Where to get rabbet for 1/16"?

[jmansion]

Hi,

I've had a Bosch POF500A in the cupboard for a long time (don't laugh! I'm a noob!) and I'd like to use it for something. Its first use in fact.

I have acquired an old pair of AR19B speaker cabinets, no drivers etc.

The woofer holes are rebated, and I think about 169mm on the outer rebate and 144mm on the cutout.

I'd like to be able to enlarge them, slightly, ideally by about 1/16".

Can anyone tell me where to find a 1/16" rabbet with 1/4" shank (is that the right term?) so I can rabbet the driver cutout, and then cut off the bit that was left as guide?

I figure that having done so, a 1/2" rabbet would then make the outser rebate a bit wider, and I'd be all set for more modern speaker drivers. I have a choice at 170/171mm outside diam. I have some old Seas drivers nominally 170-mm+-0.4mm and they don't fit.

As an alternative, same strategy with a 1/8" rabbet would probably be OK.

I'd rather not buy one of the multi-part sets with adjustable collars etc if I can help it, not least they all seem to have 1/2" shanks and they are costly. The cabinets were only 10 quid, after all.

Thanks
james

 

[marcros]

i would clamp a straight edge as a fence and then just use a straight plunge bit- if what i think that you are trying to do is correct. A sketch and take a photograph of it would help explain it a bit better.

 

[sploo]

You could make a simple circle cutting jig and cut a hole of the right size out of a scrap bit of MDF. Then clamp it to the baffle and use that with a bearing guided template bit to open out the hole.

Make sure the cabinet is right for the drivers and crossover. The box volume matters, as does the front baffle width.

EDIT: typo

 

[Phil Pascoe]

"I've had a Bosch POF500A in the cupboard for a long time (don't laugh! I'm a noob!)I've had a Bosch POF500A in the cupboard for a long time (don't laugh! I'm a noob!)"

Cracking little tool - no one who's had one will laugh at you. You might end up with a bigger one, but I bet you'll keep that one.

 

[Eric The Viking]

Wot 'ee said.

I have both a POF 500A and a POF 600A - the 500 A was my first router, too. I still use it from time to time, and the 600A is the go-to one for small handheld jobs.

Sploo had the right idea, I think: make a template from MDF. Otherwise, you might actually block the hole up first, by clamping or screwing a slab of, say, MDF behind, so you can get a real centre point to use a trammel or centre with the router. Assuming you have all the accessories, I think a circle-cutting whotsit came with the POF500A, or at least a pin that you can swing it around.

Also check that you can't remove the whole baffle board. It will make the job a lot easier if you can just fix it down to something sacrificial underneath, but, better still, you could make an entire replacement baffle from scratch pretty easily. The advantage being that you haven't ruined the original if you change your mind later or you want to sell them on. MDF is a pretty good choice as it doesn't resonate much and it's dense.

Regarding a rebate cutter, Wealden do nice ones (I have one in 1/2" shank). You might get the rebate you need just with a bearing running round the inner hole. Work out how big the rebate needs to be - Wealden has a table on their web site telling you which bearing you need for a given cutter and required rebate. You have a roughly 1/2" rebate already, so you might find it difficult to get a big enough cutter in 1/4" shank, but the POF 500A will take an 8mm collet too, so Wealden might have a larger cutter available. Try Miles Tools (Yeovil) for the collet as a spare part.

E.

 

[sploo]

One other idea (if you do have even a cheap rebate cutter) - companies such as Bearing Boys do a huge range of bearings - you might find one just the right inner and outer dimension to give your cutter the right "width" of rebate. At a push, you could even wrap a little bit of tape around the bearing to adjust the width. Not ideal, but for a small number of carefully done cuts it might work.

I've just spotted your comments about the cabinets being 10 quid. TBH you're probably at the point where you might be better off purpose building a pair of well braced cabinets from MDF. It's not expensive, and you can get exactly the right size/width for the drivers (matched to your crossover's baffle step).

What are the Seas drivers you've got, and what tweeters are you using?

 

[Phil Pascoe]

An 8mm collet for a POF 500A? I learn something every day. Mind, the collet would probably cost what I paid for the router.

