Calling all gurus with simulators

[5th element]

Hi there your brainy bunch lol

OK Im having a little bit of trouble with LSPcad and would appreciate it if any of you other people out there with simulators could use the measured zma and txt files that ive taken, and apply the supplied xover and see what you arrive at.

What it should look like is about a 2nd order LWR @ 2500 with full baffle step compensation. Atleast thats the prediction.

Many thanks Matt

ZMA and TXT files

 

[5th element]

noone?

 

[pxr5]

chart

made a chart, with speaker workshop,
cant post it though, says file to big, about 1 meg,
looks ok though, slight bass hump

 

[5th element]

yeah bass hump is from the driver in the enclosure, slight rise in response.

 

[pxr5]

bass hump about plus 2.5 db
-6db xo point sub 2000hz
roll of rate good
baffle step is a nightmare
i always have to use lower frequencies than those suggested by olsons students

 

[pxr5]

chart

just made file smaller, here it is

 

[Zaph]

I popped those numbers into SE and got pretty much the same thing as pxr5.

Seems like a well behaved breakup node. Have you considered trying a slightly smaller single inductor with a RLC notch paralleled across the driver terminals?

 

[Zaph]

Seems like a well behaved breakup node. Have you considered trying a slightly smaller single inductor with a RLC notch paralleled across the driver terminals?

For example, I just threw this together for an alternative topology. It might need some tweaking.

 

[5th element]

Interesting I hadnt considered that you learn something new everyday. I often try to stay away from notch filters to keep cost down, but that has other advantages over just a smaller inductor and would probably come out about the same price.

 

[5th element]

I think ive found the problem as to where some of the problems ive been having have come from.

OK here goes, I had already used the provided network above in a speaker design in LSPcad which worked well. So I exported the SPL curve for the bass driver so I could use it as a target for zaphs network.

OK what happens is this. I export the bass drivers SPL (with filter) as a target. Then to check that its exported correctly import it back in as the target for the bass driver, in the same design. So theoretically they both should match perfectly. But they dont.

Here is what happens. The red line is the SPL for the driver with xover. The grey line is suppose to be the same as the red line. But for some reason the grey line is what LSPcad exported from the redline....

I think this is why I can getting problems because this is not just limited to the export function but everything else.

Im asking all these questions because I have designed two pairs of loudspeakers, well three if you count my active set. But they all were highly accurate, the predicted simulated response matched the measured response once built.

However my last pair were completely off. COMPLETELY not just some small little error everything was a mess. The prediction and actual were way off. Somethings not quite right. Ill email ingemar now but if any of you have any ideas?

 

[Zaph]

I wish I could help out but I'm pretty much an SE guy.

You might want to ask Dennis Murphy on the Mad board. Aside from Ingemar, I don't think anyone knows LspCAD better than him.

 

[5th element]

Im sure Ill get it figured out but its annoying to have this problem because it means the $500 I payed for it isnt working at the moment.

I simple reinstallation might solve it but ill wait and see Ingemar might have seen this before because its obviously an error somewhere in the programming which shouldnt occur, but for some reason has.

I think version 6 is out soon so Ill be able to update with that.