After a long busy summer, I'm back to this project. I gave up on doing the woodwork myself, I started on it but it wasn't going to be living room acceptable. I found a local CNC plywood shop to make baffles for me. So...here I am about to order my mindsp.
@perrymarshall - if I go with the Flex 8, I have the option to get Dirac Live room correction, but then lose the FIR. Can't have both FIR control and Dirac. According to Minidsp, the Dirac is a more powerful tool that will yield better results than FIR alone. I know you used the older 2x4HD with your project. If you were updating the design would you go for Dirac or stay with the FIR alone. Thanks!
@perrymarshall - if I go with the Flex 8, I have the option to get Dirac Live room correction, but then lose the FIR. Can't have both FIR control and Dirac. According to Minidsp, the Dirac is a more powerful tool that will yield better results than FIR alone. I know you used the older 2x4HD with your project. If you were updating the design would you go for Dirac or stay with the FIR alone. Thanks!
Either way is fine. I'm a little bit of a control freak so I like the DIY aspect of rolling my own FIR. But for most people I think Dirac would be easier and better. Good option.
thanks all. I think I'll go Dirac just because I have some room issues it will probably help with and....I'm likely to do more harm than good brewing my own files.
Looking at the schematic, I see the following passive parts for the x-overs, and would love input on parts selection:
Woofers:
360uF caps - I can't find any single cap in this value, should I use multiple caps in parallel? And am I reading the schematic correctly, it seems like a quite high cap value. Any suggestion on the part(s) to use?
6mH inductors - thinking Jantzen air core https://www.parts-express.com/Jantzen-1958-6.0mH-18-AWG-Air-Core-Inductor-255-286?quantity=1 or https://www.madisoundspeakerstore.c...ay/solen-6.0-mh-perfect-lay-inductors-14-awg/ which has a much lower DCR (.68 DCR vs 1.45 DCR for the Jantzen) - Any preference?
Tweeters
80uF caps - 82uF +/- 5% seems ok? https://www.madisoundspeakerstore.com/px-cap-250vdc/claritycap-82-mfd-px-range-polypropylene-caps/
2.7uF caps - clarity cap https://www.madisoundspeakerstore.c...ycap-2.7-mfd-csa-copperconnect-poly-cap-250v/
5R - thinking of a 2R and 3R in series from this line? https://www.parts-express.com/Dayto...ision-1-Audio-Grade-Resistor-007-2?quantity=1
.5Mh inductors - thinking https://www.madisoundspeakerstore.com/goertz-12-awg/goertz-cf.5-12-awg-copper-foil-inductor/ (DCR .075)
Any other part suggestions that you would make or anything I am missing from the schematic? Thank you!
Looking at the schematic, I see the following passive parts for the x-overs, and would love input on parts selection:
Woofers:
360uF caps - I can't find any single cap in this value, should I use multiple caps in parallel? And am I reading the schematic correctly, it seems like a quite high cap value. Any suggestion on the part(s) to use?
6mH inductors - thinking Jantzen air core https://www.parts-express.com/Jantzen-1958-6.0mH-18-AWG-Air-Core-Inductor-255-286?quantity=1 or https://www.madisoundspeakerstore.c...ay/solen-6.0-mh-perfect-lay-inductors-14-awg/ which has a much lower DCR (.68 DCR vs 1.45 DCR for the Jantzen) - Any preference?
Tweeters
80uF caps - 82uF +/- 5% seems ok? https://www.madisoundspeakerstore.com/px-cap-250vdc/claritycap-82-mfd-px-range-polypropylene-caps/
2.7uF caps - clarity cap https://www.madisoundspeakerstore.c...ycap-2.7-mfd-csa-copperconnect-poly-cap-250v/
5R - thinking of a 2R and 3R in series from this line? https://www.parts-express.com/Dayto...ision-1-Audio-Grade-Resistor-007-2?quantity=1
.5Mh inductors - thinking https://www.madisoundspeakerstore.com/goertz-12-awg/goertz-cf.5-12-awg-copper-foil-inductor/ (DCR .075)
Any other part suggestions that you would make or anything I am missing from the schematic? Thank you!
Wire capacitors in parallel to add up to 360uF.
The inductors should have less than 1 ohm DCR.
82uF is fine; any of these xover values within +/-5% will work.
You can wire resistors in series as you describe.
The 0.5mH inductors are just for a notch filter on the rear tweeters. Any average quality inductor will do just fine, don't need exotic or expensive parts for that.
The inductors should have less than 1 ohm DCR.
82uF is fine; any of these xover values within +/-5% will work.
You can wire resistors in series as you describe.
The 0.5mH inductors are just for a notch filter on the rear tweeters. Any average quality inductor will do just fine, don't need exotic or expensive parts for that.
Thank you so much. I'm 100% not an EE so it takes me a while to see what you might see as obvious (like the parallel caps) lol. Thanks again, will order parts!
