Flex Your PCD Mettle:

[HeartFixr]

Re: Need help with crossover

Jumping in with both feet, actually I'm all in and trying to build the Econowave Altec GPA 414-8B with QSC waveguide and BMS 4552 compression drivers. In looking at Zilch's picture he had a 3/4" baffle with the waveguide front mounted and the Altec rear mounted. If I use a 1 1/2" front baffle and rear mount the Altec, it will reduce offset z measurement. How big of a deal is this with the crossover and if so, how do I compensate for it?
BTW, is this speaker ported (which is what I am building)?

[avincenty]

Re: Need help with crossover

When you say you have the L-pad maxed out, you mean no attenuation at all? From the looks of that plot, I'd say you need to dial it back just a bit.

Here's what Zilch got for the Deluxe measurement.

http://techtalk.parts-express.com/sh...70#post1624970

The tweeter response is quite a bit flatter in his measurement. The 2KHz bump is due to the baffle edge protruding. That goes away when the baffle is flat.

Yes, max HF, no attenuation. I am using a DIY panasonic mic, with the WM-61a capsule. I had applied a generic calibration file, if i remove the calibration, the response flattens out.

I think i will take them oustside to the yard for a new measurement.

[Pete Schumacher]

Re: Need help with crossover

Thanks Pete.

I just tried it with the 3012LF crossover values and took some measurements in my listening area. The first thing I noticed is that I need to keep the Lpad maxed out (to your point about sensitivity). However it sounds good. I think the low jaggedness is due to room interaction, the speakers are all the way to the back wall, three feet from the side walls. I tested swapping the polarity of the woofer and the FR shows a huge dip around 1.3Khz. The mic was positioned 1 meter away (spl not calibrated) and between the CTC of the woofer and horn. I believe Zilch decided to forgo horn compensation for the Deluxe crossover. Based on what I am getting, should I include one? I guess If I do that I would need to change the resistor values to have better response from the LPAD. Is there a recommended trick for changing the values?

Alfredo

When you say you have the L-pad maxed out, you mean no attenuation at all? From the looks of that plot, I'd say you need to dial it back just a bit.

Here's what Zilch got for the Deluxe measurement.

http://techtalk.parts-express.com/sh...70#post1624970

The tweeter response is quite a bit flatter in his measurement. The 2KHz bump is due to the baffle edge protruding. That goes away when the baffle is flat.

[avincenty]

Re: Need help with crossover

I'd be paying attention to the differences in sensitivity between the 3012 and the JBL. You may need to adjust the L-pad on the tweeter.

As far as setting the LF cutoff and slope, you do have some contribution from the natural response of the woofer that must be accounted for. I'd start with a 3rd order slope on the DSP . . .

Thanks Pete.

I just tried it with the 3012LF crossover values and took some measurements in my listening area. The first thing I noticed is that I need to keep the Lpad maxed out (to your point about sensitivity). However it sounds good. I think the low jaggedness is due to room interaction, the speakers are all the way to the back wall, three feet from the side walls. I tested swapping the polarity of the woofer and the FR shows a huge dip around 1.3Khz. The mic was positioned 1 meter away (spl not calibrated) and between the CTC of the woofer and horn. I believe Zilch decided to forgo horn compensation for the Deluxe crossover. Based on what I am getting, should I include one? I guess If I do that I would need to change the resistor values to have better response from the LPAD. Is there a recommended trick for changing the values?

Alfredo

[Pete Schumacher]

Re: Need help with crossover

Thanks Pete,

I do have a minidsp so maybe i can use it to tune the xo points. Based on the 3012lf xo posted in this forum, is the F6 of the HF 1558hz? So i would set a LR2 at 1558 and then play with the LF till i get the flattest curve. Thing is I already have the components for the 3012LF. I can easily modify the values of the inductor as i have a LCR bridge.

Alfredo

I'd be paying attention to the differences in sensitivity between the 3012 and the JBL. You may need to adjust the L-pad on the tweeter.

As far as setting the LF cutoff and slope, you do have some contribution from the natural response of the woofer that must be accounted for. I'd start with a 3rd order slope on the DSP . . .

[avincenty]

Re: Need help with crossover

Since the woofer you're using has a more extended top end than the 3012LF, the only thing you really need to modify is the woofer XO.

You may find that you need to increase the filter from 2nd order, to 3rd order overall, so that the woofer acoustic slope follows LR4 in the crossover region.

You have measurement capability, so modification of the woofer XO shouldn't be too difficult to get dialed in.

Thanks Pete,

I do have a minidsp so maybe i can use it to tune the xo points. Based on the 3012lf xo posted in this forum, is the F6 of the HF 1558hz? So i would set a LR2 at 1558 and then play with the LF till i get the flattest curve. Thing is I already have the components for the 3012LF. I can easily modify the values of the inductor as i have a LCR bridge.

Alfredo

[Pete Schumacher]

Re: Need help with crossover

I decided to give the econowave deluxe a try and purchased the DE250 CD, the QSC waveguide and found locally some vintage JBL2204H woofers. I put the woofer into a 2.8cf cabinet tuned to 52 hz and captured the FR outside in the backyard. I also purchased the crossover board and the componets for the Econowave deluxe for the 3012LF. However, looking at the frequency response from the woofer it is no near the respose of the 3012 LF, Pretty sure that the values for the 3012LF wont work. By looking at the FR the Baffle Step is obvious below 170hz. What crossover point looks good based on the graph? Is 1K doable or is it too close to the limit. My understanding for the values for the 3012LF is that they have been chosen to compensate for the rising midrange response past 1K. I put the values into a RLC calc and it gives a Fc of 684hz. Should I keep the HF crossover intact and just play with the LF inductor? Or should I try LR2 values that are matched for both HF and LF?

