My Swansong speaker build

Dedicated to those large boxes at one end of the room
User avatar
IslandPink
Amstrad Tower of Power
Posts: 10041
Joined: Tue May 29, 2007 7:01 pm
Location: Denbigh, N.Wales

#31 Re: My Swansong speaker build

Post by IslandPink »

Traditionally the problem has always been what's happening with the amp at small signal levels. Many SS amps had increased distortion at low levels.
ps. Nick - saying that your Quasars sounded OK with Chris's PA amps at Owston doesn't give the combo a clean bill of health for home use !

Scott/Colin - do the bass drivers in this scheme not use the whole volume of the (big ) cabinet, then ?
"Once you find out ... the Circumstances ; then you can go out"
User avatar
Toppsy
Shed dweller
Posts: 2400
Joined: Mon Jul 23, 2007 12:10 pm
Location: red rose country

#32 Re: My Swansong speaker build

Post by Toppsy »

Scott/Colin - do the bass drivers in this scheme not use the whole volume of the (big ) cabinet, then ?
Mark,
Not quite the whole volume of the cabinet. There is a approx. 20 litre closed box section for the midrange units and tweeter and then the small sealed volume for the Hypex amp. Apart from those 2 sections the rest of the remaining volume is used for the woofers. This applies to both the original symmetrical layout and my revised layout.

In the original symmetrical layout, the woofers each had there own separate vented volume the total equating to the whole cabinet volume less the volume for the midrange/tweeter. This remaining volume was divided equally for each the woofers.

In the new layout both the woofers share the same volume, so there is a shaft the full height of the speakers at the rear. Scott was concerned this arrangement causing a longitudinal standing wave. A Double Chamber Reflex arrangement solves this potential issue.
User avatar
Nick
Site Admin
Posts: 15706
Joined: Sun May 06, 2007 10:20 am
Location: West Yorkshire

#33 Re: My Swansong speaker build

Post by Nick »

Many SS amps had increased distortion at low levels.
Maybe, but that would, I assume require being heavy biasing into B, that's the only way I can see distortion increase at low levels. But I may be missing something. And that's a feature of push pull instead of SS as such.
Nick - saying that your Quasars sounded OK with Chris's PA amps at Owston doesn't give the combo a clean bill of health for home use
I wasn't suggesting that it did, However in the past I know when I have tried an amp with a low output impedance (I assumed that was part of the issue) with something like a Fostek in a back loaded horn the result has been peaky and apparent lack of bass compared to a valve amp driving the same speaker. That was what I was expecting and I heard far less or none of that.
Whenever an honest man discovers that he's mistaken, he will either cease to be mistaken or he will cease to be honest.
User avatar
Toppsy
Shed dweller
Posts: 2400
Joined: Mon Jul 23, 2007 12:10 pm
Location: red rose country

#34 Re: My Swansong speaker build

Post by Toppsy »

In my post #4 I mentioned the Magnesium alloy cones of the MarkAudio CHN50 drivers were a bright silver in colour and I intended to spray the cones matt black to better integrate aesthetically with the rest of the drivers.

I did spray 1 pair with a very light coating of acrylic matt black from car paint rattle cans and then knocked up 4 small sealed boxes from spare off-cuts of 6mm birch play. I mounted a pair the unpainted drivers in 1-pr cabs and the painted pair in the other.

Image

And a photo of my test bed set up.

Image

After spending some time listening comparing the 'before' and 'after' speakers last night and some further comparison listening this morning I can say the blackening of the cones to be a success. Both aesthetically and audibly. The only audible difference I can detect (and this is a really very minor difference in sound) is the unpainted speakers are a little sharp in the HF especially with female vocal. This is not evident on the painted cones. The painted cones also sound a tad richer (nay mellow). Just what one would expect with adding a very small amount of mass to the cones. Nothing to be concerned about as we are only using the drivers as midrange units. So a win situation as far as I'm concerned.

I am now going to spray the other 26 drivers. Should be fun!
User avatar
IslandPink
Amstrad Tower of Power
Posts: 10041
Joined: Tue May 29, 2007 7:01 pm
Location: Denbigh, N.Wales

#35 Re: My Swansong speaker build

Post by IslandPink »

There may be a bit of damping on the cones in the HF too though. Sounds like a good result all round.
"Once you find out ... the Circumstances ; then you can go out"
vinylnvalves
Old Hand
Posts: 1074
Joined: Sat Jan 26, 2013 12:08 pm

#36 Re: My Swansong speaker build

Post by vinylnvalves »

If you interested I have a number of sheets of 5/8” birch BB grade ply in 5 ft x 5 ft sheets. I could let you have for a good price. I know it’s not quite 18mm, but when veneered wouldn’t be far off. Depending on how quickly you want to start this project, could drop off at Owston ( cut down of course)
User avatar
Toppsy
Shed dweller
Posts: 2400
Joined: Mon Jul 23, 2007 12:10 pm
Location: red rose country