One more thing - air core vs iron core. HiFi Compass did VERY interesting distortion tests on iron vs. air core inductors here:
https://hificompass.com/en/projects/experiment/test-couple-iron-core-inductors
Conclusion:
"All this says about the iron cored coils like Sledgehammer Steel Laminate AWG15 and Mundorf MCoil FERON can certainly be used in LF circuits of high quality crossovers of loudspeakers, which incorporate the most linear LF speakers, without any risk to spoil the sound quality."
So I'm not convinced you need high priced air core coils. I do use expensive air core foil inductors in these designs. The 6mH inductor in the Bitches Brew has DCR=0.6 ohm. But there are some good alternatives.
Low DCR is especially important because it's a series crossover and the higher the DCR, the more bass leaks into the midrange driver. I've run simulations and getting much below 0.3 ohm doesn't buy you that much extra performance. But you definitely don't want it to be much higher than 0.5 Ohm.
US Coils has super-low DCR iron core inductors, for example this 5.5mH inductor has 0.115 ohm DCR:
https://uscoils.us/product/ussq55-12-5000/
At $39 it's an excellent choice.
@JMacII
https://hificompass.com/en/projects/experiment/test-couple-iron-core-inductors
Conclusion:
"All this says about the iron cored coils like Sledgehammer Steel Laminate AWG15 and Mundorf MCoil FERON can certainly be used in LF circuits of high quality crossovers of loudspeakers, which incorporate the most linear LF speakers, without any risk to spoil the sound quality."
So I'm not convinced you need high priced air core coils. I do use expensive air core foil inductors in these designs. The 6mH inductor in the Bitches Brew has DCR=0.6 ohm. But there are some good alternatives.
Low DCR is especially important because it's a series crossover and the higher the DCR, the more bass leaks into the midrange driver. I've run simulations and getting much below 0.3 ohm doesn't buy you that much extra performance. But you definitely don't want it to be much higher than 0.5 Ohm.
US Coils has super-low DCR iron core inductors, for example this 5.5mH inductor has 0.115 ohm DCR:
https://uscoils.us/product/ussq55-12-5000/
At $39 it's an excellent choice.
@JMacII
Last edited:
@perrymarshall
Would this series of caps be ok for the 360uF across the woofers? https://www.parts-express.com/330uF-100V-Non-Polarized-Capacitor-027-374?quantity=1
I can put one of these in parallel with a 30uF. I'm not sure how critical the woofer caps are to the overall sound. These large cap values from Jentzen et al will be $100s. Thoughts?
/edit - also thanks so much for the link to US Coils. I'm getting the 6mH from them. I'm going to add the 7mH z-pad for the tweeters that you posted on the other thread but using a Jentzen for that which is out of the signal path and using the Jentzen for the .5mH too which is also low DCR at .34R. The 7mH is 1.6 DCR but the US coils is the best at .189 DCR for 6mH
Would this series of caps be ok for the 360uF across the woofers? https://www.parts-express.com/330uF-100V-Non-Polarized-Capacitor-027-374?quantity=1
I can put one of these in parallel with a 30uF. I'm not sure how critical the woofer caps are to the overall sound. These large cap values from Jentzen et al will be $100s. Thoughts?
/edit - also thanks so much for the link to US Coils. I'm getting the 6mH from them. I'm going to add the 7mH z-pad for the tweeters that you posted on the other thread but using a Jentzen for that which is out of the signal path and using the Jentzen for the .5mH too which is also low DCR at .34R. The 7mH is 1.6 DCR but the US coils is the best at .189 DCR for 6mH
Yep 330 + 30 works great.
I’m not sure how critical they are either. If you are suspicious of electrolytics you can do 300uf electrolytic + 60uf film and save some dinero.
This vendor sells good 100uF film caps which I have been happy with for $15.
https://www.ebay.com/itm/2958816928...ar=594087634905&widget_ver=artemis&media=COPY
I’m not sure how critical they are either. If you are suspicious of electrolytics you can do 300uf electrolytic + 60uf film and save some dinero.
This vendor sells good 100uF film caps which I have been happy with for $15.
https://www.ebay.com/itm/2958816928...ar=594087634905&widget_ver=artemis&media=COPY
Hi Perry - I've got all my wood cut and will be putting this together this weekend and I'm off work next week so I've got time to play. I've also got the miniDSP Flex8 Dirac edition. As you noted in the build notes, and as I've discovered, the 2x4HD file will not load directly into the Flex8. However, I can load the 2x4HD in Device Counsel template and see the settings. I plan to copy over the PEQ settings manually. Questions/confirmation:
B&C Tweeter section:
B&C Tweeter section:
- The high-pass for the tweeters is 300hz 6 db/octave Butterworth. That seemed low for x-over of the tweeters and I wanted to confirm? The notes say "The tweeters have 80uF capacitors to prevent damage. They roll off below 300Hz so they do not effectively factor into the crossover design" so is the 80uF cap really doing the work here vs the DSP for cross over purposes?
- Gain -8...is that where you would adjust for the different gain of your amps if you are using two different amps for the high and low F? I assume so. I plan to run the tweeters off my Aleph J and use an old McIntosh amp I have for the bass sections.
