1. It is normally recommended that the Sd of a passive radiator be at least 1.5x the Sd of the active speaker(s), and preferably 2x. You indicated that you were using a pair of FR58EX, which I took to mean a pair in each box. Maybe I misunderstood. For a pair, 17.35*2*1.5=52.05, which is the minimum recommendation. 50 cm sq. is close enough, 70 would be better. If you are only using one driver in each box then you can halve that. But it doesn't hurt to go larger.
2. You essentially design a PR system like a ported system, but the PR replaces the port. The calculations for PR tuning replace the calculations for port tuning. If you end up with reasonable numbers then the design should work. If you end up needing an unrealistically light or massively heavy cone then you have a problem and need to change something. Also remember that a PR design will rolloff quicker than a ported system, so you won't get quite the same low requency response.
My recommendation of the Peerless 830855 might work or it might not. While the moving mass appears to be correct the Fs is rather high, and adding mass to the cone to change that messes up the tuning, which has almost no room for adjustment. Going with a larger PR gives you some adjustment room. You must add mass to get the tuning right, the exact amount should be determined by experimentation and testing, but I think this will have a better chance of success. However, you want a very small box and that means limited room to fit a larger PR to it. You can use multiple PRs, but then you have to tune multiple PRs too.
See the attached freq. resp. sample.
While I don't really know the physics behind it or why this occurs in PR systems I understand that lowering Fs of the passive radiator lowers the notch frequency.
Attachments
Hmmm... not sure.
Regardless the notch is irrelevant for the most part since it is below any frequency of interest and down too many dB to be heard.
Are we looking at composite frequency response, driver + PR?
This does not look to my eyes quite what I expect to see... since I can't tell from this if this is an alleged butterworth type alignment - it's impossible to tell where the PR is tuned.
In practice at some frequency below the PRs tuning the box will begin to behave more like a sealed box than an open vent/port. This may be what is being shown.
Regardless, it's unclear to me why this is important or of interest.
Otoh, it is possible to tune a PR "off" the frequency that the driver + box volume "wants" to tune to. That being the frequency where the port is usually added to extend the response flat - or approximately flat. The PR can be tuned below that frequency, by a considerable amount, although the response will drop in level until the PR comes in. In which case when plotted, you would see a dip like the one shown followed by a bump up and back down (reverse parabola) for frequency response.
While not a flat response, in some cases this is a useful way to make a speaker - albeit not a flat one, but when the choice is not much bottom end and some bottom end, it might be an ok way to go. I've done it, and it rocks when done cleverly. 😀
_-_-
Regardless the notch is irrelevant for the most part since it is below any frequency of interest and down too many dB to be heard.
Are we looking at composite frequency response, driver + PR?
This does not look to my eyes quite what I expect to see... since I can't tell from this if this is an alleged butterworth type alignment - it's impossible to tell where the PR is tuned.
In practice at some frequency below the PRs tuning the box will begin to behave more like a sealed box than an open vent/port. This may be what is being shown.
Regardless, it's unclear to me why this is important or of interest.
Otoh, it is possible to tune a PR "off" the frequency that the driver + box volume "wants" to tune to. That being the frequency where the port is usually added to extend the response flat - or approximately flat. The PR can be tuned below that frequency, by a considerable amount, although the response will drop in level until the PR comes in. In which case when plotted, you would see a dip like the one shown followed by a bump up and back down (reverse parabola) for frequency response.
While not a flat response, in some cases this is a useful way to make a speaker - albeit not a flat one, but when the choice is not much bottom end and some bottom end, it might be an ok way to go. I've done it, and it rocks when done cleverly. 😀
_-_-
I did some additional reading and found that the "notch" is where the PR output is 180 degrees out of phase with the woofer. Ported systems avoid this condition because they exhibit less phase shift.
In most PR systems, using a larger box, larger woofer, and larger PR, the resonant frequency of the PR goes very low, so low that it can be ignored. In most cases you are starting with a PR which has a free air resonance on 20 Hz or so, and are adding a considerable amount of mass to get the box tuning correct. Final Fs may be 10 Hz or lower.