 

[Eric The Viking]

From Miles' site:

703 - COLLET Ø 8 MM / Ø 10 MM (BSD-2608570049.A20) £13.07 £15.68 (incl.)

http://www.mtmc.co.uk/Spare-Parts/B...__p-240-12733138-12733209-12734811-88432.aspx

I bought one ages ago when I refurbished mine, but haven't used it yet. Note that there are two sizes of collet nut - only the newer, 18.5mm one will take 8mm collets (10mm shaft socket for the collet). I think there is an older model that only goes to 1/4".

There is also a 6mm collet for both sizes of nut/shaft.

 

[jmansion]

I think on further reading that the 'cut a circle of scrap to fill the hole, screw it to the box, and then route a bigger outside rebate' is what I'll have to do. I do have one of those Trend multi-base things with circle jig feature. From my reading, the Jasper jigs won't fit this router.

And I guess I can be more brutal in enlarging the cutout into the cabinet, especially if its only a couple of milimetres to remove..

> by clamping or screwing a slab of, say, MDF behind,

Note that the cabinets are complete and glued together - can't easily attach anything to the back of the holes.

> Make sure the cabinet is right for the drivers and crossover. The box volume matters, as does the front baffle width.

Well, I'm fairly cool about a driver that will give a Q anywhere between 1 and 0.5. And the baffle width is what it is and I'll see how that affects the crossover later. The crossover will be DSP, so I can correct Q relatively easily, and I'm not working to an existing crossover so the width and driver positioning aren't something Have to replicate. Its annoying the tweeter is central, I'd have offset it.

> you might be better off purpose building a pair of well braced cabinets from MDF

Well I plan to eventually, when I have a little more experience. And in fact I may yet start with my Big Cheat Idea, which is just a piece of MDF with a hole in it and a Beyma VM100 midrange cup on the back. And a bit of scrap to prop it up. The driver is a Seas P17REX coax. I seem to have a cupboard full of them, with a dodgy-looking series crossover glued on by Shaun Williams (from who I bought them, a good 10 years ago - but we decided not to have ceiling speakers in the end). I believe the Beyma has about 2.6l which is OK for a surprising range of drivers, especially if a Linkwitz Transform is available. (I have ordered a bass/sub cabinet pair from Wilmslow for a couple of rather mean looking ScanSpeak basses I pickup from ebay too, so I'll practice glue and screw on those).

For the AR cabs, I don't have any suitable drivers really, but I could either widen them and use some old Vifa or Fountek drivers I have, or just buy some XT19s. In fact I quite fancy the XT19s and some SB-Acoustics drivers - they're not expensive and both have almost flat impedance, so I might even be brave enough to try a passive crossover sometime. I woke up one day and decided that 'OK' sound was good enough, but I liked playing with new stuff, so I figured a DSP crossover would let me have New Stuff for beans if I could learn how to cut wood. ;-)

 

[Eric The Viking]

> by clamping or screwing a slab of, say, MDF behind,

Note that the cabinets are complete and glued together - can't easily attach anything to the back of the holes.

Of course you can!

The main purpose is as an immovable centre point for scribing circles. Any piece of MDF that will fit through will do, if it's longer than the hole is wide!

A wise Hobbit will also arrange a pad stuck onto the middle of said bit of MDF, exactly the same thickness as the baffle. This supports the other edge of the router base nicely whilst routing your circle.

Given you ought to be able to reach through the tweeter aperture to get to the back, you might also use a small bolt in the centre to pivot the router base on, thus removing most of the reasons for slipping and making a mess of it. AND you don't necessarily need to be exact about the centre if the new hole and rebate will be larger...

... and if you use something like D/S sticky tape and the overlap is enough, you might just be able to rout the new aperture in its entirety without removing the support piece.

Aside: are these things ported? I put new bass drivers in my Monitor Audio MA7s* recently, and when I did, I ran a bearing-guided roundover bit around the port mouth. It's made the bottom end noticeably smoother and more subtle.That may not be what you want, however...

E.

*Now almost totally rebuilt - subtly upgraded xover, Audax tweeters and now new bass units. I can't remember how they sounded when new, circa 1979.

