TMM in Response modeler and PCD

squamishdroc

I am branching out and want to design a TMM - I have not decided if it will be a two way or 2.5 way yet. What i'm wondering is how to input the frd and zma data in to RM and then into PCD. Do I model both woofers together in RM and enter the frd file into PCD? Do i use the zma file for one driver in PCD or use the file from RM for both drivers? Hope this makes sense what i'm asking.....

PS I should add that i am just working with sims - don't have drivers yet

dynamo

Model them individually. PCD allows you to toggle for two woofers in a two way and enter the frd and zma for each one. You can then toggle for series or parallel, and enter x,y, and z locations for each. 

jhollander

+1 I'd use response modeler for the impedance file for each individually also.  I like the response blender for adding the baffle to each driver.  Save the impedance file for later if your woofers will be in the same enclosure when built if you go the 2.5 route.

squamishdroc

Thanks guys - i will work on that

squamishdroc

So I have been working on this the last week - just had a question about the phase response. In the sim should the area where the phase almost lines up be closer to the crossover point or does it not matter?

dynamo

Watch for others responses as mine may be wrong. The phase wraps (vertical lines around 2k) don’t need to line up perfectly. Most importantly is that the lines are together over the crossover point and for an octave or more each way from the crossover point. Sorry if this is wrong.

PWRRYD

Dynamo is correct.  Your phase doesn't look too bad.

jhollander

To me the phase wrap still show being off by 60 degrees. 

wrt to the phase lines offset from the driver I've always thought that's due to the driver slopes not being the same, asymmetric.

Assuming LR4ish, I'd still try to get a reverse null. The reverse null would be between the lines crossing and the phase wrap. 

PWRRYD

I don't see what you mean John.  The xo point looks to be about 2700 Hz, at which point the driver phases are laying on top of each other (zero relative phase difference).  The point of phase wrap (two vertical lines) is not important.

jhollander

Assuming an even order x-o, my experience has been when there's partial alignment excluding the phase wrap the reverse null will be weak, which indicates to me there's a better fit, or if left as is, a bump off axis.

D1PP1N

I would like to pick Paul C's brain on this. From what i remember he told me he does the 2nd woofer in a  tmm by ear because measurements look terrible close up but back at the listening position its fine. I noticed up close to the build he did for me it was super bass heavy (1-3ft away). The one he did for me just had 1 extra coil on the woofer. 

squamishdroc

The crossover is 2nd order electrical on the two woofers (wired in series) and 4th order electrical on the tweeter. In the above sim the tweeter is reverse polarity. Here is the tweeter positive (reverse null).



I tried a 2.5 way crossover as well but couldn't get a response curve that I was happy with. I have the drivers on order and I will probably try both crossovers (2 way and 2.5) - as @D1PP1N said the one thing with the 2.5 way was a large bass peak.

jhollander

That's nice alignment, I would have not expected that from what I saw in the previous post.

squamishdroc

jhollander said:

That's nice alignment, I would have not expected that from what I saw in the previous post.

Thanks John - I will roll with this crossover if the actual responses match the sim.

Ed_Perkins

squamishdroc said:

So I have been working on this the last week - just had a question about the phase response. In the sim should the area where the phase almost lines up be closer to the crossover point or does it not matter?

What seems strange to me is that the woofer's phase is actually increasing from 10 Hz to about 800 Hz.  That just doesn't seem possible to me.  It appears the woofer's low end was spliced like when using Response Modeler.  Did you create a minimum phase version after the splice?

jr@mac

Two comments from this part of the peanut gallery...

1. Wait until actual in-box measurements show up. 

2. That "large" peak will be nothing but beneficial on small drivers, especially at that frequency. You may find you prefer it, I have a theory why and it isn't necessarily related to compensating for lack of extension.

ani_101

why will the Tweeter XO at 2700 have anything to do with the 2.5 or 2 way decision?

dynamo

I’m guessing 2700 is a bit high for the c to c distance from tweeter to bottom woofer to be ideal in a 2-way, where a 2.5-way that would not be as much of an issue.

I just personally prefer the two way because I was able to get the same set of drivers to sound so much better on my first attempt at a two way with them than I ever was with multiple 2.5way attempts. This I’m sure reflects my lack of design experience however as well.

squamishdroc

Ed - the woofer response was modeled in RM and I extracted min phase from that. The orange phase line is for the tweeter though if you are referring to that one.

JR - that's my plan - drivers are in the mail and then I will measure in box. Oh - I guess that means I have to make a box now

Ani - the 2.5 way didn't really affect the tweeter response or crossover point - it created some problems below the crossover point that I didn't like. But that could also have been operator error too

Dynamo - will see how it plays out once I get the box built and can measure in box.

ani_101

for 2 way, I was hoping it would be MTM, not TMM. 2.5 way would be TMM, so the CTC should not be a deal breaker

squamishdroc

Well I got my speakers last week and I made a test box to take some measurements in. My options for these guys was either a MTM or a TWW 2.5 way. After taking some measurements it looks like the MTM will be the better choice of the two. The TWW crossover ended up with too low of an impedance load (under 4 ohms average) to be stable with my amp. 



I will be using a piece of Suar for the baffles and the rest will be mdf with some veneer on it.



I did some crossover work tonight and some OmniMic sweeps and came up with these results which were pretty close to my sims

JasonP

That is an impressive amount of dip at 4k, and of course, might be baffle interaction. I'd love to see a measurement at 15 and 30 degrees off-axis horizontal. I don't know if that is too much work for you!

squamishdroc

JasonP said:

That is an impressive amount of dip at 4k, and of course, might be baffle interaction. I'd love to see a measurement at 15 and 30 degrees off-axis horizontal. I don't know if that is too much work for you!

Here you go   The tweeter had a 5db dip at 4k and then a huge spike (10db) at 1800. That dip I guess is from trying to tame the spike and the woofers are 5db down at 4k too.



And here is driver response in Xsim.