- The "crossover" section in the file seems incomplete. What do you recommend for the low pass? I know the midrange coax is on this circuit too and it needs to go up to at least 1100 so I'm not sure how to set this up. Maybe this was done in your design with the FIR filters which I can't use w/ Dirac.
- The notes say "6dB/octave PASSIVE series crossover between the subwoofers and the midwoofer. It is series instead of parallel because a series crossover has much cleaner interactions with impedance peaks of the woofers." - looking in the schematic it looks to me like the inductor is in parallel with the woofer not series? I am not an expert in reading these schematics however, so I wasn't sure how if I am reading this correctly and wanted to confirm. Should I wire the 6mH inductor in series with the mid-woofer or in parallel across the mid-woofer? And I guess same question for the 360uF cap, I was reading this as wiring the two woofers in parallel with the 360uF cap in parallel as well, then wiring the mid-woofer in series with the two subwoofers, with the inductor in parallel across the -/+ of the mid woofer.
MiniDSP Flex Eight files are already provided here: https://tinyurl.com/bitches-brew-dsp-flex8 You don't need to copy over from 2x4HD.
B&C Tweeter section:
In the former flex 8 version I used shelf filters for most of the tweeter EQ and added a 300hz high pass for protection.
But in the later version I use a 6Khz 6dB high pass which compensates for the downward tilt of the compression horn response. Then it is Combined with some shelf filters around 1000Hz. I think it’s a more elegant solution. The effective crossover is still around 1400Hz as I recall. And yes there is an 80uF capacitor in series with the tweeters which protects them from accidents. All of the real "crossover" work is done by the shelf filters combined with the 6dB filter at 6KHz.
Yes, you adjust the gain sections based on whatever combo of amps you're using, which may also have their own level controls.
SB Woofer:
There is no "crossover slope" used in the woofer section of the Flex Eight at all. It is all done with shelf and peak filters in the PEQ section, combined with the passive SERIES crossover between the two 15" subs and the B&C 15" mid. Wire it exactly as shown in the diagram, 6mH inductor in parallel with the B&C Mid; that section is wired in series with the 2 Subs which have a 360uF capacitor in parallel with them.
I used very little conventional crossover filters and structures in this design. The series crossover between the 15s, and the use of shelf filters instead of standard crossover slopes all minimizes phase shift.
If you're using the FLEX version that has no FIR provisions, I assume you're using the DIRAC version instead. Tweak the shelf filters in the existing design as necessary to get it to flat, then apply DIRAC after everything else is done and you should be OK.
Really looking forward to hearing your report and pictures.
B&C Tweeter section:
In the former flex 8 version I used shelf filters for most of the tweeter EQ and added a 300hz high pass for protection.
But in the later version I use a 6Khz 6dB high pass which compensates for the downward tilt of the compression horn response. Then it is Combined with some shelf filters around 1000Hz. I think it’s a more elegant solution. The effective crossover is still around 1400Hz as I recall. And yes there is an 80uF capacitor in series with the tweeters which protects them from accidents. All of the real "crossover" work is done by the shelf filters combined with the 6dB filter at 6KHz.
Yes, you adjust the gain sections based on whatever combo of amps you're using, which may also have their own level controls.
SB Woofer:
There is no "crossover slope" used in the woofer section of the Flex Eight at all. It is all done with shelf and peak filters in the PEQ section, combined with the passive SERIES crossover between the two 15" subs and the B&C 15" mid. Wire it exactly as shown in the diagram, 6mH inductor in parallel with the B&C Mid; that section is wired in series with the 2 Subs which have a 360uF capacitor in parallel with them.
I used very little conventional crossover filters and structures in this design. The series crossover between the 15s, and the use of shelf filters instead of standard crossover slopes all minimizes phase shift.
If you're using the FLEX version that has no FIR provisions, I assume you're using the DIRAC version instead. Tweak the shelf filters in the existing design as necessary to get it to flat, then apply DIRAC after everything else is done and you should be OK.
Really looking forward to hearing your report and pictures.
Last edited:
I have one built, but not wired. More to come….
Do you still recommend adding those little impedance balancing parallel circuits to the x-overs?
They make it an easier load to drive. Yes, I do like them. Totally optional (and I can't prove they make it sound better).
Since you're doing a fully active 3 way instead of 2 way, make sure you configure the DSP crossover according to this tech note:
I’m building the standard two way with the passive mid range cross over so following the original schematic plus the impedance parallel circuits. Should be playing music tomorrow!
I finally got some time last night to sit and listen to these. I’m just using the base DSP files, have not tuned anything yet and have not run the Dirac setup. Even so, these are the best speakers I have ever heard. I was particularly struck by how much detail they have even at low volume, imaging is fantastic, and the bass is really impressive. Not just the bass output, but the tonality and texture of the bass is better than anything I’ve ever owned. I’ll post more detailed thoughts after I spend more time with these. But even now I can say they were absolutely worth the effort and cost.