Thanks for the above.
Have you settled on a baffle and woofer placement? Can you please give me baffle dimensions, woofer driver size (piston area aka Sd will do) and woofer centre height on the baffle?
Have you settled on a baffle and woofer placement? Can you please give me baffle dimensions, woofer driver size (piston area aka Sd will do) and woofer centre height on the baffle?
Are you able to post the 1 meter woofer (farfield) response as the raw REW MDAT file ? (rename with a .txt). There are room reflections in the measurement and without the impulse data (plan old FRD / TXT is only amplitude and phase) I am unable to remove the reflections.
One more request - both port and nearfield response only go down to 20Hz. I need 10Hz given your tuning to align port and woofer nearfield outputs to sum. Maybe just post the raw MDAT measurement files? (if too big - have you got a google drive?)
Id have to retake the measurements, I only did 20 to 20. I'll get baffle size, driver placement, and SD tomorrow. Driver is dayton sd315a.
If they are the same measurements, a gate of 3,6ms does fine. For port tuning, look at the phase, not at the amplitude of both peaks.
Here's the mdat of the woofer from 0-20khz.
https://drive.google.com/file/d/19J2lSmlA6r0536b7eDLZyM63pEEXkEQg/view?usp=sharing
I managed to bang out one of the ports that I had cut, it was going to either need to be replaced or just removed entirely depending on what happens with the 180~hz thing I was hearing. That will let me play with porting as well. Kinda surprised my port is tuned so low on these, I guess my math was off, looks like it's tuned to 27hz, I was getting 34hz but oh well.
Trying a bunch of other dampening to see if I can deal with the impedance blip at ~160-170 but my impedance rig has decided to just not work right anymore. Not really sure what's up but REW won't calibrate with it, just used it yesterday.
https://drive.google.com/file/d/19J2lSmlA6r0536b7eDLZyM63pEEXkEQg/view?usp=sharing
I managed to bang out one of the ports that I had cut, it was going to either need to be replaced or just removed entirely depending on what happens with the 180~hz thing I was hearing. That will let me play with porting as well. Kinda surprised my port is tuned so low on these, I guess my math was off, looks like it's tuned to 27hz, I was getting 34hz but oh well.
Trying a bunch of other dampening to see if I can deal with the impedance blip at ~160-170 but my impedance rig has decided to just not work right anymore. Not really sure what's up but REW won't calibrate with it, just used it yesterday.
Whew, got it working again, had to rewire some bits.
Thanks Dave for your help on the driver data splicing.
For funsies I measured my Amiga's, they have a similar bump at basically the same spot, they have similar internal heights. The port is on the rear though, not sure how much that contributes to what I was hearing. My big speakers impedence peak is quite narrow while the Amiga impedence peak is more like a little hump. Just some observations made out of boredom.
Thanks Dave for your help on the driver data splicing.
For funsies I measured my Amiga's, they have a similar bump at basically the same spot, they have similar internal heights. The port is on the rear though, not sure how much that contributes to what I was hearing. My big speakers impedence peak is quite narrow while the Amiga impedence peak is more like a little hump. Just some observations made out of boredom.
I think I made some progress on narrowing down the impedence blip down low. sitting on the side of the cabinet helps a lot, as does pressing down on the baffle. I think I can rule out the port being a problem.
not really sure how to tackle this problem, no more room for braces really. Probably better off with a new cabinet. I look at some large-ish 3 ways like the JBL l100 classic and I see no problems in impedance/resonances. Those just have a single V brace to the baffle, crazy.
not really sure how to tackle this problem, no more room for braces really. Probably better off with a new cabinet. I look at some large-ish 3 ways like the JBL l100 classic and I see no problems in impedance/resonances. Those just have a single V brace to the baffle, crazy.
Last edited:
Sorry I had a busy day yesterday. I'll work on the MDAT files today.
Re impedance bumps - I often use F clamps with pieces of wood across suspect panels to see if this changes things (indicating another brace may be required).
I've found lining panels with gym floor rubber tile helps (reasonably cheap here for a 1sqm which does a heap of speakers... also useful for resoling motorcycle boots 😀). This stuff is not "open cell" so will reduce cabinet volume so take that into account
I assume you are doing the impedance sweep under power if you are hearing farting type noises.... this indicates to me leakage around the driver mount point, or possibly leaky port fitting.
Re impedance bumps - I often use F clamps with pieces of wood across suspect panels to see if this changes things (indicating another brace may be required).
I've found lining panels with gym floor rubber tile helps (reasonably cheap here for a 1sqm which does a heap of speakers... also useful for resoling motorcycle boots 😀). This stuff is not "open cell" so will reduce cabinet volume so take that into account
I assume you are doing the impedance sweep under power if you are hearing farting type noises.... this indicates to me leakage around the driver mount point, or possibly leaky port fitting.