In this particular application we are beginning with a small box, higher tuning frequency, a lighter cone. The Fs of our PR is going to start at 50Hz or above. It is a different animal. My concern is if the Fs is too close to the tuning frequency, which is a real possibility here. If the 830855 is used, with a Fs of 63 Hz and box tuning of 95 Hz, then we won't gain as much from adding the PR. Using a 5-incher as a PR we can begin with a lower Fs, perhaps 50 to 55 Hz and still need to add mass, lowering the Fs even more while reaching the target tuning of 95 Hz. Ideally we want to drive the notch as far down in frequency as we can, which is best achieved with a larger PR with a lower Fs to begin with.
In most PR systems, using a larger box, larger woofer, and larger PR, the resonant frequency of the PR goes very low, so low that it can be ignored. In most cases you are starting with a PR which has a free air resonance on 20 Hz or so, and are adding a considerable amount of mass to get the box tuning correct. Final Fs may be 10 Hz or lower.
In this particular application we are beginning with a small box, higher tuning frequency, a lighter cone. The Fs of our PR is going to start at 50Hz or above. It is a different animal. My concern is if the Fs is too close to the tuning frequency, which is a real possibility here. If the 830855 is used, with a Fs of 63 Hz and box tuning of 95 Hz, then we won't gain as much from adding the PR. Using a 5-incher as a PR we can begin with a lower Fs, perhaps 50 to 55 Hz and still need to add mass, lowering the Fs even more while reaching the target tuning of 95 Hz. Ideally we want to drive the notch as far down in frequency as we can, which is best achieved with a larger PR with a lower Fs to begin with.
Last edited:
I'd not worry about this notch for a few reasons.
First, these are idealized simulations. Chances are your real world box is going to be dominated by QL more than anything else, so the apparent notch will not be so deep nor audible.
Then, there is going to be significant doubling for bass notes in most cases, so 40 will appear as 80, and most bass notes will have a fair level of harmonics up the spectrum as well, the doubling will add in.
Unless the notch falls for example on low "E" which iirc is 44 Hz. (?) it probably is invisible.
The level down there for either PR or ported is so very low that what the system is doing is not going to be noticed much.
Imo, anyone who is tuning PRs down as low as you suggest has probably lost their minds, since the ability of the main driver to control that much mass is very much in question. As is the ability of the surround(s) to withstand the pressures for long.
First, these are idealized simulations. Chances are your real world box is going to be dominated by QL more than anything else, so the apparent notch will not be so deep nor audible.
Then, there is going to be significant doubling for bass notes in most cases, so 40 will appear as 80, and most bass notes will have a fair level of harmonics up the spectrum as well, the doubling will add in.
Unless the notch falls for example on low "E" which iirc is 44 Hz. (?) it probably is invisible.
The level down there for either PR or ported is so very low that what the system is doing is not going to be noticed much.
Imo, anyone who is tuning PRs down as low as you suggest has probably lost their minds, since the ability of the main driver to control that much mass is very much in question. As is the ability of the surround(s) to withstand the pressures for long.
actually, come to think of it, the idea that that notch is the 180degree out between the woofer and the PR is incorrect. At that point the PR has MAX output and the woofer minimum. So, if the simulation shows the main driver only, then it is correct, and the deeper the notch the better the transfer to the PR.
_-_-
_-_-
Hi,
A 1" driver with 4mm of Xmax is the same as a 2" with 1mm
and the same as a 4" with an utterly pathetic 0.25mm, none
of them will do any semblance of bass, just distortion.
Forget PR's, here it is pointless, but FWIW if you insist then :
GRS 3FR-4 Full Range 3" Speaker Driver 4 Ohm
Will do the job admirally open circuit - and by adding paralel
resistors to it you can adjust the damping of the alignment.
It will be fine at 3" x 2mm excursion being exactly
double 2 x 1" x 4.5mm, and Fs is the right ballpark.
rgds, sreten.
