The good driver I have is like Zaph's measurement, as shown here:
And the bad one has the bass distortion hump centered at 80Hz that is twice as high.
similar project
Very interesting thread.
I have a pair of oak cabinets that I made a few years ago for some Royd drivers. size is 390cu".
After reading your thread and other comments about the AC130F1, particularly the Continuum positives, I decided to assemble a pair with some Dayton ND16 tweeters. Crossover is presently LR4 at 3500Hz and sound really, really good.
The AC130 is super smooth up to 5khz, and then easily managed with a minidsp. Sealed and stuffed with Polyfill.
Probably the best bookshelf size speaker I have built. I will also try the FR58 and XT25 and maybe, if Santa is kind enough a pair of RS28A.
Thanks for the inspiration.
Peter
Very interesting thread.
I have a pair of oak cabinets that I made a few years ago for some Royd drivers. size is 390cu".
After reading your thread and other comments about the AC130F1, particularly the Continuum positives, I decided to assemble a pair with some Dayton ND16 tweeters. Crossover is presently LR4 at 3500Hz and sound really, really good.
The AC130 is super smooth up to 5khz, and then easily managed with a minidsp. Sealed and stuffed with Polyfill.
Probably the best bookshelf size speaker I have built. I will also try the FR58 and XT25 and maybe, if Santa is kind enough a pair of RS28A.
Thanks for the inspiration.
Peter
Attachments
Nice work! Thanks for sharing. I have used the ND25FA tweeter and it would probably be good match too for a 3.5k XO. As long as you have a minidsp - try the Harsch XO for quasi transient perfect. Set a target curve of BW4 for woofer and Bessel 2 for tweeter. Set tweeter delay equal to half period of XO frequency.
Glad the Harsch worked out. There is a small delay offset due to the location of the voice coil of the woofer relative to tweeter. This will be about 1in behind tweeter. That's about 74micro seconds - so you may have to add that to the textbook Harsch delay. If you measure the response you can see that the combined response dips a little below the woofer only response - that is correct. Also the IR and Step Response will look good.
Stereo Pair Completed
I finally had some time to put together the stereo pair for this speaker. Some motivation was provided by seeing Pjblakey make his nice pair. Did not take long at all and glad I did. They image really well and the sound stage is super realistic. I may move them to my computer table next as near field monitors.
The stereo pair:
Here is the measurement of the response and phase on unit 2:
with 5ms gating - phase is about +/- 15deg from 350hz to 14kHz:
IR and SR:
Mondogen had no faith in me for making a stereo pair...
I finally had some time to put together the stereo pair for this speaker. Some motivation was provided by seeing Pjblakey make his nice pair. Did not take long at all and glad I did. They image really well and the sound stage is super realistic. I may move them to my computer table next as near field monitors.
The stereo pair:
Here is the measurement of the response and phase on unit 2:
with 5ms gating - phase is about +/- 15deg from 350hz to 14kHz:
IR and SR:
Mondogen had no faith in me for making a stereo pair...
Attachments
Last edited:
I have a pair of DX25 and XT25 that are currently doing duty in a couple of other projects.
Santa will hopefully bring the RS28A.
In the meantime I'll play with the FR58. I used them for a while in my OB speaker and was impressed.
BTW, I tried out the Harsch XO on a pair of small speakers with RS180 and waveguide XT25 crossing at 1800hz. Got rid of some "honkiness" that I noticed previously.
Thanks for all the info.
Peter
Santa will hopefully bring the RS28A.
In the meantime I'll play with the FR58. I used them for a while in my OB speaker and was impressed.
BTW, I tried out the Harsch XO on a pair of small speakers with RS180 and waveguide XT25 crossing at 1800hz. Got rid of some "honkiness" that I noticed previously.
Thanks for all the info.
Peter
The RS28A will be back in stock this week. I have a rather nice walnut board that is air drying that should be ready for planing in a month or so. I am thinking about a Continuum size cabinet with a leather covered baffle. Does leather offer any sonic benefits, say similar to scotchbrite or felt? Maybe insufficient absorbent qualities.
Leather is more absorbent than hard wood, but the felt/scotchbrite works by having "fuzz" that breaks up the diffraction.
I just moved my little monitors to my PC desk. I had to EQ out some desk bounce with a -4dB Q=0.7 at 300Hz now it is flat (with some reflection hash from keyboard, lamp, etc) at listening position. Sounds very nice - precise like mix monitors.
This is going to tie up one of my miniDSP 2x4's at the desk, but well worth it. Time to buy another unit.
I am listening to Buddy Guy and Tracy Chapman singing "Ain't no sunshine" - wow. Bass with desk enhancement now reaches 35Hz.
I just moved my little monitors to my PC desk. I had to EQ out some desk bounce with a -4dB Q=0.7 at 300Hz now it is flat (with some reflection hash from keyboard, lamp, etc) at listening position. Sounds very nice - precise like mix monitors.
This is going to tie up one of my miniDSP 2x4's at the desk, but well worth it. Time to buy another unit.
I am listening to Buddy Guy and Tracy Chapman singing "Ain't no sunshine" - wow. Bass with desk enhancement now reaches 35Hz.
Good question - basically go for a FAST rather than a 1.3kHz more conventional 2 way XO. I don't notice an issue with not having an unified point source right now. I think the overlap of the BW1 XO is so wide it's not readily apparent when one switches on and the other off. I might give it a try, but will just enjoy listening to current setup for a while at the desk. I never had computer speakers sound this good.
Welcome and great thanks sharing build.
Because asked at all is as said FR58EX can do it verse a real tweeter and because you have nice fast configurable tool miniDSP.
Picture 1 is vertical lobe at a single point frq (same as XO point) ctc spacing is set 106mm:
1. BW1 woofer/tweeter XO at 1300Hz.
2. BW1 woofer/tweeter/woofer XO at 1300Hz.
3. HARSH woofer/tweeter XO at 1300Hz.
4. HARSH woofer/tweeter XO at 600Hz.
First plot suggests if you turn speaker upside down should make some difference.
Picture 2 is not only at single point frq but also at 2 octaves below and above frq point. While all three configs show perfect same spl being 100% on axis and thereby linear phase, ctc spacing do make frq dependant lobe sectors. This was fun study and maybe show why Dunlavy use W/T/W config for his BW1 based speakers.
Because asked at all is as said FR58EX can do it verse a real tweeter and because you have nice fast configurable tool miniDSP.
Picture 1 is vertical lobe at a single point frq (same as XO point) ctc spacing is set 106mm:
1. BW1 woofer/tweeter XO at 1300Hz.
2. BW1 woofer/tweeter/woofer XO at 1300Hz.
3. HARSH woofer/tweeter XO at 1300Hz.
4. HARSH woofer/tweeter XO at 600Hz.
First plot suggests if you turn speaker upside down should make some difference.
Picture 2 is not only at single point frq but also at 2 octaves below and above frq point. While all three configs show perfect same spl being 100% on axis and thereby linear phase, ctc spacing do make frq dependant lobe sectors. This was fun study and maybe show why Dunlavy use W/T/W config for his BW1 based speakers.
Attachments
Last edited:
Nice work Byrtt. How many monitors are out there that are two ways with a tweeter on top and XO in the 1.5k to 3k range? Just about all of them - and this suggests they have vertical lobing issues. I guess people don't know what they don't know. A coaxial is perhaps one way to really clean this up.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.