Page 1 of 2

project x diy speaker build

Posted: Sun Sep 24, 2017 1:26 pm
by jayman67
ok after reading the roy allinson papers and understanding some of it i bought these at the local charity shop to have a bash at diy speaker building.
at the moment i have no firm plans as to what i will end up with so some guidance is needed from you chaps please.

here's the donors,a pair of richard allan pavane speakers.
12inch woofer,8 inch mid and a 4 inch tweeter.
i dont profess to know or understand speaker crossovers or calculations so perhaps it would be best to stick to a 2 way design,i can on the other hand solder,iam also an ex cabinet maker and now im a steel fabricator so making the cabinets is the easy bit lol.
oh and dont hold back if you think the drivers arn't worth the hassle please say so,i had a quick listen and they are very reminiscent of ditton 44's

Image

Image

Image

so what do you think guys?
cube type 2 way?
metal cabinets?
doc modded drivers?

Re: project x diy speaker build

Posted: Sun Sep 24, 2017 1:55 pm
by Simon Hickie
Looks a most interesting project. Why do I never find these things in charity shops? Stu's Marantz 7 project looks like it might be relevant. The drivers look in pretty decent order and the crossover looks pretty simple. The tweeter may suffer a bit at the top end in terms of beaming due to the cone size. Top end was limited to around 17khz.

T&S parameters for some of the drivers here: http://www.diyaudio.com/forums/multi-wa ... ached.html

Your drivers are here http://ukhhsoc.torrens.org/makers/Richa ... index.html somewhere!

Re: project x diy speaker build

Posted: Sun Sep 24, 2017 1:57 pm
by jayman67
cheers,you see this is where im going to struggle,ive not got a clue what T&S means? lol

Re: project x diy speaker build

Posted: Sun Sep 24, 2017 3:47 pm
by karatestu
Richard Allen drivers are well respected, especially by the doc. Maybe worth keeping them but I would wait for the doc to confirm that.

Crossover looks quite simple. I don't know if doping is suitable for all drivers, they look like paper. Doc uses Kevlar jobbies but I think people have doped other materials. I have doped aluminium cones but have not got enough on yet so cant say whether it has been successful yet. You can only remove the crossover once the mid & bass have been doped. 12" driver will get you REAL bass and the mid is larger than I have seen. I don't know about the tweeter, technology has come along way as far as tweeters are concerned.

Check out if you can what the cabs are made of and how thick, any bracing ? Maybe steel lining the insides would be worth it and throwing away what ever it is stuffed with. Looks like it has an oblong port ? The doc recommends blocking ports and any air leaks to turn it into acoustic suspension.

Good luck what ever you do. Messing about with speakers is a lot of fun.

Stu

Re: project x diy speaker build

Posted: Sun Sep 24, 2017 4:15 pm
by jayman67
cheers stu,
yes i realize the tweetesr will have to go,it looks very dated to me as i sit hear looking at the ones in my yamaha ns1000's,
all the richard allan drivers are dated 1968.
re the cabinets , i would like to try the cube type cabinet or omni style with say downward facing bass diver or other variations with that in mind the original cabinets are not needed.
this is of course all ideas floating around in my head so i will wait for the richard to comment as to the best way forward. :)

Re: project x diy speaker build

Posted: Sun Sep 24, 2017 4:22 pm
by Simon Hickie
Ah! Thiele/Small parameters https://en.wikipedia.org/wiki/Thiele/Small_parameters although these drivers may predate their brochure measurements. Someone may have done some measuring. In a nutshell they offer guidance to cabinet size and type (sealed, bass reflex etc.) and usually one is interested in just the bass driver. Your cabinet looks like it has some sort of port. If you make your cabinet sealed: too small and you'll get a mid/upper bass boost but lower frequencies will be cut in output sooner (underdamped - "one note bass" - impressive on first hearing at a dealers, but ultimately tiring and inaccurate); too large and you get more but slower bass roll-off at higher frequencies, but the bass is tighter (overdamped). Critically damped is where the box size is somewhere between the two. With sealed boxes, you have quite a bit of leeway: e.g. if critically damped was 25 litres, then anything from 15 to 50 is probably fine.

I've done a quick sim of the 12 inch driver from the limited data available: It looks like around 15 litres net is optimal, but anything up to 40 looks fine. That suggests to me a two box solution like Stu's might work well: Bass driver at the bottom of say a 40 litre cab (useful floor reinforcement from a slightly overdamped enclosure, upward facing 8 incher and front facing tweeter in smaller top cabinet. It may be worth trying a different tweeter too (e.g. on of the Doc's).

All that said the raw drivers might fetch quite a bit on the world's favourite auction site. I came out comfortably ahead when I sold the drivers from a pair of B&W DM4s (which was stuffed to the gunwales and sounded awful).

Re: project x diy speaker build

Posted: Sun Sep 24, 2017 4:31 pm
by jayman67
wonderful info cheers!
ive just read the same wiki article :)
what did you use to do the sim?

Re: project x diy speaker build

Posted: Sun Sep 24, 2017 5:06 pm
by Simon Hickie
winisd beta http://www.linearteam.dk/?pageid=winisd

I find these things to constitute "broad guidance" rather than being tightly prescriptive. It just helps avoid putting a driver that needs a really large enclosure into something stupidly small and then wondering why it's all boomy with little definition. This is why the Doc's 8 inch driver works well in the Cube 1s and also a range of other donor cabinets of varying sizes.

Re: project x diy speaker build

Posted: Sun Sep 24, 2017 5:50 pm
by karatestu
The thing that puzzles me is how doping affects the T&S parameters of a driver. Surely they must change ?

Re: project x diy speaker build

Posted: Sun Sep 24, 2017 6:03 pm
by Simon Hickie
I believe it increases Qms a bit but this doesn't do much to the Qts: Qes seems to have a far greater influence on Qts. As for the other parameters, I don't know - Vas perhaps as a result of the heavier cone?