#37 Re: My Swansong speaker build

Post by Toppsy »

vinylnvalves wrote: Mon Jul 25, 2022 12:56 pm If you interested I have a number of sheets of 5/8” birch BB grade ply in 5 ft x 5 ft sheets. I could let you have for a good price. I know it’s not quite 18mm, but when veneered wouldn’t be far off. Depending on how quickly you want to start this project, could drop off at Owston ( cut down of course)
Thanks for the kind offer, much appreciated. I am mindful of the current cost of birch BB grade ply. However the local place I buy all by sheet wood offer a full machining service - CNC router table and CNC controlled flat bed beamsaw - all at very reasonable cost, or it was last I used their services. For CNC cutting a full 8' x 4'' sheet into panels on their flat bed beamsaw they only charge me £10. A bit of a no brainer really. Saves me hours and I get very accurate cut panels. Makes the building much more pleasurable.
vinylnvalves
Old Hand
Posts: 1074
Joined: Sat Jan 26, 2013 12:08 pm

#38 Re: My Swansong speaker build

Post by vinylnvalves »

At only £10 to cut the sheets on a beam saw is very cheap. If that includes CNC cutting the holes, how are they still in business. Yes the extra cost of material becomes insignificant when all the hard work is done for you. I have to “pay” my offspring to help me move 10” x 5” sheets, probably more than £10
User avatar
Toppsy
Shed dweller
Posts: 2400
Joined: Mon Jul 23, 2007 12:10 pm
Location: red rose country

#39 Re: My Swansong speaker build

Post by Toppsy »

Might not have described the service they provide very well. :(
The £10 charge is for only the cutting of a single sheet into panels. Perhaps there was an element of 'mates rates' as I have a good rapport with the guys there. Takes less than 10 mins to program and cut a full 8' x 4' sheet on their beamsaw. No cutting the holes or any other cut-outs just straight line cuts. I just give them a cutting schedule of the panel sizes and number of each. I do have to buy full sheets and the cutting cost is per sheet. At todays rates I fully expect the cost of the cutting service to be at least double?

For driver cut-out holes and other cut-outs these are done on the flat bed CNC router table. The cost of this service is charged on an hourly basis based on machining time. I send them a DXF copy my CAD drawing and they quote me a price for the work.
User avatar
Scottmoose
Needs to get out more
Posts: 1802
Joined: Sat Sep 08, 2007 11:03 am
Contact:

#40 Re: My Swansong speaker build

Post by Scottmoose »

Nick wrote: Mon Jul 25, 2022 12:02 pm...in the past I know when I have tried an amp with a low output impedance (I assumed that was part of the issue) with something like a Fostek in a back loaded horn the result has been peaky and apparent lack of bass compared to a valve amp driving the same speaker. That was what I was expecting and I heard far less or none of that.
Sorry couldn't do much earlier, half-wrote a post & had to go on taxi duties. And apologies for the staccato prose below, as I'm writing mostly off-the-cuff.

We're simply in the realms of system matching. An amplifier being capable of high power output taken purely in & of itself (and setting everything else aside for a minute) is not an issue for high efficiency speakers. 'High efficiency' in this context is also just a generic / unitless term for a speaker capable of a certain output level. Say, for e.g., 95dB 1m/2.83v. Exact figure doesn't really matter, it's just a general principle.

A lot of HE[ish] single-driver systems are derived from early drivers going back to the '40s which had huge motors. Main reason was because as most here know better than me, power was expensive and you didn't generally have a whole lot of it. The big motors also usually (not invariably, but usually) meant the drivers had a huge amount of electro-mechanical damping, i.e. they had low - very low Q values. Partly innate, partly by intent because a lot of the available amplifiers had either a relatively high or adjustable output impedance. Since the amplifier output impedance, wire series R and loudspeaker form a system, the result is what you'd expect: the one counters the other, so the over-damped bass characteristic on a pure voltage source (which the speakers weren't designed to be used with) suddenly goes away -it's basically raising the effective driver Qts, and thereby increasing bass output.

About 20 years back, Fostex demo'd the then-new FE208ESigma + its horn on the end of some 100w Accuphase poweramps; not a great natural match as the Accuphase had way too low an output impedance. So the driver designer built a pair of lead-shielded tungsten speaker wires which put about 4ohms of series R into the circuit, providing the same effect of lowering the damping / raising Q, balancing the response to what was intended (the drivers were designed with the expectation that they'd mostly be used on the end of amplifiers with an output impedance in the 2.5ohm - 4ohm region, give or take).