It wasn't any of this, there was an issue with the cabling somewhere. It was very obviously a signal problem and not a cabinet related thing.
Can I please have the baffle size and placement / location of the SD315A on the baffle?
Thanks - I've assumed your tweeter height (measurement position) as shown. This is using VituixCAD. Blue line is the baffle step and diffraction response.
I'll add a series of posts showing what I've done, mainly for others to critique incase I get something wrong (quite possible)
I'll add a series of posts showing what I've done, mainly for others to critique incase I get something wrong (quite possible)
Thank you Dave, if you do share steps I will surely keep them bookmarked for future projects. I appreciate you taking your free time to help.
Ok - the above mic position post is wrong. What we want is an on-axis quasi anechoic response of woofer with baffle step / diffraction applied THEN merged with the port response. The reason we want on-axis, is you then apply a Y offset on the driver in the crossover simulation in tools like VituixCAD which will then apply a high frequency roll off for the off-axis relative to mic now on the tweeter axis for design.
Here's the steps I took. I've put all the data in the MDAT on my drive so you can download.
I then loaded the Dayton factory curve for the SD315A and scaled the response in 6. above to match so you get a 1w/1m response level. Make sure you rescale this response to match midrange and tweeter when you measured them. The absolute SPL does not matter - it's the relative SPL that counts for measuring for crossover design.
https://drive.google.com/file/d/1dVhIHKn3MO7D5RPmdkfMB31x6Y6PLkSs/view?usp=sharing
For others benefit - you can see the Dayton factory response (blue curve) with my quasi-anechoic "baffle step loss" response "purple" merged with gated farfield.
Here's the steps I took. I've put all the data in the MDAT on my drive so you can download.
- I applied a 3.64msec gate to your 1 meter woofer (farfield response) to remove reflections
- I then brought down in SPL the nearfield port and woofer responses to match the 1 meter response
- I loaded the woofer diffraction response - with mic taken on woofer axis (i.e. woofer centre)
- I then merged the woofer NF response (without port) with the baffle diffraction response (this is A times B in REW - check the notes for the measurements as REW kindly captures the arithmetic). We now have a quasi anechoic woofer response - that is what we should expect for a ground plane measurement
- I then merged / spliced the farfield (1 meter woofer) with Woofer NF and BSC response = our full range response for the woofer
- Finally - I merged the port response to the woofer response to get the combined in box low frequency response.
I then loaded the Dayton factory curve for the SD315A and scaled the response in 6. above to match so you get a 1w/1m response level. Make sure you rescale this response to match midrange and tweeter when you measured them. The absolute SPL does not matter - it's the relative SPL that counts for measuring for crossover design.
https://drive.google.com/file/d/1dVhIHKn3MO7D5RPmdkfMB31x6Y6PLkSs/view?usp=sharing
For others benefit - you can see the Dayton factory response (blue curve) with my quasi-anechoic "baffle step loss" response "purple" merged with gated farfield.
I'm no expert, but with VituixCAD Merger Tool you can take the Port nearfield and Woofer nearfield into account at the same time.
- You don't need to do the AxB to merge the baffle step loss as it can be taken into account if you have saved the diffraction model and then opened it up using the little folder icon. In the screenshot note that the Diffraction Response radio button is selected and I opened the file "wafflesomd woofer diffraction.txt" that I had save. (Note, it is not the same as he posted a few posts ago, I had used the Pit Viper spacing.
- In the top section "Low Frequency Part" there is a blue check box that you can apply BSC from the diffraction model to the woofer AND to the port if it is front-ported, which it is in this case. Thus both nearfield files are opened and I have selected the BS radio button for both.
A4eaudio - did you scale down your port response to match the woofer nearfield? I believe one technique is to align output at the tail (i.e. between 10 and 20 Hz) - otherwise you might be over emphasizing the port contribution in your response:
Edit: having said the above - I'm doubting my response as the F3 appears to be only 47Hz - which seems way too high for a ported SD315A
Edit: having said the above - I'm doubting my response as the F3 appears to be only 47Hz - which seems way too high for a ported SD315A
Last edited:
wafflesomd... what was your woofer enclosure volume and port dimensions please? (diameter and length) and number of ports?
No, but only because I'm pretty sure VituixCAD does it automatically "somehow". I do know what you are referring to, from using Jeff Bagby's response blend and his white paper.
When I put both the port and woofer nearfield in the top section "Low Frequency Part" of VituixCAD, they are merged in the lower portion where you adjust the SPL to merge. You can check the "show nearfield" box and it does show the woofer and port lines separately. Also, checking and unchecking the "BS" box on port file shows the effect of applying the baffle diffraction to the port.
As I said, I'm no expert so may not be completely correct, just pointing out some possibilities that I'm aware of.