LspCAD and enclosure design

[Saurav]

A friend asked me to help him design enclosures for some HT speakers he plans to build (something I've never tried before ). He's chosen his drivers, and I thought this would be a good opportunity to see how this works in LspCAD. So I tried to follow the tutorial. I picked the BR template, add the source and ground, switched to 'simulate', entered the T/S parameters, used the wizard to get some values for box volume and port length / diameter. The calculated response looks reasonable.

Now comes the part I'm stuck at. Let's say I want to shoot for something like a B2 @ 80Hz, to work with a regular HT receiver. So I put that in as the target, check all the boxes I can find for the optimizer, start the optimizer, and... nothing happens. The optimizer just sits there. I see the box volume and port length bolded, so they should be set up right. Does the LspCAD optimizer not work with enclosures? I can play around with values by hand, of course, but at that point I might as well use something like Unibox.

So maybe that's what I should do, or am I missing something and LspCAD should be able to do this?

[5th element]

I cannot speak for LspCAD6 as I've only got 5.25, however the optimiser works fine with boxes. You just have to tell it what parameters it's allowed to fiddle with. Box size, stuffing, filter etc. But it sounds like you've got that sorted anyway.

From what I've read though, there is a button you can press to make 6 run as a 5.25 lookalike. Maybe you could try that?.

[Saurav]

I think I had all of those checked, and the top box also checked which turned on the whole box + port group. I'll try 5.25. I think that binary was bundled in, or maybe 6 runs as 5.25, not sure how that works.

Maybe I just had a stupid config where the optimizer decided there was no way to get from here to there The odd thing is, it worked one time, but I think it ended up varying the T/S parameters of the driver to get there. Which wasn't really useful.

[BobEllis]

You have a system that naturally has a 4th order cutoff and you are asking the optimizer to suggest ways to get a second order cutoff? It might work better with a sealed box.

I don't have any experience with LspCAD, though.

[Saurav]

Heh, you're right, I mixed that up, didn't I. I was thinking about sealed boxes when I was trying tor a BW2 @ 80Hz to line up with the receiver's XO. The problem with a sealed box with this driver is that it ends up rolling off around 150Hz (at least for reasonable sizes).

Actually, that makes more sense now. Probably the only way to get a 2nd order rolloff in my design is to use a different driver, which is why the optimizer runs if I let it optimize the driver, but not the box.

*facepalm*

OK, so here's my next "I haven't read Dickason and don't know anything about designing boxes" question: Should I just stick with whatever the wizard comes up with? I could play around a little by hand and see what I get. BB4, QB3, Q = 3, 7, 15... this is probably all explained somewhere, and I need to find the right reading material.

My friend also said that he's going to be using this with a receiver / sub, but he's also building something for a (non-critical-listener) neighbor, and that one will run full-range. In that kind of situation, would it make sense to add a little bump to the bass around the rolloff region? I think some (many?) commercial bookshelf speakers do that to hide the fact that there's not much output below 60Hz?

[Dennis H]

You might be able to get a B2-80-like response out of the ported box, at least around 80 Hz, if you limit the frequency range of the optimizer. Try to get a shoulder that looks sorta B2-80 and don't worry about the steeper roll-off down low. The steeper roll-off would affect group delay but you should be able to compensate with the receivers sub distance control.

[Saurav]

Thanks, that's a good suggestion. Now that you mention range, I remember noticing that the optimizer picked an overall level that was around the port's output instead of the driver's. So there may have been something else that it was confused about. In the driver optimization window, there are check boxes with no text next to them, so I have no idea what those check boxes do.

[5th element]

Hmm this does sound a little odd. Is the optimiser doing anything at all?

Even if the end result isn't actually feasible it should have a jolly good try at getting it right. In other words you should see some action going on somewhere - it might be horribly wrong action, but you should see something.

In 5 with boxes, I simply click system optimisation, select the box volume and Qa and away it goes, trying to reach some target.

One thought - do you have 'live update' selected, if it's available? I know 6 looks quite a bit different from 5, having windows all over the place. In 5 only the optimisation window updates if live update is deselected. With it selected all the other windows update at the same time.

[Saurav]

Yes, 'live update' is enabled. No, the optimizer doesn't do anything, none of the curves or numbers move. It just switches to the lower resolution graph (which is barely noticeable for this, the curves just seem to twitch once) and sits there until I hit 'stop', at which point the curves twitch back.

I'll spend some more time with 5. I can't run it directly, I get a "Let's tell Microsoft that something crashed" window. It runs when launched through 6. Could be a security thing. Anyway, the project file syntax is different, so I made a new project, but I don't think the optimizer in 5 did anything either.

I'll double check all the numbers I entered too, but I think I have them right. I clicked the "don't change this" box for all the T/S parameters I typed in, and unchecked the 'optimize' box for all the driver parameters.

[Saurav]

5.25 is a no-go. It won't load the .unt driver file that I created with 6, and clicking on driver -> new causes it to crash. So I never got far enough to run the optimizer.

I'll see what I can get with manual adjustments and then check that using the other tools. I think my first goal will be to get all the applications saying the same thing

[Saurav]

OK, this is what I end up with:



Does that make sense? Can I fit a 14" port into a 12l box? I used outside dimensions of 17x11x7.5, so it would appear that it should barely fit, but I have no idea if this is an absurd configuration or not.

I also realized that changing the 'series resistance' changes the calculated response quite a bit, so I'll probably have to re-visit this, I have no idea what he's planning for a crossover.

FWIW, in case anyone's wondering - I think he picked the Beta-8A because he plans to cross to a CD on a relatively small horn. So it's a mix of sensitivity and response.





That's mostly eyeballing. It kinda looks like 2nd order @ 80Hz. The excursion chart says it exceeds Xmax between 80 and 180Hz. Is that bad? How bad is it? What about the port air velocity? Reasonable or too high?

And I read all these warnings about how sealed boxes were much more forgiving of errors, and with a vented box you really have to get the build right. The T/S parameters are from the manufacturer datasheet. If my friend builds this, how far off might the end result be? Maybe I should ask him to first build a test enclosure and then I should measure it?