@fluid thanks for that info. This matches previous driver/port measurements (near field… see previous graphs posted up above somewhere).
So if the impedance looks good. The. I just need to collect cleaner responses of the drivers. Not too happy with the cleanliness of the impulse. I think I will need to pick up an amp better than my old hitachi.
So if the impedance looks good. The. I just need to collect cleaner responses of the drivers. Not too happy with the cleanliness of the impulse. I think I will need to pick up an amp better than my old hitachi.
The impulses in post #45 look quite suspect to me, all over the place in the first millisecond which is strange. You may want to look at doing some loopback measurements of the measurement chain to see if it is clean. Be careful measuring a power amp, very easy to blow up a soundcard doing that without an attenuator or zener clamp circuit.
Pretty sure you can save a cal with a loopback through the amp that will remove all issues with almost any amp.
I also move the mic stand as far to the right or left and use the top bar to come back to the center. That puts the stand out of the 0 deg. 15 deg, and possibly 30 deg. measurments.
Calibration files in REW are applied to the SPL and Phase window not the impulse response. A corrupted impulse response won't be fixed in the displays that rely directly on the impulse response.
Taking care of what happens behind the mic including the clip/mounting and stand is important.
An old B&K image shows some of the effects
What is the optimal microphone stand/holder configuration?
I agree. The impulses measured for the drivers were not very clean. I need to redo them.
I agree. The impulses measured for the drivers were not very clean. I need to redo them.
A long tube that extends directly behind the mic with some kind of adapter to smooth the transition from the mic body to the tube.
Couple of good examples here
https://www.diyaudio.com/community/threads/purifi-waveguide-project.394174/post-7383721
At some point I will build a subwoofer. Is there an amp I can use for measurement now, which will then be suitable to use as the amp for the subwoofer, when I finally get to that stage?
I have tried my old Hitachi, and a circa 1990's Onkyo...both seem to have really dirty results in the impulse (even at the very low frequency range, and even at near field measurements).
I am also contemplating sending me microphone out to be calibrated (it is a Umik-1, with factory calibration).
@fluid I fixed my microphone stand, and now use the microphone to pipe method. Pipe length is just shy of a meter (I am measuring at/around 22" for far field.
There are many, depends on your budget and how much power you think you will need. There are some quite well measuring Chinese Class D chipamp amplifier from Topping, Aiyima, Fosi, SMSL, Allo etc. These will be more than good enough for measurements but may not have enough power. Hypex plate amps are good, Icepower modules, the list could go on.
The UMIK will ultimately be a limiting factor in taking time based measurements which are really needed for modern speaker design. If I was you I would buy another calibrated analogue microphone like the Sonarworks XREF or similar and an audio interface with decent mic preamp. There are other ways but I would recommend you to save the headache and do it right from the start if you can afford to.
The Pipe will certainly help but 0.5m is not going to be in the far field of very many speakers. Usually you want to be measuring at least 3 to 4 times the longest dimension of the cabinet if you can.
0,5m gives you about 3ms delay, while in most domestic measurement environments you can reach 4 to 5ms, which should be enough for two way speakers with crossovers above 1kHz. I use a rod of about 3’, that not only enlarges the delay but also weakens the delayed reflection signal.
I would not calibrate the Umik. Most if not all of these mikes are perfectly flat from 100Hz up to 4-5kHz. As fluid stated, rather save for a discrete setup.
I would not calibrate the Umik. Most if not all of these mikes are perfectly flat from 100Hz up to 4-5kHz. As fluid stated, rather save for a discrete setup.
For amplifiers, this is one of the less expensive turn-key options
https://www.parts-express.com/FX-Au...eo-Amplifier-160-WPC-Black-230-281?quantity=1
Performance is good, noise and distortion is reasonably low, frequency response is flat, and it comes with a volume control knob to control the gain. It is advertised as 160 WPC, but that is very optimistic. With the 32V power supply (included), it is more like 110 WPC. Still, for $83 that is a bargain.
A step up in performance would be Ice power board amps. An ICEpower 200ASC will give you 1 channel of 200W for about $130. It is not turn-key, you will need to supply chassis and connectors, and probably a power cord.
PVC water pipe makes a great microphone tube, and it is very inexpensive. You should have no trouble finding a pipe with an internal diameter that matches your mic, usually 3/4", 7/8", or 1". Use sandpaper to round the edges where the mic sits so there is no flat reflecting surface.
https://www.parts-express.com/FX-Au...eo-Amplifier-160-WPC-Black-230-281?quantity=1
Performance is good, noise and distortion is reasonably low, frequency response is flat, and it comes with a volume control knob to control the gain. It is advertised as 160 WPC, but that is very optimistic. With the 32V power supply (included), it is more like 110 WPC. Still, for $83 that is a bargain.
A step up in performance would be Ice power board amps. An ICEpower 200ASC will give you 1 channel of 200W for about $130. It is not turn-key, you will need to supply chassis and connectors, and probably a power cord.
PVC water pipe makes a great microphone tube, and it is very inexpensive. You should have no trouble finding a pipe with an internal diameter that matches your mic, usually 3/4", 7/8", or 1". Use sandpaper to round the edges where the mic sits so there is no flat reflecting surface.
@hifijim I used 3/4" PVC pipe (for use with electrical wires) that I had from a wiring project. Here a few pictures. The distance from the vertical part of the stand to the tip of the microphone is 46" (1.16m). the USB wire is snaked through the tube and exits at the hole I drilled before attachment to the stand. Luckily the stand handle had a diameter that fit perfectly into the tube, so I got lucky. The microphone level goes up to about 5'-5" from the ground (incase you were wondering).
