I have not actually listened to it yet. I only did the Miflex bypass cap test. That’s a great price. Go for it.
Hi Florian,
I am not sure why you think I don’t respect what you presented. I have used very fair and objective words and responses with you. I may not be jumping up to try it because I have done this sort of stuff already years ago when I had DSP and could adjust it at will. Don’t forgot this thread has been going for about 5 years now. I have tried many crossovers and many listening tests and thousands of measurements. I have found that the speakers sound best to me when the step response looks like a sharp edged right triangle - a classic indication that one has a good time alignment with a 1st order crossover. Indeed, when I move away from the measurement sweet spot, maybe it is delayed differently and closer to what you show. My measurement is at 0.5m to avoid room reflections. At 2.5 meters away, that alignment is less aligned and less critical. Room reflections start to dominate.
I am actually not clear what you are trying to recommend that we do differently. Something about adding more mechanical delay to the tweeter relative to the woofer and putting the tweeter on top in a separate box. I think Pcbab and Perceval both made something like this. It might be easier for him to try this. As for me - I am stuck with this flat baffle and two drivers. But what you recommend, is it clearly different than the step responses we have now? Is it worth the complexity of a second box vs a single rectangular easy to make enclosure?
So here was Perceval’s, looks like about 2.5in setback?
And here was PCgab’s:
So you are, stepped baffle with extra delay isn’t exactly new in this thread.
But years ago, I had a separate full range mounted on a separate Dagger and this sat on top of a separate woofer. And I indeed slid it back and forth probably over 6in in 1/2in increments. I can’t find the thread now but recall it was a trapezoidal baffle with a Dagger rear chamber. I had a passive 1st order crossover with a huge motor run cap. Can’t find that post now. But basically, we have been there and done it already. Years ago - so it’s not exactly a new discovery - it may even be in this thread.
So I would suggest you restate your objectives, steps taken, show data, in a clear description. Please look at previous similar work where fullrange tweeter was separate and adjusted in position for best time alignment. Show how yours is different from my Dagger and Perceval’s and Pcgab’s efforts at same thing.
Cheers,
X
Edit, I found mine:
Subjective Blind ABX Test of EnABLed FF85WK - Round 6
Looks like about 2.5in maybe 3in setback. So you see, I have done this! And the step response is sharp right triangle:
I am not sure why you think I don’t respect what you presented. I have used very fair and objective words and responses with you. I may not be jumping up to try it because I have done this sort of stuff already years ago when I had DSP and could adjust it at will. Don’t forgot this thread has been going for about 5 years now. I have tried many crossovers and many listening tests and thousands of measurements. I have found that the speakers sound best to me when the step response looks like a sharp edged right triangle - a classic indication that one has a good time alignment with a 1st order crossover. Indeed, when I move away from the measurement sweet spot, maybe it is delayed differently and closer to what you show. My measurement is at 0.5m to avoid room reflections. At 2.5 meters away, that alignment is less aligned and less critical. Room reflections start to dominate.
I am actually not clear what you are trying to recommend that we do differently. Something about adding more mechanical delay to the tweeter relative to the woofer and putting the tweeter on top in a separate box. I think Pcbab and Perceval both made something like this. It might be easier for him to try this. As for me - I am stuck with this flat baffle and two drivers. But what you recommend, is it clearly different than the step responses we have now? Is it worth the complexity of a second box vs a single rectangular easy to make enclosure?
So here was Perceval’s, looks like about 2.5in setback?
And here was PCgab’s:
So you are, stepped baffle with extra delay isn’t exactly new in this thread.
But years ago, I had a separate full range mounted on a separate Dagger and this sat on top of a separate woofer. And I indeed slid it back and forth probably over 6in in 1/2in increments. I can’t find the thread now but recall it was a trapezoidal baffle with a Dagger rear chamber. I had a passive 1st order crossover with a huge motor run cap. Can’t find that post now. But basically, we have been there and done it already. Years ago - so it’s not exactly a new discovery - it may even be in this thread.
So I would suggest you restate your objectives, steps taken, show data, in a clear description. Please look at previous similar work where fullrange tweeter was separate and adjusted in position for best time alignment. Show how yours is different from my Dagger and Perceval’s and Pcgab’s efforts at same thing.
Cheers,
X
Edit, I found mine:
Subjective Blind ABX Test of EnABLed FF85WK - Round 6
Looks like about 2.5in maybe 3in setback. So you see, I have done this! And the step response is sharp right triangle:
Last edited:
I'll using your provided crossover for the Rs225-4 &ps95 linked in the first post. That's correct?
Yes, it was for RS225-4 and RS100-8, but the PS95-8 can be substituted just adjust the attenuation resistor to taste.
Sounds good. I wanted to use the rs100 but wife likes look of ps95 lol. Looking forward to this build- I'll get to purchase enough for 7 speakers in two weeks. These will be my new "floor" channels in our 7.4.4 theater
I have 10 RS225-4 already sitting around collecting dust. I used factory FRD for both in Xsim and applied your crossover design and it looked similar. Wasn't sure what to do with them until I seen this thread lol
Heres what it simulates as in Xsim- I had to reverse polarity on PS95- doing so on RS225-4 is different phase.
