My father's speakers. A more general opinion of them was accurate. Also a cute trick mention in an independent anechoic review. Select high tweeter setting and add a bit of bass boost. F6 moves from 40Hz to 30. No objectionable general bass boost. I've heard all sort though them and also attended all sorts of concerts even opera. Rock 😉 well someone buys their new hi fi gear and many go to listen. Pink Floyd might come out. In some respects the very lows are rather underwhelming. They needn't be but they have never been intended to be loud.. Our ears are not sensitive enough. I have read that 110db doesn't sound loud at 20Hz and in any case needs 1kw to achieve it. Probably what was going on at a firework event I attended recently. The music sounded awful and well short of getting down to 20Hz,
🙂Anyway these speakers formed my opinion of what is needed for music. 😉 I didn't know that crossing over at 1500Hz might not be a good idea. F3 at 30Hz may be better but chasing 20Hz ????
Accurate. Tricky term. My father in law, an excellent pianist wanted something to listen to his records with. He chose a Pioneer set up. They play with the response. More extreme than Wharfedale often did - badly. It's a tonal thing. Warmer, richer, low etc. Yamaha still do similar in a different way. This sort of "distortion" was often noticeable when we went around listening to loads of different speakers even some rather expensive ones.
Boxsim V2 optimiser is a bit of a handful. I've not tried the earlier version but have it from your upload. So far I have failed to get the allowed frequency range per speaker to work. Some sliders offer no - max level may need this setting unless a deep bass shelf is ok. It will as you mention optimise parts out by providing crazy values. Educational really. Say I go one order higher on an xover due to cone break up. The reason it gets optimised out is that it still causes problems and offers no clear advantage. Probably spoils phase as well. 😉 The abort button can be useful. It might make a very large change to impedance for very little gain over what it achieved before this point. Phase. When I get to something to actually build I'll post for opinions. This is for one I don't intend to build. Interested in the tweeter and a particular xover design for it by Visaton.
Xover 2154Hz. My main problem is what to add to improve - such as more phase comp and say notch filters when needed. Idea's crop up from the boxsim projects site 😉 but sorting what does what isn't that easy and different ideas get used.
@AjohnL.
Various BPJ files, of varying quality are found here:
https://boxsim-db.de/kategorie/systeme/zwei-wege/
Just put them in the projekte file and run them. A quick start!
Various BPJ files, of varying quality are found here:
https://boxsim-db.de/kategorie/systeme/zwei-wege/
Just put them in the projekte file and run them. A quick start!
Yes thanks I picked that up from your earlier post in the boxsim thread. I've found the mention of chassis you have used useful as well.
I've tried another use of the projekte files. Use one of it's xovers for a tweeter. LOL Standard values. I find that instead of crossing at 3k it crosses a bit over 2. This turns out to be down to the need to reduce the output to match this woofer. Maybe an L pad would fix that rather than just a series r. The optimiser has no problems tuning the woofer to finish up with a close to a 2nd order LR as the driver allows. Phase as I posted which appears to be pretty good. How to improve. If the woofer is too fast easy - stick a capacitor across it. If too slow the dominant effect is the woofers xover's cap. Reduce it and a better match can be achieved and the optimiser used to set the inductor. End result a small change in xover f. There will still be a departure at ~4kHz or so down to the woofer. Notch filter or is the response flat enough anyway?
Later comes reality when the speaker is built and tested. Will the phase etc be as good as the simulation suggests? A probably wise bird suggested buy greater inductance values than the sim suggests also be prepared to tune your reflex box.
There are 2 problems with the optimiser. It wont add bits and it may set xover points where ever it likes by chasing phase. It wont be the only F where that is good enough, This can be partly cured by calculating values but due to impedance they wont be in the correct place especially on a woofer. A tweeter is likely to come out closer.
A conversation with @AllenB who uses an xsim type approach suggests that it's possible to train yourself to think in the right way - ie what mouse wheel tweak on a part will do. There is also a need to know what parts to add over the basics to fix issues. He recently posted a rule of thumb concerning factors to change in an xover when the impedance of the driver is changed. The only way I know that this can be done is due to that post. 😉 It leaves me wondering if there are others say to change an xover F based on one that works but isn't where it ideally should be. The optimiser can polish things up pretty well and manual changes can still be made and it used again to polish it up even more.
So I am trying to train my thinking to allow an optimiser to work. Looks like it can be done and IMHO quicker than trying to learn to use the Xsim way. 😉 Not that I have a design that I intend to build yet. There is some more learning to do. Copy a projekte - no way. Use these for xover ideas fine. I've picked up a couple of ideas for correcting things with additional bits so far. There are probably others.
This is a Selah Audio design of some merit, IMO. I have heard this sort of 6" MTM Scan plus ribbon speaker (with separate subwoofers to make excursion less of an issue for the 6" units, and the speakers were wall mounted and the walls lined with 3" of gyproc) and it can be superb:
Not cheap! But does very good things with dispersion and room acoustics. Of course you must best listen on axis.
Probably two regular tweeters in the middle would do much the same. An MTTM.
Not cheap! But does very good things with dispersion and room acoustics. Of course you must best listen on axis.
Probably two regular tweeters in the middle would do much the same. An MTTM.