Alfredo

Since the woofer you're using has a more extended top end than the 3012LF, the only thing you really need to modify is the woofer XO.

You may find that you need to increase the filter from 2nd order, to 3rd order overall, so that the woofer acoustic slope follows LR4 in the crossover region.

You have measurement capability, so modification of the woofer XO shouldn't be too difficult to get dialed in.

[avincenty]

Need help with crossover

I decided to give the econowave deluxe a try and purchased the DE250 CD, the QSC waveguide and found locally some vintage JBL2204H woofers. I put the woofer into a 2.8cf cabinet tuned to 52 hz and captured the FR outside in the backyard. I also purchased the crossover board and the componets for the Econowave deluxe for the 3012LF. However, looking at the frequency response from the woofer it is no near the respose of the 3012 LF, Pretty sure that the values for the 3012LF wont work. By looking at the FR the Baffle Step is obvious below 170hz. What crossover point looks good based on the graph? Is 1K doable or is it too close to the limit. My understanding for the values for the 3012LF is that they have been chosen to compensate for the rising midrange response past 1K. I put the values into a RLC calc and it gives a Fc of 684hz. Should I keep the HF crossover intact and just play with the LF inductor? Or should I try LR2 values that are matched for both HF and LF?

Alfredo

[Pete Schumacher]

Re: Flex Your PCD Mettle:

You're adding a time delay which shows up on the graph as added phase shift.

[dtruck]

Re: Flex Your PCD Mettle:

Thanks, but the summation vs. angle is all looking correct to me, and manageable with a crossover. What I don't understand, is how it makes sense that I see these two graphs when I take the tweeter z-offset from zero to -.12m, while moving the woofer offset accordingly to keep the relative offset the same (it seems to be pretty much exactly +10cm woofer from straight off the midpoint between drivers). I'm not sure if I'm just being dense about phase, or there is a problem with my frd generation? (please ignore the two measurement glitches - I just didn't bother repeating the last round when those showed up).

[Pete Schumacher]

Re: Flex Your PCD Mettle:

I have another PCD question: Can anyone explain to me the changes in driver phase that result from changing the offset? For example, if I put a tweeter at -9cm and woofer at 0cm, the woofer phase will be as imported but the tweeter phase goes crazy. If I put the woofer at +9cm and the tweeter at zero, the woofer phase is the one that changes. Logically, it seems like the "real" numbers are both negative, one more than the other, in which case both change. This makes no difference to the summed response, but it does make a big difference to the total phase.

I guess the question is: is there a way to go about this that's most likely to result in a realistic simulation of total phase? My almost complete lack of understanding of what minimum phase means may have something do with this....

Anybody have an opinion on whether this is too optimistic for a DE250?
[ATTACH=CONFIG]27225[/ATTACH] [ATTACH=CONFIG]27226[/ATTACH]

Minimum phase is derived directly from the driver response. When you use minimum phase files, PCD will allow full simulation, including driver offsets and off-axis response given flush mounted drivers. The relative phase will be calculated for an offset driver if accurately positioned (x, y, z) in PCD. With large offsets, like 9cm, there will likely be summation issues to overcome, especially for higher crossover points. You may find that mounting the woofer to the back of the baffle could help, or inverting tweeter polarity, so that the relative phase at the crossover point lines up between drivers.

1500Hz for the DE250, which is about what you have there, seems about par for the course.

[dtruck]

Re: Flex Your PCD Mettle:

Anybody have an opinion on whether this is too optimistic for a DE250?

[dtruck]

Re: Flex Your PCD Mettle:

I have another PCD question: Can anyone explain to me the changes in driver phase that result from changing the offset? For example, if I put a tweeter at -9cm and woofer at 0cm, the woofer phase will be as imported but the tweeter phase goes crazy. If I put the woofer at +9cm and the tweeter at zero, the woofer phase is the one that changes. Logically, it seems like the "real" numbers are both negative, one more than the other, in which case both change. This makes no difference to the summed response, but it does make a big difference to the total phase.

I guess the question is: is there a way to go about this that's most likely to result in a realistic simulation of total phase? My almost complete lack of understanding of what minimum phase means may have something do with this....

[dtruck]

Re: Flex Your PCD Mettle:

The z offset is the difference in straight line distance from the acoustic centers to the mic position. The "zero plane" or baffle is not considered. When you set the tweeter offset to "0", you're then establishing the woofer relative to that position.

Excellent, that clears that up. Thanks.

[Pete Schumacher]

Re: Flex Your PCD Mettle:

Yeah, I suppose just building something is the best way to find out what's what. I would still like to know what the math is based on, though. Is z-offset as entered in PCD intended to be the difference in straight line distance from measurement position to each acoustic center? Or, is it relative to a hypothetical zero plane that is perpendicular to the drivers? Thanks.

The z offset is the difference in straight line distance from the acoustic centers to the mic position. The "zero plane" or baffle is not considered. When you set the tweeter offset to "0", you're then establishing the woofer relative to that position.