Flex Your PCD Mettle:

[Zilch]

Re: Flex Your PCD Mettle:

I wonder if anyone is bringing an Econowave to InDIYana?

How long do you have to build a pair...? :D

[Pete Schumacher]

Re: Flex Your PCD Mettle:

How long do you have to build a pair...? :D

Not sure the budget permits at the moment. But it is something that really does have my interest piqued.

[dlr]

Re: Flex Your PCD Mettle:

That's the crossover for the Geddes Abbey design, which uses the same B&C DE250-8 compression driver on a 12" round Geddes axisymmetric waveguide with B&C 12DBX100 12" woofer. I'm using the KappaLite 3012LF because it can accept a Q=2 assist at 30 Hz to make extended bass in this small cabinet. Used with sub(s) in HT or for SR, they don't need that, of course....

Keep in mind that Geddes also has the dual density acoustic foam insert in his OS waveguide and that his design is for listening significantly off-axis. Coupled with the crossover designed to maintain response throughout a +/- 45 degrees within 2db, IIRC, for more uniform power response, it's not surprising that his crossover is more complex.

dlr

[dlr]

Re: Flex Your PCD Mettle:

Geddes sure uses a lot of real estate for a crossover.

Yeah, and he's definitely a stickler about the inductor spacing and layout.

dlr

[Skywave-Rider]

Re: Flex Your PCD Mettle:

Nice work on this design. I am interested in building it or following similar procedures to optimize a crossover for woofers I have on hand.

Question, does "relative offset," as discussed in this thread, imply offset of drivers measured from a fixed designed-acoustic-center? Or what?

[Pete Schumacher]

Re: Flex Your PCD Mettle:

Nice work on this design. I am interested in building it or following similar procedures to optimize a crossover for woofers I have on hand.

Question, does "relative offset," as discussed in this thread, imply offset of drivers measured from a fixed designed-acoustic-center? Or what?

When using the FRD files for XO modeling in PCD, acoustic offsets need to be programmed in unless the FRD files have the absolute phase incorporated. The FRD files Zilch provided have minimum phase information, which is directly related to the frequency response and can be derived from it. Using minimum phase information requires that the acoustic offset be estimated to restore the relation to the drivers relative phase in the design. Zilch also supplied the combined, unfiltered response, which would contain the relative acoustic offset, and can be derived by adjusting the value in PCD while combining the individual FRDs until they closely match the measured combination. His measured data was at the midpoint between woofer and tweeter at a distance of 44". By replicating the driver positions in PCD, I was able to pretty closely recreate the plot from the individual FRD files and deduce the acoustic offset.

Having the drivers set in 3 dimensions allows for some decent simulations of polar response and allows for more accurate modeling of response in the XO region.

[dlr]

Re: Flex Your PCD Mettle:

Nice work on this design. I am interested in building it or following similar procedures to optimize a crossover for woofers I have on hand.

Question, does "relative offset," as discussed in this thread, imply offset of drivers measured from a fixed designed-acoustic-center? Or what?

To expand a bit on what Pete said, using minimum-phase responses is required to easily investigate the off-axis response without manually calculating the change in distance when the mic/listening position is moved. If you also specify the driver diameters (needed for directionality calculations), the one thing that is not taken into account will be baffle diffraction. Even so, it's very useful to see the influence of driver offsets.

dlr

[stgdz]

Re: Flex Your PCD Mettle:

Zilch to ask a silly question,

could the ""-waveproject that you are building be modified into a three way with say this driver for a midrange?

http://www.parts-express.com/pe/showdetl.cfm?Partnumber=294-651

You could use the dayton classic as a the woofer, cross it lower and be able to soak up some of the mid range. It would have to be mounted on baffle as its depth is about 3.2" compared to the 5 1/4" of the dc300.

Would that tilt the axis a little bit?

[Zilch]

Re: Flex Your PCD Mettle:

Could the ""-waveproject that you are building be modified into a three way with say this driver for a midrange?

http://www.parts-express.com/pe/showdetl.cfm?Partnumber=294-651

You could use the dayton classic as a the woofer, cross it lower and be able to soak up some of the mid range. It would have to be mounted on baffle as its depth is about 3.2" compared to the 5 1/4" of the dc300.

Would that tilt the axis a little bit?

The axis location and forward lobe will change depending upon the acoustic offsets, center-to-center distance, driver directivity, and filters. (I forget what else. ;) )

In detail, everything must be adjusted to optimize the performance of each combination.

[The only thing more fun is listening to the result.... :D ]

[Skywave-Rider]

Re: Flex Your PCD Mettle:

Thanks for the explanations.

Let's see if I understand this correctly....
Minimum phase subtracts flight time, thereby essentially placing separate drivers at the same distance to the measurement microphone, which is not good for modeling systems which have actual offsets.

If I was to generate FRD files to import into PCD, I would have to be sure not to include minimum phase information. I am using Fuzzmeasure right now, I have to meddle around to see what I can do.

If I understand the fundamental concept, let me know and I'll try to move up from there.

[Pete Schumacher]

Re: Flex Your PCD Mettle:

Thanks for the explanations.

Let's see if I understand this correctly....
Minimum phase subtracts flight time, thereby essentially placing separate drivers at the same distance to the measurement microphone, which is not good for modeling systems which have actual offsets.

If I was to generate FRD files to import into PCD, I would have to be sure not to include minimum phase information. I am using Fuzzmeasure right now, I have to meddle around to see what I can do.

If I understand the fundamental concept, let me know and I'll try to move up from there.

If you don't have minimum phase information, then you can't use PCD to model how things change as you go off axis. You do need phase information of some kind in order to correctly model how the responses combine at the crossover point.

[Skywave-Rider]

Re: Flex Your PCD Mettle:

Thanks, Pete.
"By replicating driver positions in PCD..."
Is this done with a tape measure or equivalent, and input into the software?

I will have to try to get PCD running before I ask any more questions.

[Pete Schumacher]

Re: Flex Your PCD Mettle:

Thanks, Pete.
"By replicating driver positions in PCD..."
Is this done with a tape measure or equivalent, and input into the software?

I will have to try to get PCD running before I ask any more questions.

You can use a tape measure as an estimate, but doing what Zilch did, by measuring the sum of the drivers along with each individually, offsets can be derived from PCD and then used with the minimum phase files to simulate the off axis response.

The driver relative positions are input at the control section near the top of the PCD spreadsheet. In that same area are spinner controls for dialing in different listening positions, both horizontally and vertically, as well as driver diameters for even more accurate simulations, taking into account driver directivity as a function of diameter.

[Zilch]

Distortion Measurements:

Pete requested distortion measurements:

Disclaimer:

I never run distortion measurements and have zero skill in doing them. In this instance, I ran the gated sinusoidal sweep using the settings shown in the legend with autodelay at 1 khz and clicked the buttons. Curves are raised 20 dB.

Let Zaph and Geddes argue out whether distortion measurements are at all meaningful in the larger sense; I'm just not a player in this.... :rolleyes:

http://www.gedlee.com/distortion_perception.htm

[Pete Schumacher]

Re: Distortion Measurements:

Pete requested distortion measurements:

Disclaimer:

I never run distortion measurements and have zero skill in doing them. In this instance, I ran the gated sinusoidal sweep using the settings shown in the legend with autodelay at 1 khz and clicked the buttons. Curves are raised 20 dB.

Let Zaph and Geddes argue out whether distortion measurements are at all meaningful in the larger sense; I'm just not a player in this.... :rolleyes:

http://www.gedlee.com/distortion_perception.htm

That tweeter distortion level is mighty small indeed.