I suppose it's high time I posted an update! With the weather here in the UK finally reaching acceptable levels and my car working quite nicely again I got back to the speaker project. The other day I did the final cuts on my pieces of wood and glued the baffles up. The 2nd waveguide took some sanding to make it fit centrally but it was worth the effort (in terms of looks).
I'm very pleased with how they look so far. I'm going away for a week but soon I'll get them closer to completion and finally listened to.
Excuse the low image quality, these were taken with a £50 digicam.
Simon
I'm very pleased with how they look so far. I'm going away for a week but soon I'll get them closer to completion and finally listened to.
Excuse the low image quality, these were taken with a £50 digicam.
Simon
Attachments
Very nice!
You may remember me as the guy who suggested those Skytronics to you, all the way back in (I think) 2008!
You may remember me as the guy who suggested those Skytronics to you, all the way back in (I think) 2008!
Reminds me of a saying I've heard:
After you finish the first 90%, all you have left is the other 90%!
After you finish the first 90%, all you have left is the other 90%!
Well-proportioned baffles. Looking really nice.
However the WG seemed a little bit odd at the throat. The driver looks like APT. If so, that's a 2-slit phase plug which protrudes out of the throat of your WG. And then the sound waves coming out of the outer slit would be splitted very hard (facing a rapid flared surface), while those of inner slit would escape straight out ahead without the 'support' of the WG (too far away from the nearby surface) -- that's what I'd imagine.
So, normally the phase plug (which used for synchronizing the wavefront) is placed slightly behind the throat, so the flat wavefront (in theory) can show up right at the throat.
Just my 2c.
However the WG seemed a little bit odd at the throat. The driver looks like APT. If so, that's a 2-slit phase plug which protrudes out of the throat of your WG. And then the sound waves coming out of the outer slit would be splitted very hard (facing a rapid flared surface), while those of inner slit would escape straight out ahead without the 'support' of the WG (too far away from the nearby surface) -- that's what I'd imagine.
So, normally the phase plug (which used for synchronizing the wavefront) is placed slightly behind the throat, so the flat wavefront (in theory) can show up right at the throat.
Just my 2c.
That's funny Adam, and is really true!
CLS, thanks for your comments, I'm so glad you brought this issue up because my instincts tell me something's wrong with my waveguide-throat-driver transition.
I was hoping to upgrade to the compression drivers recommended by Dr Geddes at a later date, which I think look quite different...
What, in your considered opinion, should I do to rectify this situation? Should I build up a spacer to make the outer part of the phase plug flush with the throat of the waveguide?
Thanks,
Simon
CLS, thanks for your comments, I'm so glad you brought this issue up because my instincts tell me something's wrong with my waveguide-throat-driver transition.
I was hoping to upgrade to the compression drivers recommended by Dr Geddes at a later date, which I think look quite different...
What, in your considered opinion, should I do to rectify this situation? Should I build up a spacer to make the outer part of the phase plug flush with the throat of the waveguide?
Thanks,
Simon
You're welcome and yes you may try it, but unfortunately I think it's almost impossible to make it fit both your current APT and future other compression drivers🙁
It's because if a standard 1" bore is to maintained through the 'adaptor' for APT, then it'd be a straight tube. So the transition between the end of that tube to the WG will be inevitably very abrupt, which is a bad thing for sure.
If you try to make that transition smooth, then the entrance of WG will be larger than 1", it'll be a problem for your future upgrade.
To make things 'perfect', you'll need 2 different type of adaptors - one for APT, the other for standard 1" CD. I believe they'd be in very different depths and shapes. Is that effort worth it? It's up to you.
Good luck and tell us your progress🙂
Thanks for your response. Given the effort put in so far, yes it's worth it, at least to get as close to ideal as I can without wrecking the waveguides we've made.
If I understand correctly I cannot make a perfect transition for the APT without cutting into the current throats so I will have to "make do" with a compromised throat for the APTs and reinvestigate the situation when I have the B&C drivers.
Firstly (when I get back from holiday) I'll need to listen to the speakers and see if I can tolerate compression tweeter sound. It seems so much more lively than direct radiating soft dome tweeters but that's probably a good thing!
Simon
If I understand correctly I cannot make a perfect transition for the APT without cutting into the current throats so I will have to "make do" with a compromised throat for the APTs and reinvestigate the situation when I have the B&C drivers.
Firstly (when I get back from holiday) I'll need to listen to the speakers and see if I can tolerate compression tweeter sound. It seems so much more lively than direct radiating soft dome tweeters but that's probably a good thing!
Simon
Well I've finally sanded my baffles a bit and put all the drivers in place. I've bolted everything in and wired up a basic crossover just to make some sound.
Instantly there appeared to be some promise of a very clean sound but as we should all realise a skewed frequency response can give a very wrong impression of a speaker's true character.
I've spent the last days fettling like mad, exclusively by ear for now.