A 1" driver with 4mm of Xmax is the same as a 2" with 1mm
and the same as a 4" with an utterly pathetic 0.25mm, none
of them will do any semblance of bass, just distortion.
Forget PR's, here it is pointless, but FWIW if you insist then :
GRS 3FR-4 Full Range 3" Speaker Driver 4 Ohm
Will do the job admirally open circuit - and by adding paralel
resistors to it you can adjust the damping of the alignment.
It will be fine at 3" x 2mm excursion being exactly
double 2 x 1" x 4.5mm, and Fs is the right ballpark.
rgds, sreten.
Hi,
(edit : double post, I thought i'd lost the previous one)
FWIW in the context of the original question this is near ideal :
GRS 3FR-4 Full Range 3" Speaker Driver 4 Ohm
2 x 1" = 0.02cuft Vas, and the 3" has 0.03cuft Vas.
The Fs ratios of 220Hz to 135Hz are promising.
I'd guess a 0.04 to 0.05 cuft box is the best.
It also has double* the Vd of the two 1" drivers.
If its not obvious do not connect it to be a PR.
rgds, sreten.
* Well actually its got a lot more as a PR but
it has enough such that you can adjust the
damping by placing resistors across it.
(edit : double post, I thought i'd lost the previous one)
FWIW in the context of the original question this is near ideal :
GRS 3FR-4 Full Range 3" Speaker Driver 4 Ohm
2 x 1" = 0.02cuft Vas, and the 3" has 0.03cuft Vas.
The Fs ratios of 220Hz to 135Hz are promising.
I'd guess a 0.04 to 0.05 cuft box is the best.
It also has double* the Vd of the two 1" drivers.
If its not obvious do not connect it to be a PR.
rgds, sreten.
* Well actually its got a lot more as a PR but
it has enough such that you can adjust the
damping by placing resistors across it.
Last edited:
No.
Minimum excursion of the woofer occurs at the box tuning frequency, where the contribution from the PR is at a maximum. That is not where the notch is located, that is the "knee" in the response curve. Below the box tuning frequency the PR output decreases and the output of the woofer increases. At some point the output of the two cones (woofer and PR) is 180 degrees out of phase, and that is the notch in the response curve.
So the question which still lingers in my mind is ascertaining if the driver would not exceed its X-max limit.
One thing which I learned about tuning frequencies is that the brain can regenerate the fundamental freq from the harmonics mostly lying above 80Hz. So for a small box without DSP, tuning below 80hz will be an overkill. It will work just fine.
One thing which I learned about tuning frequencies is that the brain can regenerate the fundamental freq from the harmonics mostly lying above 80Hz. So for a small box without DSP, tuning below 80hz will be an overkill. It will work just fine.
In any ported or PR system the woofer will be in danger of exceeding Xmax below the tuning frequency. For example If you tune the port or PR to 80 Hz but attempt to reproduce lower frequencies there is little to prevent high excursion of the woofer, except for the limitations of the woofer's suspension. In a sealed box the air inside the box cannot escape, so it provides additional restoring force which helps limit excursion. On other words sealed systems help "stiffen" the woofer's own suspension at high excursions. Ported and PR systems only exhibit this effect near the tuning frequency, where the port/PR is producing maximum output and in effect applying "back pressure" to the active woofer. PR systems might be slightly better than ported systems in this regard, since the PR is not the same as an open port to the outside air. Still, in a small system like yours any significant signal below perhaps 60 Hz would very quickly drive the woofer to or past Xmax even at low listening levels.
Ignoring the frequencies below the box tuning, the frequency of maximum excursion depends on the system alignment (Butterworth, Quasi-Butterworth, or other). In general it is near 1.5x the tuning frequency, so a box tuned to 80Hz will have maximum excursion somewhere near 120 Hz. Good box design software can provide a more accurate calculation for a specific alignment. Small drivers like the FR58EX will also have limited ability to produce volume in the octave above tuning. In a closed box a pair of FR58EX drivers will reach Xmax at about 100dB SPL at 200 Hz, or 88dB at 100 Hz. In a PR system with 80 Hz tuning I estimate your max SPL will be perhaps 90dB. Tuning to 90 Hz or 95 Hz will raise this slightly, but might be experienced more as a decrease in distortion rather than a noticeable increase in volume. You will trade a few Hz of low frequency extension for less excursion and less distortion above 120 Hz.