Thanks for that one. I could have knocked up something similar in boxsim but nice to see some one else's work. The results are interesting. 30L to achieve what the AR-6's could do in ~15 thanks to a low Fs woofer. 🙂 Must admit if I could simulate an AR-6 I'd probably want to do more but then comes ears. I've attached the report for anyone who is curious. Interesting magazine. Reliable circuit designs and more interest in audio than some - also decent tests
Whoops attachment too big. The article is in this edition
https://www.worldradiohistory.com/UK/Electronics-Today-UK/70s/Electronics-Today-1972-11.pdf
It suggests 3 way could be better,
@AjohnL, enjoyed reading that. 1972. It was a glorious time!
My college room-mate built a Sinclair 60 kit system, in a cardboard box as it goes with no regard for safety! It was awful, but fun to build... cheap too!
You wouldn't see adverts like this these days:
He's thinking: "She smokes, she drinks, she likes listening to my Rock records in Quadraphonic! If she likes football too, she will be perfect!" 😆
Nor this level of sexism:
Good review of the AR6. The AR4a was a goodie too. Obviously done by a proper engineer, not some hack taking a bribe for a good review on a HiFi magazine. I knew these people. 😎
My college room-mate built a Sinclair 60 kit system, in a cardboard box as it goes with no regard for safety! It was awful, but fun to build... cheap too!
You wouldn't see adverts like this these days:
He's thinking: "She smokes, she drinks, she likes listening to my Rock records in Quadraphonic! If she likes football too, she will be perfect!" 😆
Nor this level of sexism:
Good review of the AR6. The AR4a was a goodie too. Obviously done by a proper engineer, not some hack taking a bribe for a good review on a HiFi magazine. I knew these people. 😎
LOL I'm not so sure knowing about the glorious days of the 60's and most of 70's in the UK is a good idea really but we can adjust.
I built a Sinclair Scientific calculator kit. Interest in optics. Took me a while to realise the trig functions were out sadly still needed 6fig log tables at least.
Causing me to have a rethink. If I add a subwoofer W170S become feasible, even sealed. 20L goes more than low enough. Lounge amp is class D though which to me means no extreme impedance excursions. Using a 4ohm woofer may help with that.
I have immense respect for my good friend, @Lojzek:
https://www.diyaudio.com/community/...sealed-simple-loudspeaker-build-plans.352046/
I didn't want to detract from his obvious disordering of Capacitors. 3.3uf and 4.7uf made far more sense to me on impedance grounds.
https://www.visaton.de/en/products/drivers/woofers/ws-17-e-4-ohm
But an excellent effort, which I recommend as an MTTM wired in series. Cheap too. 🙂
https://www.diyaudio.com/community/...sealed-simple-loudspeaker-build-plans.352046/
I didn't want to detract from his obvious disordering of Capacitors. 3.3uf and 4.7uf made far more sense to me on impedance grounds.
https://www.visaton.de/en/products/drivers/woofers/ws-17-e-4-ohm
But an excellent effort, which I recommend as an MTTM wired in series. Cheap too. 🙂
Last edited:
Another interesting link.
At the moment I am pondering my problems. Excellent phase tracking from some hundred to near 5k
Impedance peak ??????? As it currently stands
The dotted line shows the effect of running the optimiser to favour >2ohm. It was a fair bit higher following just changing 8ohm W170S to 4, 😉 surprised me but the crossover was deigned for the 8.
The rise above 10k annoyed me and as the hump is down to both chassis, maybe tweeter worse I added a roll off cap to the resistor on the input to the tweeter crossover. That reduced the hump to as shown. Tried rc across the woofer and no real change. I have no idea what my 😉embarrassing Pioneer AV receiver will make of that. 4 to 8ohm according to them.
The boxsim method appears to be to move the crossover roll over points out so that the actual cross over point is lower than the usual crossovers have. Adding the tweeter roll off has done a bit of that as well as reduced the impedance at the hf end and the hump.
Xmax limits spl to 90db at 50Hz. Oddity with Boxsim it always sets a value lower than the spec sheets say even on a small woofer where they state less than 10% distortion limits. However I expect a woofer to contribute in that area so suck it and see at the moment. Also the above is a sealed box. Vented may give more control of the low end cut off. Sealed it's 20L. Seems I could reduce it but have no idea what Q changes would do Visaton suggest 20L Q .707. A number of designs use it suggesting the xmax limit is similar.
At the moment I am pondering my problems. Excellent phase tracking from some hundred to near 5k
Impedance peak ??????? As it currently stands
The dotted line shows the effect of running the optimiser to favour >2ohm. It was a fair bit higher following just changing 8ohm W170S to 4, 😉 surprised me but the crossover was deigned for the 8.
The rise above 10k annoyed me and as the hump is down to both chassis, maybe tweeter worse I added a roll off cap to the resistor on the input to the tweeter crossover. That reduced the hump to as shown. Tried rc across the woofer and no real change. I have no idea what my 😉embarrassing Pioneer AV receiver will make of that. 4 to 8ohm according to them.
The boxsim method appears to be to move the crossover roll over points out so that the actual cross over point is lower than the usual crossovers have. Adding the tweeter roll off has done a bit of that as well as reduced the impedance at the hf end and the hump.
Xmax limits spl to 90db at 50Hz. Oddity with Boxsim it always sets a value lower than the spec sheets say even on a small woofer where they state less than 10% distortion limits. However I expect a woofer to contribute in that area so suck it and see at the moment. Also the above is a sealed box. Vented may give more control of the low end cut off. Sealed it's 20L. Seems I could reduce it but have no idea what Q changes would do Visaton suggest 20L Q .707. A number of designs use it suggesting the xmax limit is similar.