Possible to model this configuration in PCD 7 ?

[percy]

Low pass filter, either series or parallel, for 3 identical drivers connected in series.
If so, please advise how.
Thanks!

[percy]

I am attaching an example schematic about the configuration I was trying to describe.
and by PCD 7 I meant Passive Crossover Designer (spreadsheet) version 7 by FRD Consortium.

[wintermute]

Hi Percy, I haven't got PCD handy (new pc) but from memory you can do two woofers, but I'm not sure about three.

Do you have real measurements taken on the baffle? What I did for my MTM was take the impedance measurement for both woofers in box, and also the freq response for both (on axis with the tweeter) and then used this as a single driver in PCD (and later speaker workshop). It worked a treat. ie measure with all three running.

The diagram you have posted looks like it was done in speaker workshop. The crossover modeller in SW is really powerful (and in some respects more accurate than PCD) The area I found it more accurate was when modelling notch filters due to being able to put in the DCR of the coils.

Where PCD is great though for being able to quickly increase or decrease caps or coils values and see an immediate effect on the Frequency response, Speaker Workshop is a bit of a pain in that regard.

Tony.

[percy]

I have on-baffle measurements but of only one driver by itself, not with all 3 wired in series like that. Is it ok to do measurements with all 3 drivers wired in series, regardless of which application I use to model the crossover ? Would it work fine ?
Yes its in Speaker Workshop and I am getting a hang of it, its just that like you mentioned the ease with which you can see the response shaping instantly in PCD.

[AllenB]

Originally Posted by percy

Is it ok to do measurements with all 3 drivers wired in series,

For simplicity, use the single driver measurement for all three. For something more accurate, do three separate measurements. Before you go to that extent it may be worth idealising your measurement environment with regards to reflections.

regardless of which application I use to model the crossover?

Yes.

Speaker Workshop

I created three woofers to use in the crossover.


Here is the response of your crossover (in black) compared to the old response (red).

[wintermute]

IMO yes it is fine to measure all three running at once, the only tricky part is where do you place the mic?

For me the decision was fairly easy since I have an MTM configuration. I took all measurements on axis with the tweeter. If you want to measure all three together then you should place the mic in the position that would be on axis with your ears when seated listening. Measurements of all speakers in the system can (and should be) done without moving anything. Also the impulse response (assuming you are using impluse) should be locked so that all drivers start at the same time zero.

If you do this, then all offsets (X Y and Z plane) can safely be ignored. I figure why complicate things by having to simulate the offsets and the resulting interactions if you can measure the actual response in the first place

However taking individual measurements will allow you to experiment with what the effect would be if you moved things around...

As well as doing the measurements with the raw drivers, do one with the crossover in place as well (again without moving anything). Then compare the result of the simulated response with the actual measured response. This is a good sanity check, and should tell you if your measurements are working correctly or not.

Here is what I posted on my MTM thread with respect to the above: the original graph is attached.

Hows this for a correlation between sim and actual measured response? Black is the sim, and Blue is the actual measured response. all measurements were done with the mic and speaker in the exact same position, crossover was already made, just swapped the wires for the three measurements, and then simmed the actual built crossover with the individual tweeter and woofer responses to get a comparison. Pretty damn close!! If I can get that good a correlation for the 4th order network I'll be very happy

I can't remember the exact crossover circuit but I suspect it was 4.3uF and 2.2 Ohm in series with the tweeter, and 4K notch filter on the midbass units 13uF 100uH ~16 ohms).

Tony.