Let me know what you think- Not sure what else to do with these 10-RS225-4 i have sitting around collecting dust.
Attachments
There is a time delay if the woofer is on top and same as ear level. The delay will fix the phase flip that you had to add. It shouldn’t be reversed as that would not be a transient perfect speaker anymore. There is no reason that the PS95 would need to be reversed. Try adding equivalent of about 2.5in of time of flight delay to the PS95. Then try the XO tweaks.
I'm trying to remember what I settled on before I stored them. In my original, one piece baffle the set back was ~1.5" to the Fostex FF85, then I made another set of baffles with ~3". The 3" as I remember it sounded much more coherent, and measured really well. The 3" was more a function of 4 layers of mdf, as opposed to 2 layers. It didn't measure as well as the one you posted certainly, but it was close.
I've been thinking about resurrecting these again, I made some boxes for the ff85's that sit on top, similar to B&W 801, or so I would like to think.
tl;dr:
~2.5" is likely the sweet spot.
That looks better (both drivers positive phase) but there is a lot of dips and peaks in the midrange. Is that just the natural non-flat response of the PS95?
Your cap on PS95 is too big, it will lead to distortion as it makes the PS95 play too low. Make that 60uF.
Look at the Step Response graph and check your time alignment that way. The phase should be relatively flat from 100Hz to 10kHz.
This woofer is more sensitive, make R1 a short (0ohm jumper) and make R2 open (not populated) and that should balance the treble with bass more.
Aim for 900Hz crossover frequency by adjustment of R3 and C3.
Your cap on PS95 is too big, it will lead to distortion as it makes the PS95 play too low. Make that 60uF.
Look at the Step Response graph and check your time alignment that way. The phase should be relatively flat from 100Hz to 10kHz.
This woofer is more sensitive, make R1 a short (0ohm jumper) and make R2 open (not populated) and that should balance the treble with bass more.
Aim for 900Hz crossover frequency by adjustment of R3 and C3.
Last edited:
FAST array?
I'm kicking around the idea of using the RS225 with an "array" (I use that term lightly; I'm talking 3 speakers in a vertical configuration to start with) of CHP-70 Mark Audio drivers (see the link below) in a FAST design. I checked with Dave Dlugos, the designer of the array, and he said the CHP-70 is a drop-in replacement for the CHR-70. I'd like to try an array, but I really don't know enough to "design" one that will sound good. So, I have some questions.
1)I might be way off base, but does this sound reasonable?
2) Would the crossover need to change much?
3) I haven't used a DSP, but would that be a better idea than a crossover?
4) Would it be better to go with a pseudo-array of the PS95 or TC9?
http://www.madisound.com/pdf/TheCHRArray-plan-250811.pdf
Thanks,
Mike
I'm kicking around the idea of using the RS225 with an "array" (I use that term lightly; I'm talking 3 speakers in a vertical configuration to start with) of CHP-70 Mark Audio drivers (see the link below) in a FAST design. I checked with Dave Dlugos, the designer of the array, and he said the CHP-70 is a drop-in replacement for the CHR-70. I'd like to try an array, but I really don't know enough to "design" one that will sound good. So, I have some questions.
1)I might be way off base, but does this sound reasonable?
2) Would the crossover need to change much?
3) I haven't used a DSP, but would that be a better idea than a crossover?
4) Would it be better to go with a pseudo-array of the PS95 or TC9?
http://www.madisound.com/pdf/TheCHRArray-plan-250811.pdf
Thanks,
Mike
Sorry, you lost me with why the use of 3x CHP70 and an RS225?
The three drivers are in parallel I assume? Sensitivity will be way too high for the single RS225 after baffle step losses (circa 82.5dB at 2.83v). You will just be padding down the three CHPs. You will also get vertical comb interference at closer distances. Now maybe 3 CHPs with 4x RS225-8 in series parallel might make more sense. But it’s a different speaker now. If you kept the baffle width the same maybe the same XO could be used but the full range needs to be 8ohms and about 88dB.
So really, just one is needed. For the price of three CHPs? A single 10F/8414 might make more sense and IMO, sound clearer and more coherent.
The three drivers are in parallel I assume? Sensitivity will be way too high for the single RS225 after baffle step losses (circa 82.5dB at 2.83v). You will just be padding down the three CHPs. You will also get vertical comb interference at closer distances. Now maybe 3 CHPs with 4x RS225-8 in series parallel might make more sense. But it’s a different speaker now. If you kept the baffle width the same maybe the same XO could be used but the full range needs to be 8ohms and about 88dB.
So really, just one is needed. For the price of three CHPs? A single 10F/8414 might make more sense and IMO, sound clearer and more coherent.
Thanks X,
With your comment in mind, I'll forego the array route; thus, the reason I asked. I'll keep my FAST as is unless I try one of the other drivers you posted. I'll look into a pair of the 10F/8414's.
For now, I think I'm going to get a mini DSP to start playing with. I have one in the Madisound cart and I'm trying to decide if it's worthwhile. I couldn't get my mind wrapped around crossover design, so I'm thinking maybe a mini DSP will be more intuitive in my mind.
Thanks again!
Mike