That would help.
You should buy yourself a copy of Toole's Sound Reproduction. A huge amount of useful info in it.
Toole makes 70 measures in an anechoic chamber to get a FR metric that he has shown correlates with sonics. Directivity index is part of that and you are looking for something smooth & monotonic.
dave
Thanks for the tip. i could do with some more reading matter, i shall be sure to look into it.
Now onto the other simmed plots, which in fairness i should have bothered to post before!
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
note: left polar is horizontal, baffle is on North point; right polar is vertical, and baffle is on the East point
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
The phase graph is the one which i dislike the most, but then that is the sacrifice one makes when using multiple drivers. However, the 2nd order filters are not in antiphase which helps IMHO. The polar plots show pretty much what i would expect, with increased 'beaming' especially in the 12 & 16k plots. Again, this is all simmed so i dont fool myself into thinking this is 100% accurate.
As always, comments both pro and con welcomed
Last edited:
The W170S can be impressive, but requires a 40L box to be at it's best (f3 of 28Hz), so it takes more space than 2xW130S, which only require a 30L box, f3 of 42Hz, but have much greater sensitivity and power handling.
As for the XO on the W130S, I used first order (components, particularly inductors get expensive when you're crossing so low), and 2nd order on the FR.
I did experiment with a notch filter on the peak, but they sound much more lively without the filter in place.
I did notice something annoying about the Visaton published specs - they're all mechanical maximum excursion, from rest to peak. There is no Xmax figure. Still, tuned around 40Hz, there's little chance of running out of excursion with most music...
Looking good so far
Chris
PS - sorry about referring to old posts - you guys have been busy typing.....
As for the XO on the W130S, I used first order (components, particularly inductors get expensive when you're crossing so low), and 2nd order on the FR.
I did experiment with a notch filter on the peak, but they sound much more lively without the filter in place.
I did notice something annoying about the Visaton published specs - they're all mechanical maximum excursion, from rest to peak. There is no Xmax figure. Still, tuned around 40Hz, there's little chance of running out of excursion with most music...
Looking good so far
Chris
PS - sorry about referring to old posts - you guys have been busy typing.....
I'm talking about the roll-coaster off-axis response. The response should be monotonic (ie 1st derivative always negative and as flat as possible. This is more important than flat on-axis response.
dave
i see what you mean. however, i dont think that the saddle between 500 and 3k of about 1.5-2dB on the 30 and 60 degree plots is all that much to be concerned about. If i were using a larger bass driver, or FR driver for that matter, i would think that it would be worse. If it the the roll off of HF at off axis positions you refer to, i gather this would also be worse with a larger driver, and an FR. Granted a rising response on axis would reduce this in an FR, it would still be present, albeit largely covered by a rising response. at angles greater than 60 degrees i dont really think much could be done for this, or any other driver. however i may yet be wrong about that.
Assuming this saddle is as much to do with baffle step diffraction as anything else, i will however, look at messing with the baffle radius and see if i can improve things at all.
Assuming this saddle is as much to do with baffle step diffraction as anything else, i will however, look at messing with the baffle radius and see if i can improve things at all.
rollercoaster?
lol well after a few mins of messing with baffle edges, driver placements etc etc, i have concluded that i have the best overall compromise between on and off axis plots obtainable, with this design. you should see the real 'rollercoaster' with zero baffle radiusing both with on and off axis plots.....to this affect, a 20mm radius seems optimum, which to be honest surprises me alot. i would have though larger one would improve things, but according to the plots on the sims, it does not. the saddle decreases at the expense of creating ripples at 30, 60 and on axis around 3k and up, of double the amplitude of the evil saddle
lol so you could say, im quite satisfied with the outcome of the sims and should it be accurate to the finished design in reality, i will be a very happy man indeed.
lol well after a few mins of messing with baffle edges, driver placements etc etc, i have concluded that i have the best overall compromise between on and off axis plots obtainable, with this design. you should see the real 'rollercoaster' with zero baffle radiusing both with on and off axis plots.....to this affect, a 20mm radius seems optimum, which to be honest surprises me alot. i would have though larger one would improve things, but according to the plots on the sims, it does not. the saddle decreases at the expense of creating ripples at 30, 60 and on axis around 3k and up, of double the amplitude of the evil saddle
lol so you could say, im quite satisfied with the outcome of the sims and should it be accurate to the finished design in reality, i will be a very happy man indeed.
With a FR the off axis is monotonically decreasing. That is what you are looking for. Toole & Olive have shown that the off-axis behaviour is more readily corellated to sonics than off-axis.
dave
yes due to the cone moving from omni operation past the transitional area and into breakup, and the resultant beaming of a single cone. im not sure what relevenace monotonic has here, but the relative smoothness of the decrease in omni operation would largely be a factor of the controlled or otherwise, nature of the HF extension via breakup. In a good case then it would be largely smoother i accept, as long as the cone breakup starts around a similar area in FR as the cone omni-directionality drops off, and hence compensate. however i still have doubt that anything but a 'ideal' FR driver could deliver exponential/loagrithmic(whatever rate would apply, though i doubt linear very much) movement away from omni operation
Hence in FR driver no 'saddle as such should exist, although im sure there would be some small compromise Somewhere in the response. controlled flexture here would be very intolerant of minor variations i would guess. with 2 drivers, one will never achieve this though, although use of a cone tweeter MAY help to improve it a little; so i dont see the point in criticising what is essentially a flaw in multiway topology (FAST NOT included in that definition) as from my experience(which is good but im no pro) the off axis is largely better than most i have seen for that topology.