Today I've kept one speaker the same (tolerable sound) and purely worked on the other. I used a switch to test different coils on the woofers more directly, I've messed with roll-off slopes on the treble and mids, I've played with attenuation of mid and treble a lot and I've altered x-o points between all drivers and I've done one or two other things too. I've recently had some large breakthroughs in progress.
First thing of interest is, as expected, the tweeter and midrange (Beta 8a) needed huge amounts of padding to meet the mid-bass efficiency of the woofers (Skytronics in parallel on a 12mH coil). When this was set more or less correctly (something like 12dB worth of L-pad on each - done purely by ear for now) the sound became tolerable and some bass came out of hiding.
The next thing of interest was the disturbing nasal quality I was getting from the Beta 8a / APT-80 tweeter combo. Vocals and piano were not close to good enough tonally!
Looking at the Edge simulations for drivers mounted on an open baffle and considering what seems to be "common knowledge" (as I take it) of the falling response of compression tweeters mounted in waveguides it's not surprising to find problems with a lack of midrange warmth.
I tried switching the woofers to a 4.7mH coil to fill the gap up to the midrange but the sound was bloated and lacking in deep bass. This was exactly a trade-off I had issues with in my "practice" speakers. Some switching between coils revealed the 4.7mH scenario wouldn't be acceptable. I decided to use Edge's suggestion for a passive "baffle step correction" filter comprising of 4.7mH in parallel with about 7R - these are placed in line with the mids. I thought this would make for a much flatter response from the mid driver and make the transition to bass much smoother.
The result was certainly good, the midrange is now much fuller and almost spot-on. I may try a two-pronged approach and also add plenty of absorbing material behind the mid driver to attenuate the rear-wave. This should also smoothen the response, reducing the open baffle losses.
Instantly there appeared to be some promise of a very clean sound but as we should all realise a skewed frequency response can give a very wrong impression of a speaker's true character.
I've spent the last days fettling like mad, exclusively by ear for now.
Today I've kept one speaker the same (tolerable sound) and purely worked on the other. I used a switch to test different coils on the woofers more directly, I've messed with roll-off slopes on the treble and mids, I've played with attenuation of mid and treble a lot and I've altered x-o points between all drivers and I've done one or two other things too. I've recently had some large breakthroughs in progress.
First thing of interest is, as expected, the tweeter and midrange (Beta 8a) needed huge amounts of padding to meet the mid-bass efficiency of the woofers (Skytronics in parallel on a 12mH coil). When this was set more or less correctly (something like 12dB worth of L-pad on each - done purely by ear for now) the sound became tolerable and some bass came out of hiding.
The next thing of interest was the disturbing nasal quality I was getting from the Beta 8a / APT-80 tweeter combo. Vocals and piano were not close to good enough tonally!
Looking at the Edge simulations for drivers mounted on an open baffle and considering what seems to be "common knowledge" (as I take it) of the falling response of compression tweeters mounted in waveguides it's not surprising to find problems with a lack of midrange warmth.
I tried switching the woofers to a 4.7mH coil to fill the gap up to the midrange but the sound was bloated and lacking in deep bass. This was exactly a trade-off I had issues with in my "practice" speakers. Some switching between coils revealed the 4.7mH scenario wouldn't be acceptable. I decided to use Edge's suggestion for a passive "baffle step correction" filter comprising of 4.7mH in parallel with about 7R - these are placed in line with the mids. I thought this would make for a much flatter response from the mid driver and make the transition to bass much smoother.
The result was certainly good, the midrange is now much fuller and almost spot-on. I may try a two-pronged approach and also add plenty of absorbing material behind the mid driver to attenuate the rear-wave. This should also smoothen the response, reducing the open baffle losses.
Last edited:
Hi Simon. That soft wadding behind the midrange was what was needed to correct the in-room sound of the cheapies I made for friends yonks back. Of course I just used what I had available but as an old sleeping bag worked for me I can suggest a starting point.
Start with about 25mm of BAF and use some cheap cotton poly blend sheeting on both sides of the BAF batting and make if a loose fit.
Be interested in your results
Start with about 25mm of BAF and use some cheap cotton poly blend sheeting on both sides of the BAF batting and make if a loose fit.
Be interested in your results
Another considerable gain was achieved by connecting the T-bass circuit, comprising of a 500VA toroid, 8mH coil, 500uF? capacitor and 2.5R? resistor to Graham Maynard's recipe. This goes in line with the woofers before the 12mH coil and has the effect of slightly attenuating mid/upper bass but boosting deep bass from Fs by as much as 6dB. It also seems to improve the tunefulness of bass.
The only issue I have is that with my current test amp - a rather weak and grim 50WPC Sherwood receiver - playing too loud causes the amp to switch itself off. A favourite test track - a drum kit recording - can't be played at realistic levels at all: it just switches itself off 😱
The incredibly cheap woofers will distort quite readily but I think when both speakers are playing on stronger amps in my small room I'll not have issues with maximum level. If I do, then I'll have to talk my mum into allowing me to build the attic infinite baffle subwoofer I've been pondering lately... 😉
I connected the Beta 8 chassis to negative for a little extra free tweak, but maybe it should be connected to safety earth - ?