That's just one set of conditions though, mostly relating to a sub-set of high efficiency wideband drivers with very low Q values (say, below < 0.2). They don't all have that of course. And there are plenty of other types of HE systems, especially multiway, that aren't designed for or don't benefit from a high output impedance amplifier. Most '70s era JBLs for e.g., even the big ones. Your baffles as I recall have a set of Supravox 285s and Lowther PM6As. Neither have overdamped Q values (and the Lowther is being used as a mid-tweeter anyway, so that isn't overly relevant) -according to the published data 0.33 and 0.343 assuming the 8ohm silver coil version of the latter. That's at the lower end of middling, similar to a lot of modern midbass units, and fine for a regular solid state amplifier with a low output impedance, high power or otherwise. HE speakers tend to put a microscope on flaws up the chain like noise etc., and other issues might crop up. But as a broad point, it's really just about matching the driver and amplifier characteristics (and preferably designing enclosures as relevant with that in mind).
'"That'll do," comes the cry of the perfectionist down the ages.' (James May The Reassembler)
Website www.wodendesign.com
Community sites www.frugal-horn.com & www.frugal-phile.com
User avatar
Nick
Site Admin
Posts: 15706
Joined: Sun May 06, 2007 10:20 am
Location: West Yorkshire

#41 Re: My Swansong speaker build

Post by Nick »

and the Lowther is being used as a mid-tweeter anyway
Though it is being driven full range at the low end at least. Just FYI. And it was the 16R one, though that is partly a hangover from trying to get the series xover working.
Whenever an honest man discovers that he's mistaken, he will either cease to be mistaken or he will cease to be honest.
User avatar
Scottmoose
Needs to get out more
Posts: 1802
Joined: Sat Sep 08, 2007 11:03 am
Contact:

#42 Re: My Swansong speaker build

Post by Scottmoose »

Underlines the point even more then, since the 15/16ohm siver coil PM6A has a nominal Qts of 0.381 (well into bass reflex loading territory) so isn't inherently overdamped & doesn't need any help from the amplifier. Running them as a 1.5 way with just a low-pass on the Supravox will also help in this as it's additive below the corner frequency. So I'm not all that surprised it worked OK / there weren't vast changes. Nice when it happens!
'"That'll do," comes the cry of the perfectionist down the ages.' (James May The Reassembler)
Website www.wodendesign.com
Community sites www.frugal-horn.com & www.frugal-phile.com
chris661
Shed dweller
Posts: 2556
Joined: Mon Feb 22, 2010 7:29 am
Location: Sheffield

#43 Re: My Swansong speaker build

Post by chris661 »

I'd written up a reply, and then got logged out before I managed to send it.

Here's V2.0.


A quick note on the amp at Owston, Nick's OBs, etc:
- The amp was a Crown MA5002VZ, and has been well looked after. In its day, it was a very good amplifier. I wasn't particularly surprised, then, that it held up well.
- I fired up the crossover simulator, and the OBs stay above 8ohm above 20Hz, with the exception of a narrow dip to 7ohm around 100Hz. ie, a bit of series resistance shouldn't bother them much compared to speakers that drop to 3ohm in places.


With regards to Colin's speakers, I have a few things I'd suggest. Colin, you're welcome to make use of these, or discard, as you like:

- If it was mine, I'd go for a sealed box for the LF drivers. Rooms in the UK can be very helpful in terms of LF gain (wall/corner loading & cabin pressure), so I don't think a bass reflex (with the compromise of a sharply-defined cutoff frequency) would bring any benefit.
In short, you don't need the extra output, and you might be sacrificing very-low-frequency extension.

- A coaxial tweeter/mid-array might be worth looking at. It'd reduce the effective centre-to-centre distance compared to the MTM design currently shown.

- The Hypex DSP amps look good, and I'm sure programming them will be straightforward enough.


Looking forward to seeing these in person once they're built.

Chris
User avatar
Toppsy
Shed dweller
Posts: 2400
Joined: Mon Jul 23, 2007 12:10 pm
Location: red rose country

#44 Re: My Swansong speaker build

Post by Toppsy »

- If it was mine, I'd go for a sealed box for the LF drivers.
Apologies for not responding earlier Chris.

Having discussed with Scott sealed v vented for the LF drivers for these speakers I will build them initially as vented. This gives the option of running them sealed if desired. I have the option to damp (fill or even blank off) the vents and convert to sealed / lossy sealed ('aperiodic') and the LF volume we currently have would be (according to the published driver data) on a near-Bessel alignment under those conditions, so fine as far as suitability for being run sealed goes. A case of suck it and see during the listening tests once the speakers are built.
- A coaxial tweeter/mid-array might be worth looking at. It'd reduce the effective centre-to-centre distance compared to the MTM design currently shown.
Thanks for the suggestion Chris. interesting as that avenue might have been I purchased the MarkAudio mid-array drivers and the SB Acoustics Beryllium dome tweeters a few weeks ago and so am committed to the as drawn driver arrangement.

Must get back to spraying matt black the last couple of CHN50 drivers silver Magnesium cones.
User avatar
izzy wizzy
Old Hand
Posts: 1496
Joined: Fri Nov 02, 2007 7:02 pm
Location: Auckland NZ
Contact:

#45 Re: My Swansong speaker build

Post by izzy wizzy »

I have a question that arises every time I see folk mod speakers with dope , paint or other such treatments. What were the designers/manufacturers doing or thinking so that they didn't use these things to "improve" their drivers? In this case, does shiney look better in the marketing blurb i.e. Looks sell, modded drivers not so much? Or is there something else at play?

Following this thread keenly and soaking up the thoughts going into the design. Great stuff.
Post Reply