 

[sploo]

Well, I'm fairly cool about a driver that will give a Q anywhere between 1 and 0.5. And the baffle width is what it is and I'll see how that affects the crossover later. The crossover will be DSP, so I can correct Q relatively easily, and I'm not working to an existing crossover so the width and driver positioning aren't something Have to replicate. Its annoying the tweeter is central, I'd have offset it.

...

Well I plan to eventually, when I have a little more experience. And in fact I may yet start with my Big Cheat Idea, which is just a piece of MDF with a hole in it and a Beyma VM100 midrange cup on the back. And a bit of scrap to prop it up. The driver is a Seas P17REX coax. I seem to have a cupboard full of them, with a dodgy-looking series crossover glued on by Shaun Williams (from who I bought them, a good 10 years ago - but we decided not to have ceiling speakers in the end). I believe the Beyma has about 2.6l which is OK for a surprising range of drivers, especially if a Linkwitz Transform is available. (I have ordered a bass/sub cabinet pair from Wilmslow for a couple of rather mean looking ScanSpeak basses I pickup from ebay too, so I'll practice glue and screw on those).
...

A Q of 0.5 to 1 is quite a wide range, but you've confused me a bit; the P17REX contains both a midbass and tweeter, so there's no requirement for a separate tweeter.

If you're worried about tweeter placement for acoustic reasons then I'd be more concerned about baffle step issues (than the tweeter placement). If you're worried about tweeter placement for aesthetic reasons then ok, but then it does sounds like a new box would be easier.

If you're doing an active electronic (rather than passive electrical) crossover then you do have much more scope - but you would then need separate amps for the midbass and tweeter, surely?

An open baffle system is good (check out Linkswitz's designs), but there are issues with drivers in free air that need to be understood (to avoid damage). You also tend to need a lot of driver surface area to get decent bass. That said, I hacked together a fun open baffle prototype using a Peerless XLS10 and a HiVi B3s some years ago - it sounded better than it had any right to.

Wilmslow cabs pretty good. ScanSpeak drivers generally very good. Ironically, my main speakers are a Wilmslow design, using ScanSpeak drivers (though I built the cabs myself).

 

[jmansion]

> A Q of 0.5 to 1 is quite a wide range, but you've confused me a bit; the P17REX contains both a midbass and tweeter, so there's no requirement for a separate tweeter.

Yes it does. I'm not planning to use that driver, in the box I bought. But I thought it would be representative of the size of frame that Seas used for their 17cm products, some of which are still available as NOS. I will try the coax in the 'slab of MDF with midrange cup'.

That Q range is quite easily corrected with a Linkwitz transform. These drivers all run out of xmax before they run out of power handling (or my amps run out of power).


>If you're worried about tweeter placement for acoustic reasons then I'd be more concerned about baffle step issues (than the tweeter placement). If you're worried about tweeter placement for aesthetic reasons then ok, but then it does sounds like a new box would be easier.

I expect to deal with bafle step.

>If you're doing an active electronic (rather than passive electrical) crossover then you do have much more scope - but you would then need separate amps for the midbass and tweeter, surely?

I have a collection of some-decades-old stereo power amps, yes. In fact the speakers I currently have hooked up are running off an Audiolab 8000P that must be 30 years old - its grey! And at some stage I thought I'd try using multi-channel audio over HDMI and see if the jitter is really so bad as to wreck it. If its listenable, then a graphics card and a receiver actually provides quite a cheap way into multichannel active.

And I have a Terratec DMX 6-Fire I can use to run a crossover now.

 

[sploo]

> A Q of 0.5 to 1 is quite a wide range, but you've confused me a bit; the P17REX contains both a midbass and tweeter, so there's no requirement for a separate tweeter.

Yes it does. I'm not planning to use that driver, in the box I bought. But I thought it would be representative of the size of frame that Seas used for their 17cm products, some of which are still available as NOS.

Ah. Understood.

That Q range is quite easily corrected with a Linkwitz transform. These drivers all run out of xmax before they run out of power handling (or my amps run out of power).

That was my worry (extreme filtering pushing the driver when it's better to get the right sized box). Sounds like you know all that already so no worries.

I have a collection of some-decades-old stereo power amps, yes. In fact the speakers I currently have hooked up are running off an Audiolab 8000P that must be 30 years old - its grey! And at some stage I thought I'd try using multi-channel audio over HDMI and see if the jitter is really so bad as to wreck it. If its listenable, then a graphics card and a receiver actually provides quite a cheap way into multichannel active.