The thing that strikes me when messing with these speakers is how hard it is to get a proper-sounding midrange. It got me thinking of a fellow DIYers battles with open baffles and how he's resorted to trying all sorts of different drivers for midrange in hope of finding the missing "harmonics" that we're used to from boxed speakers. I wonder if the missing part is to some degree simply the poor frequency response from a falling mid driver response married to a rising woofer response. This can surely only be ameliorated by trying to get some kind of flattish response from the mid and woofers individually before crossing them to one another. In my case I decided that meant using a very large coil to closely mirror the woofer roll-off in reverse - just as Jamo do it, but boosted further downwards by T-bass and the expectation of the room improving matters. This is combined with the midrange being extended downwards by a compensation circuit. (The mid is still rolled off using 150uF.)
I realise I'm writing a dreary essay here so I will try to draw up a crossover schematic to debate (should anyone wish to point out daft things I've done etc.) and soon I will take some crude individual driver measurements on-axis and some overall response curves - in-room only as that's all I'm interested in. These speakers are [quite selfishly?] for my room only.
The only issue I have is that with my current test amp - a rather weak and grim 50WPC Sherwood receiver - playing too loud causes the amp to switch itself off. A favourite test track - a drum kit recording - can't be played at realistic levels at all: it just switches itself off 😱
The incredibly cheap woofers will distort quite readily but I think when both speakers are playing on stronger amps in my small room I'll not have issues with maximum level. If I do, then I'll have to talk my mum into allowing me to build the attic infinite baffle subwoofer I've been pondering lately... 😉
I connected the Beta 8 chassis to negative for a little extra free tweak, but maybe it should be connected to safety earth - ?
The thing that strikes me when messing with these speakers is how hard it is to get a proper-sounding midrange. It got me thinking of a fellow DIYers battles with open baffles and how he's resorted to trying all sorts of different drivers for midrange in hope of finding the missing "harmonics" that we're used to from boxed speakers. I wonder if the missing part is to some degree simply the poor frequency response from a falling mid driver response married to a rising woofer response. This can surely only be ameliorated by trying to get some kind of flattish response from the mid and woofers individually before crossing them to one another. In my case I decided that meant using a very large coil to closely mirror the woofer roll-off in reverse - just as Jamo do it, but boosted further downwards by T-bass and the expectation of the room improving matters. This is combined with the midrange being extended downwards by a compensation circuit. (The mid is still rolled off using 150uF.)
I realise I'm writing a dreary essay here so I will try to draw up a crossover schematic to debate (should anyone wish to point out daft things I've done etc.) and soon I will take some crude individual driver measurements on-axis and some overall response curves - in-room only as that's all I'm interested in. These speakers are [quite selfishly?] for my room only.
Hi Ted,
As I was tinkering this evening I kept having naughty thoughts involving sleeping bags and a pair of scissors, and the reason is precisely because I recall you getting good results like that.
I would like to try loft insulation as it seems to have great acoustic properties. The only issue is it's horrible stuff to touch!
Simon
As I was tinkering this evening I kept having naughty thoughts involving sleeping bags and a pair of scissors, and the reason is precisely because I recall you getting good results like that.
I would like to try loft insulation as it seems to have great acoustic properties. The only issue is it's horrible stuff to touch!
Simon
Simon I would suggest some quick cheap room treatments before you get too far.
Have you hung anything on the walls behind the speakers yet?
I have found that spending a couple of hundred dollars on room treatment is the sonic EQ of spending a thousand on drivers.
A visit to a local thrift shop and a few old woolen blankets on the wall ( or quilts) disguised by some cheap wall hangings? may be needed first.
Have you hung anything on the walls behind the speakers yet?
I have found that spending a couple of hundred dollars on room treatment is the sonic EQ of spending a thousand on drivers.
A visit to a local thrift shop and a few old woolen blankets on the wall ( or quilts) disguised by some cheap wall hangings? may be needed first.
Keep away for fiberglass if it has to be in the open, an old woolen blanket should work even better, especially if it has been washed a few times and felted down.
There are corner absorbers (quality foam) behind the left speaker, along with a pot-plant and cd racks. There's very little bare wall. Behind the right is largely open space but to the left a pot-plant and more cd storage. There's not much more I can do back there in this room.
Very soon (tomorrow?) the speakers will go into their proper positions in this room and I'll have a jolly good listen and see what's what, or hear what's what. I was really assuming I could benefit further from some rear wave attenuation, it's not simply that I have a big problem anymore...
Very soon (tomorrow?) the speakers will go into their proper positions in this room and I'll have a jolly good listen and see what's what, or hear what's what. I was really assuming I could benefit further from some rear wave attenuation, it's not simply that I have a big problem anymore...
- Status
- Not open for further replies.