I also just ordered a Fosi Audio V3 amp so I have something clean to rule out potential reasons for dirty impulses. Even when making the near field measurements with this new setup, I could see a dirty impulse, which makes me believe it is a dirty signal.
Happy Thanksgiving!
I also just ordered a Fosi Audio V3 amp so I have something clean to rule out potential reasons for dirty impulses. Even when making the near field measurements with this new setup, I could see a dirty impulse, which makes me believe it is a dirty signal.
Happy Thanksgiving!
I went outside and collected some measurements. Here are the Impulses and responses for tweeter and woofer. (Woofer has the port and near field combined, baffle diffraction adjustment, and SPL level adjusted to merge with the far field measurement).
I used the upgraded pipe microphone method for collection. I also use a Fosi Audio V3 amplifier.
for VituixCad I included my ZMA measurements.
I am not sure what is going on with the Morel tweeter in the higher frequency range... It could be the waveguide I have it mounted in (which is really just a guesstimate of a waveguide design based on some other waveguide designs). Should I put some some of filter at these frequencies to fix this? it is quite a drop off in spl (10db or so).
Here is my six-pack and crossover design. Any suggestions?
I used the upgraded pipe microphone method for collection. I also use a Fosi Audio V3 amplifier.
for VituixCad I included my ZMA measurements.
I am not sure what is going on with the Morel tweeter in the higher frequency range... It could be the waveguide I have it mounted in (which is really just a guesstimate of a waveguide design based on some other waveguide designs). Should I put some some of filter at these frequencies to fix this? it is quite a drop off in spl (10db or so).
Here is my six-pack and crossover design. Any suggestions?
Sorry to be blunt, but either your tweeter measurements are seriosly flawed, or your waveguide is really bad. No sensible tweeter, in WG or not, should have this +-6 dB sawtooth response.
Maybe do a sanity check measurement with a regular tweeter mounted on a flat baffle to verify your measurement setup.
If the waveguide is to blame, why not use one of the (excellent) @augerpro designs?
https://www.somasonus.net/waveguides
https://www.diyaudio.com/community/threads/open-source-waveguides-for-cnc-3d-printing.318190/
With crossover, general trend should be downward sloping ER/Power Response, so tweeter level should not be higher than low/mid level. A dip in axial response is preferable to a bump in ER/PR. E.g., variant 1 here:
is usually preferable to variant 2:
Anyway, I must add, it's great to see new users designing speakers with a set of off-axis measurements and modern multi-axial software, so keep up the good work!
Maybe do a sanity check measurement with a regular tweeter mounted on a flat baffle to verify your measurement setup.
If the waveguide is to blame, why not use one of the (excellent) @augerpro designs?
https://www.somasonus.net/waveguides
https://www.diyaudio.com/community/threads/open-source-waveguides-for-cnc-3d-printing.318190/
With crossover, general trend should be downward sloping ER/Power Response, so tweeter level should not be higher than low/mid level. A dip in axial response is preferable to a bump in ER/PR. E.g., variant 1 here:
is usually preferable to variant 2:
Anyway, I must add, it's great to see new users designing speakers with a set of off-axis measurements and modern multi-axial software, so keep up the good work!
@VoxCelestial thank you for the feedback. I used Augerpro's WG as the starting point for this Morel (granted, I don't have the know-how to run the full physics analysis to arrive at the shape, but rather just followed the ratios he discusses). I think I really need a flat baffle response at this point to see if that tweeter is the issue or if it is the WG. I am strongly considering selling these Morels and just buying the SB Acoustic ones from the build I am trying to replicate.
Try the waveguides from either the Scanspeak or Satori TW29TX tweeters at my website. Whichever fits the diaphragm better. The contour used on those has worked pretty well with soft dome tweeters.
I’ve seen such dips with waveguides and soft dome tweeters before. I suspect the combo of a (too) high dome profile and a certain throat size. Flatter domes (Seas, Scan Speak) seem to perform better.
@augerpro Thanks for the recommendation. The diameter of the Morel doesn't match very closely to any of the scanspeaks or the Satori. I used the same ratio as the Bliesma.
Here are some cross sections of the "custom" one(teal) I created vs. the Bliesma (Blue), and SB Acoustics (pink).... I wish I looked at this previously because I clearly see my issue now.... I don't have that nice curved shape, but rather a more conical shape.
I noticed that the Scanspeak and Satori (I believe both are silk) are much taller than these. Is this because the silk performs differently than a metal dome (SB and Bliesma are aluminum and Be respectively).
What surface modeler are you using to generate these curves? To me it looks like some surface modeling. I am thinking to adapt one of the SS or Satori waveguides to the Morel driver since my cabinets are cut and I can go bigger...but not smaller to match the SB Acoustics).
Here is the Scanspeak(gray) vs. Custom (teal).
Here are some cross sections of the "custom" one(teal) I created vs. the Bliesma (Blue), and SB Acoustics (pink).... I wish I looked at this previously because I clearly see my issue now.... I don't have that nice curved shape, but rather a more conical shape.
I noticed that the Scanspeak and Satori (I believe both are silk) are much taller than these. Is this because the silk performs differently than a metal dome (SB and Bliesma are aluminum and Be respectively).
What surface modeler are you using to generate these curves? To me it looks like some surface modeling. I am thinking to adapt one of the SS or Satori waveguides to the Morel driver since my cabinets are cut and I can go bigger...but not smaller to match the SB Acoustics).
Here is the Scanspeak(gray) vs. Custom (teal).
Is the Morel faceplate removable? If so, is it flat underneath? Can you post a close-up pic with notes showing important dimensions, like outside diameter of surround?