And I have a Terratec DMX 6-Fire I can use to run a crossover now.

IMHO The drivers themselves (non linearities, off axis response), the box (resonances), and the speaker/room interaction (comb filtering) create way more problems (orders of magnitude worse) than the difference between some old amp and something fancy, so that sounds like a perfectly sensible solution.

 

[Eric The Viking]

(Wildly off topic for general woodworking)

... the difference between some old amp and something fancy...

Rubbish is always rubbish, of course, but good old amps are much better made than modern stuff, if only because the PSUs tend to be better at transient delivery.

If I had a pound for every piece of junk I've come across that cracks and can't drive high-Z loads, etc. Good quality older electrolytics are also much longer lived than the modern high charge-density types. Newer isn't always better.

Completely agree about tweeter resonances - also caused by the clip on for the fabric cover, but easily fixed by simply bevelling the inside of the fabric frame a bit ad/or adding something like felt around the inner edge.

E.

PS: They weren't always good. For example, the Mission Cyrus amp couldn't cope with LS3/5As, because of the unusually high impedance, and there were any number of MOSFET designs out there that settled down happily at 30kHz (as transmitters). New tweeters, anyone?

 

[sploo]

Completely agree about tweeter resonances - also caused by the clip on for the fabric cover, but easily fixed by simply bevelling the inside of the fabric frame a bit ad/or adding something like felt around the inner edge.

By resonances I was actually meaning the panels of the enclosure vibrating - on poorly made boxes they can produce more sound at certain frequencies than the drivers :shock:

I suspect that tweeter issue would come under the category of diffraction (in itself a whole bag of fun).

The daft thing with home audio is that people chase minuscule (sometimes arguably imperceptible to the human ear) improvements, whilst ignoring the 30dB nulls that comb filtering is causing because they've shoved their speakers next to a wall. Sadly home audio is one of the most BS laden subjects around; though there are good guys such as Sean Olive and Siegfried Linkwitz.

 

[jmansion]

For example, the Mission Cyrus amp couldn't cope with LS3/5As, because of the unusually high impedance

As it happens, I have a pair of Cyrus Power stereo power amps, and I was going to use one per side for the mid and tweeter. I don't have the addon-PSUs so they are a bit weedy compared to the Audiolab.

I'm not expecting to have too much problem with impedance issues so long as I don't try building an MTM wih 4 ohm woofers.

In any case, if I get hooked on this stuff, its only a matter of time before I buy Hypex power amps (or LCAudio, or ColdAmp, etc).


I quite agree about the BS factor, and its why I figured I'd rather have a toolkit of toys I can use to play with things and stay out of hi-fi shops.

The whole thing was on hold for years until I watched the demo video here: http://www.lupisoft.com/ekio/ and then I thought 'that looks fun!'.

 

[sploo]

For example, the Mission Cyrus amp couldn't cope with LS3/5As, because of the unusually high impedance

As it happens, I have a pair of Cyrus Power stereo power amps, and I was going to use one per side for the mid and tweeter. I don't have the addon-PSUs so they are a bit weedy compared to the Audiolab.

I'm not expecting to have too much problem with impedance issues so long as I don't try building an MTM wih 4 ohm woofers.

In any case, if I get hooked on this stuff, its only a matter of time before I buy Hypex power amps (or LCAudio, or ColdAmp, etc).


I quite agree about the BS factor, and its why I figured I'd rather have a toolkit of toys I can use to play with things and stay out of hi-fi shops.

The whole thing was on hold for years until I watched the demo video here: http://www.lupisoft.com/ekio/ and then I thought 'that looks fun!'.

EKIO looks interesting. I've seen Linux tools but haven't come across that one.

Any benefit to using one amp for the woofers and one for the tweeters so each one sees a similar load on both channels?

 

[Eric The Viking]

Any benefit to using one amp for the woofers and one for the tweeters so each one sees a similar load on both channels?

My own main speakers are BBC LS 3/7s driven by Quad 303s. My absolute favourite speakers are LS 5/8s which originally had Quad 405s, later Chord (post 1992 approx). In both cases it's one amp per channel (i.e. per side), with an add-on electronic crossover (big, complex board for the 303s, smaller one for the 405s). The asymmetry seems to cause no problems at all.

