Arvo Part Type 3 crossover, measurements - where?

[ThomasW]

He likes his tweeters BIG.... :B

[JonMarsh]

Great Scott, no pictures today, spent my time doing laundry, cleaning bathrooms, and laying up fiberglass on the back of the first test waveguide. (just reinforcement).

Actually, this seems markedly easier to move around than the last Arvo. I think it's actually lighter. My condoloences to your spouse....

[philip_g]

Laundry? Cleaning? These are foreign concepts to me. :T

[JonMarsh]

Yeah, well they've kind of been that way for me, too- but even I can stand that only so long.


First tweeter/waveguide assembly is finished. The waveguide is MUCH deader than stock. Good old fiberglass and marine epoxy! :T

Will try to post a few pics this weekend before I head to Austria on biz next week. The REAL goal this weekend is getting cleand up enough to have my daughter over for XMAS eve dinner! And I'll get back just on Friday before the Holidays, so time is tight! Christmas shopping? I'll have to do mine in January this year.

[philip_g]

Yeah, well they've kind of been that way for me, too- but even I can stand that only so long.

I actually broke down last week and got one of those electric things that you press a button and it sprays your shower with cleaning stuff. ops:

What's next? A roomba? :rofl:

[capslock]

Jon, do I understand the drawing right, the baffle core is 16 resp. 24 in. wide, but will be 18 resp 26 once the side trims have been added?

I understand you are planning to route these side trims from 3 x 1.5 beams of some hardwood, right? Wouldn't thingies with say 1 in. rear extension give more rigidity as well as slightly higher directivity as an U-baffle as approached?

You probably laid down your thoughts on the woofer configuration in one of the many Avro threads, but I must have missed them:
- how about an arrangement that cancels some of the dynamic forces, i.e. two half W-baffles stacked on top of each other (wouldn't run out to 300 Hz, though)?
- using a U-baffle, cardoid like sub as suggested by John Kreskovky (might be preferable if you have to stay within 3 - 6 ft of rear wall, could run to 300 Hz if not combined with force cancellation)
- placing the second woofer on top of the mid/high section (would give more uniform LF wavefront at the cost of SAF)

They did the same thing to me, I have to travel to Japan on Sunday, returning Thurday night before the Xmas weekend. Will have 10 days off after returning from all the family visits, though, to get the basement workshop going again. Purchased a precision lathe and will be working on pole piece improvements...

[JonMarsh]

Jon, do I understand the drawing right, the baffle core is 16 resp. 24 in. wide, but will be 18 resp 26 once the side trims have been added?

I understand you are planning to route these side trims from 3 x 1.5 beams of some hardwood, right? Wouldn't thingies with say 1 in. rear extension give more rigidity as well as slightly higher directivity as an U-baffle as approached?

You probably laid down your thoughts on the woofer configuration in one of the many Avro threads, but I must have missed them:
- how about an arrangement that cancels some of the dynamic forces, i.e. two half W-baffles stacked on top of each other (wouldn't run out to 300 Hz, though)?
- using a U-baffle, cardoid like sub as suggested by John Kreskovky (might be preferable if you have to stay within 3 - 6 ft of rear wall, could run to 300 Hz if not combined with force cancellation)
- placing the second woofer on top of the mid/high section (would give more uniform LF wavefront at the cost of SAF)

They did the same thing to me, I have to travel to Japan on Sunday, returning Thurday night before the Xmas weekend. Will have 10 days off after returning from all the family visits, though, to get the basement workshop going again. Purchased a precision lathe and will be working on pole piece improvements...

Japan?! That's even worse! The only way I could top you is if I was going to Singapore instead, but that looks like late Jan or early Feb now...

Yeah, we've been hiding discussion about these in a number of threads- sort of like spread spectrum stealth for the design process- those in the "know" know which threads to check- sometimes I think that DOESN'T inlcude me!

Like everything in life, this is something of an experiment, let me explain the basis and my reasoning.

The flat baffle is a deliberate choice, as opposed to a W baffle (got one), U baffle (all pre-existing Arvo's), H Baffle (tip of the hat to Linkwitz and others). Why? All of these other baffle types do various things to the response above 100 Hz, including some peaking (present in all dipoles), but also even with a tapered U baffle a pretty deep notch/step down in the response, usually somewhere around 250-300 Hz. W baffles have the chamber resonance that my old slot loaded sub design of the 70's had, with the rising response starting about 100 Hz. Well, you can spend a lot of effort compensating for those effects, but only be successful to a point.

After doing some simulations and some addtional baffle tests, I decided that the next project would be a plain flat baffle (some days I think I would make more progress more quickly if I listened more to my old friend Charles Hansen, as he suggested this intuitively about three years ago). Being the slow learner that I am, I finally came around to trying this approach, doing some extensive simulations in EDGE before deciding on the baffle configuraiton and driver locations.

The edge trim will be built up from layered hardwood, having some overlap on both the front and back, and attached from the back using socket head furniture screws (the big flat head type) with threaded inserts in the baffle (same as used for the woofers). Last Sunday I installed the threaded inserts for the woofers, and that went well; I like the grip and depth, and prefer the installation and fit greatly over T-Nuts. Using 1/4-20 does require drilling out the woofer mounting holes just slightly.

Interesting variations you suggest, I'll take a look at simulation in EDGE how that might work out, would probably require different baffle shape.

For now, this seems a straight forward first pass to take on this design, it puts the mid and tweeter where I like vertically for listening height, I have a pretty tall source at 250 Hz where you tend to get issues with floor bounce dips. Already have most of the crossover parts, and a simulated LF crossover design- still need final measurements.

Also just completed last week fiberglass re-inforcement of one of the waveguides and combining with one of the C13-6. May also test the C30N in that configuration, if I'm not happy with the C13. But I think the waveguide will allow me to comfortably use the C13 down to 2 kHz, and as it extends to a bit above 30 kHz, I think I'd rather stay with that driver if I can (since my budget doesn't extend to diamond tweeters, even if they are a girls best friend). The Aurasounds are much easier to work with because of their light weight, have a very stiff dead cone, and remarkably flat response and very low distortion even in the midrange. And the C90-T5 is pretty remarkable. I have my fingers crossed on this one that it may really wind up being something special...

Gotta run for my train- in Starbucks right now!

[capslock]

Yes, W baffle is definitely out if you want to use it at that high a frequency. A double midrange might give you the chance to go lower, but you'd run into floor bounce more with a lower XO point.

My options 2 (sub at bottom and top end of baffle) and 3 (damped U-sub), which can be combined by the way, do not have that limitation. At least I hope that the notch in the response of the U-sub can be dealt with by optimizing the lining/stuffing.

Have a look at chapter 6 of www.dipolplus.de/frameset.htm -- he links to two other pieces of software that can do some things Edge can't.

Edit by moderator to add a direct link to the software downloads.

https://web.archive.org/web/20070704042249/http://www.dipolplus.de/thema6.htm

[capslock]

Oh, and did I get that right about the baffle + trims being 18 inches wide at the top and 26 at the bottom?

[JonMarsh]

Yeah, that should be the finished dimesion.

And thanks for the link, Eric!

Mit freundlichen Grüßen,

~Jon