bass extension is only 49Hz at -3dB. It’s a tradeoff as usual. I think I like 43Hz better at 90dB. Upper reach is the same at 480Hz before falloff due to 1/4wave chamber depth. As usual, important to follow a schematic when wiring something like this 4x parallel and 2x series to 2x parallel for 4ohms. A battery was used to check overall correct polarity of all 8. They all need to push into the chamber on +ve.
The mid range and tweeter are quasi transient perfect (flat phase and time aligned). The slot loading on the woofers creates an extended delay that makes it impossible to time align the woofers with the mid/tweeter unless DSP is used. In practice though, the sharpness of the percussion is still excellent because the mid and tweeter are fine aligned. The step response will show two peaks (first is time aligned right triangle with tweeter and mid) followed by a bass response hump. If DSP and 2 ch active amp (woofer plus passive mid and tweeter) then the acoustic output would be all time aligned and quasi transient perfect.
you can see the measured phase down to the crossover of the mid and woofer is flat.
Hi Plott,
I am temporarily using the Pyle paper cone on one channel until 8 more of the GRS poly cone ones arrive. The GRS 6PR-8 6.5in poly cone rubber surround woofer. But almost any low cost high Qts (circa 1+) is good for this. Parts Express used to have $5 blowout 6.5in 4ohm woofers that I used on the original cardboard prototype. The higher Qts allows deeper bass relative to the Fs. One option is to use say, 6x 8in woofers which will have a lower Fs and decrease crossover to 350Hz since chamber depth is deeper.
For midrange I am using Faital Pro 6FE100 and that is working well. There may be a higher performance option with cast frame baskets, Neodymium motors, copper shorting rings etc type drivers like Faital Pro 6PR160 that should be lower distortion. On cardboard prototype I used PRV 5MR450NDY which is also excellent. You need a 93-95dB sensitive mid to keep up with the sensitivity of the bass drivers when in series parallel.
For the tweeter, I am using the GRS PT2522-4 4ohm planar (open back dipole version).
This is turning out to be a true multi-way project. I’m wondering if it should be started as new thread in Multi-way forum. If one used a single full range driver on top then it would be a FAST SLOB.
I am temporarily using the Pyle paper cone on one channel until 8 more of the GRS poly cone ones arrive. The GRS 6PR-8 6.5in poly cone rubber surround woofer. But almost any low cost high Qts (circa 1+) is good for this. Parts Express used to have $5 blowout 6.5in 4ohm woofers that I used on the original cardboard prototype. The higher Qts allows deeper bass relative to the Fs. One option is to use say, 6x 8in woofers which will have a lower Fs and decrease crossover to 350Hz since chamber depth is deeper.
For midrange I am using Faital Pro 6FE100 and that is working well. There may be a higher performance option with cast frame baskets, Neodymium motors, copper shorting rings etc type drivers like Faital Pro 6PR160 that should be lower distortion. On cardboard prototype I used PRV 5MR450NDY which is also excellent. You need a 93-95dB sensitive mid to keep up with the sensitivity of the bass drivers when in series parallel.
For the tweeter, I am using the GRS PT2522-4 4ohm planar (open back dipole version).
This is turning out to be a true multi-way project. I’m wondering if it should be started as new thread in Multi-way forum. If one used a single full range driver on top then it would be a FAST SLOB.