I'm not sure it matters much in practice, probably exactly the opposite. There are big issues with current delivery on transients, and having one power supply, which usually drives two channels, actually doing just one, probably helps.

I also have a 'cooking' MOSFET stereo power amp which has everything but the toroidal transformer separate - two sets of rectifiers and reservoir caps, etc. It has very good imaging as a consequence, as there is no 'suck' when loud things happen on just one channel.

So personally I'd use one amp per channel.

On tweeter resonances: having a slightly recessed baffle, especially in bookshelf speakers, is a well-known source of tweeter problems - the lip of the case forms a resonant cavity (yes, really!). It's one reason for the four felt blocks round the tweeters of the LS 3/5A (the other is to modify the tweeter output I believe, so it's slightly less harsh).

Sorry if I was rude about the Cyrus amps. They're probably fine driving 8 Ohm loads or lower, just not anything that needs higher voltages.

E.

 

[jmansion]

Any benefit to using one amp for the woofers and one for the tweeters so each one sees a similar load on both channels?

[/quote]

My end game is a 3-way, the Audiolab will effectively run a stereo sub array, and there will be a Cyrus Power on each side running midbass and tweeter. The required load will be quite modest, I think.

I don't think the amps - despite being some way from state of the art - will be the limiting factor at all until I've had an awful lot of practice with crossover design.

 

[sploo]

My own main speakers are BBC LS 3/7s driven by Quad 303s. My absolute favourite speakers are LS 5/8s which originally had Quad 405s, later Chord (post 1992 approx). In both cases it's one amp per channel (i.e. per side), with an add-on electronic crossover (big, complex board for the 303s, smaller one for the 405s). The asymmetry seems to cause no problems at all.

I'm not sure it matters much in practice, probably exactly the opposite. There are big issues with current delivery on transients, and having one power supply, which usually drives two channels, actually doing just one, probably helps.

I also have a 'cooking' MOSFET stereo power amp which has everything but the toroidal transformer separate - two sets of rectifiers and reservoir caps, etc. It has very good imaging as a consequence, as there is no 'suck' when loud things happen on just one channel.

So personally I'd use one amp per channel.

On tweeter resonances: having a slightly recessed baffle, especially in bookshelf speakers, is a well-known source of tweeter problems - the lip of the case forms a resonant cavity (yes, really!). It's one reason for the four felt blocks round the tweeters of the LS 3/5A (the other is to modify the tweeter output I believe, so it's slightly less harsh).

Sorry if I was rude about the Cyrus amps. They're probably fine driving 8 Ohm loads or lower, just not anything that needs higher voltages.

E.

I suppose an active crossover achieves the decoupling of the drivers and amp (so you don't get back-EMF problems from one driver to another, as with a passive system).

I've always wanted to go active, and I understood that one of the useful benefits was being able to purpose amps for the drivers (i.e. much less power required for the trebles). That said, if you have a set of amps already, and they're good for any of the drivers then I suppose it's not a big deal.

Tweeter resonances. Interesting. I don't recall seeing any AES papers on that (doesn't mean to say there aren't any). I do wonder if it's at least partly related to diffraction though. In any event, recessed drivers or sharp edges aren't advisable - though TBH the size of radius you'd need on a cabinet edge to eliminate diffraction over a large (audible) frequency range is pretty impractical.

My end game is a 3-way, the Audiolab will effectively run a stereo sub array, and there will be a Cyrus Power on each side running midbass and tweeter. The required load will be quite modest, I think.

I don't think the amps - despite being some way from state of the art - will be the limiting factor at all until I've had an awful lot of practice with crossover design.

Good crossover design is something of an art - as is good pairing of suitable drivers. The Zaph Audio site contains good narrative on some of his designs; detailing why he chose specific drivers, and how a "good" driver with problems in certain frequency ranges can be OK if paired appropriately.

That said, I suspect it's a whole load easier using a PC based active crossover than the black magic of fiddling with passive electrical filters (where one part of a circuit can affect another). There's certainly more scope for bludgeoning a decent driver with quirks (e.g. resonance peaks) than it is with passive filtering anyway.