Hence in FR driver no 'saddle as such should exist, although im sure there would be some small compromise Somewhere in the response. controlled flexture here would be very intolerant of minor variations i would guess. with 2 drivers, one will never achieve this though, although use of a cone tweeter MAY help to improve it a little; so i dont see the point in criticising what is essentially a flaw in multiway topology (FAST NOT included in that definition) as from my experience(which is good but im no pro) the off axis is largely better than most i have seen for that topology.
Last edited:
i cite and FR example, albeit not an ENABLEd one or a fostex
out of curiosity after the logic Dave presents, i thought it prudent to try an FR driver on boxsim to see what the off axis response would look like..just for my own info more than anything.
attached is the off axis plots of the visaton bg13 FR driver. on axis it looks poor with the characteristic rising response, and a peak or two. Not as respectable as some on axis plots of other FR drivers but similar to many.
There are a series of serious nulls in the 30 and 60 plots as well as ALL the others. which tbh is to be expected, as cone breakup is occurring. This would create some strange freq dependant lobing issues and i suspect make the music appear to dance around the room, and effect stereo image stability quite severly.
next is the visaton al130 i am using, again running FR:
Again, nulls appear and the same applies. breakup creating frequency dependant nulls and a shifting stereo image.
In both cases it is not good at all, and the off axis plots of my design, although by no means perfect, are streets ahead of either example shown above:
for the record, i chose an FR driver with a whizzer like SOME of the fostex, and one with a high fs(hoping that is would perform better), and also one of the same, or very similar cone diameter to illustrate my point.
the point being: my design is not perfect, but the off axis lobing is a great deal better than the FR example, or any ive seen, and also i would imagine many if not most (or dare i say it), ALL FR drivers out there.
certainly the small saddle and smoother HF transition into directional behavior is a lesser problem than the extensive lobing in the Full range FR plots
out of curiosity after the logic Dave presents, i thought it prudent to try an FR driver on boxsim to see what the off axis response would look like..just for my own info more than anything.
attached is the off axis plots of the visaton bg13 FR driver. on axis it looks poor with the characteristic rising response, and a peak or two. Not as respectable as some on axis plots of other FR drivers but similar to many.
There are a series of serious nulls in the 30 and 60 plots as well as ALL the others. which tbh is to be expected, as cone breakup is occurring. This would create some strange freq dependant lobing issues and i suspect make the music appear to dance around the room, and effect stereo image stability quite severly.
next is the visaton al130 i am using, again running FR:
Again, nulls appear and the same applies. breakup creating frequency dependant nulls and a shifting stereo image.
In both cases it is not good at all, and the off axis plots of my design, although by no means perfect, are streets ahead of either example shown above:
for the record, i chose an FR driver with a whizzer like SOME of the fostex, and one with a high fs(hoping that is would perform better), and also one of the same, or very similar cone diameter to illustrate my point.
the point being: my design is not perfect, but the off axis lobing is a great deal better than the FR example, or any ive seen, and also i would imagine many if not most (or dare i say it), ALL FR drivers out there.
certainly the small saddle and smoother HF transition into directional behavior is a lesser problem than the extensive lobing in the Full range FR plots
Last edited:
certainly creates interest here, and for the european buyers also a great price to performance ratio as well, and ill happily admit in that respect it beats the AL130 hands down. but for the last gram of juice i went for the AL130. the next project will be with the w170s8, budget but quality system for a close friend. xovers should be a doddle!!
I'm not going to try to cover off what Toole does so well in his 2" thick book... just recommend it as a must read.
There are examples of cones & domes... if one takes a bit of extra effort it is doable,
dave
Dave i wouldnt dream of asking you to, this is a forum for debate, and that is what im doing. Im probably missing something i hadnt thought of, read or learnt yet, i accept that.
i however cannot see the issue of a small amplitude saddle versus large breakup nulls off axis that most, if not all FR driver show.
I essentially agree, IN THEORY.
A single driver is better IN THEORY, yet until i can see a FR driver with a smooth amplitude reduction off axis, (which is what i believe you are saying is desired) and with no nasty nulls, and of course no saddle, i dont believe there is cause to criticise such a small problem, since at present i havent seen anything FR thats better. to do so would just show advocation to single cones over multiway , and unnecessary critique of the latter.
i however cannot see the issue of a small amplitude saddle versus large breakup nulls off axis that most, if not all FR driver show.
I essentially agree, IN THEORY.
A single driver is better IN THEORY, yet until i can see a FR driver with a smooth amplitude reduction off axis, (which is what i believe you are saying is desired) and with no nasty nulls, and of course no saddle, i dont believe there is cause to criticise such a small problem, since at present i havent seen anything FR thats better. to do so would just show advocation to single cones over multiway , and unnecessary critique of the latter.
come visit, I have a room full of them. There are some in the UK to, EggFest looks like it is going to be middle of May... a good place to get a taste.
dave
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.