For lowest distortion you will need a subwoofer or otherwise roll off the lowest frequencies, which your little speakers cannot reproduce effectively.
Ignoring the frequencies below the box tuning, the frequency of maximum excursion depends on the system alignment (Butterworth, Quasi-Butterworth, or other). In general it is near 1.5x the tuning frequency, so a box tuned to 80Hz will have maximum excursion somewhere near 120 Hz. Good box design software can provide a more accurate calculation for a specific alignment. Small drivers like the FR58EX will also have limited ability to produce volume in the octave above tuning. In a closed box a pair of FR58EX drivers will reach Xmax at about 100dB SPL at 200 Hz, or 88dB at 100 Hz. In a PR system with 80 Hz tuning I estimate your max SPL will be perhaps 90dB. Tuning to 90 Hz or 95 Hz will raise this slightly, but might be experienced more as a decrease in distortion rather than a noticeable increase in volume. You will trade a few Hz of low frequency extension for less excursion and less distortion above 120 Hz.
For lowest distortion you will need a subwoofer or otherwise roll off the lowest frequencies, which your little speakers cannot reproduce effectively.
Hey Bob
I did get the winIsd setup, but after entering parameters for a new driver, it prompts me saying that there is a consistency error by pointing to diff groups of parameters.
I read somewhere that its better to first enter the Q parameters, after that others. But the consistency error still shows, but now only few of them. If I modify anything else, then again the same as before, lots of group.
How do you solve this, when you enter new driver parameter?
Enter as little data as possible into WinISD. Qe, Qm, Fs, Re, Sd, Vas. Let the program fill in the rest. Pe and Xmax are useful if you look at excursion and max SPL.
Bob
Bob
When entering speaker parameters WinISD calculates any other parameter it can, since they are often inter-related. If there is any discrepancy between the published specs (which you entered) and the calculated specs you will get the error message. I just re-enter the specs one at a time and watch to see which other specs WinISD fills in, and don't replace any one of the calculated specs. The numbers should be pretty close to the published specs, but if there is something way off you can assume that one or more of the specs is either a misprint or a bad estimate. And make sure the units are correct for each spec, you can toggle between different measuring units in WinISD and it is easy to try to enter liters as square inches and so forth.
Alright thanks a lot. That worked. 🙂
I got a couple of questions regarding WinISD.
1. I selected 2 drivers, normal configuration and filled the volume as 0.03ft3 for Fountek FR58EX. Does WinISD assume that I would place two fountek drivers as stereo in an enclosure volume of 0.03ft3?
2. Based on what WinISD assumed on what I was looking for, the fountek does not have much low end. I added Linkwitz transform(fp-60Hz) with a high pass filter at 55 Hz, to protect it from over excursion. Now the graph is pretty sweet for me. Will that be the actual response for fountek now, working as stereo in 0.03ft3.
3. What is static gain? Whats the electronics equivalent?
4. Now adding a suitable passive radiator improve the low end further and further protect the fountek from over excursion?
Thanks.
I got a couple of questions regarding WinISD.
1. I selected 2 drivers, normal configuration and filled the volume as 0.03ft3 for Fountek FR58EX. Does WinISD assume that I would place two fountek drivers as stereo in an enclosure volume of 0.03ft3?
2. Based on what WinISD assumed on what I was looking for, the fountek does not have much low end. I added Linkwitz transform(fp-60Hz) with a high pass filter at 55 Hz, to protect it from over excursion. Now the graph is pretty sweet for me. Will that be the actual response for fountek now, working as stereo in 0.03ft3.
3. What is static gain? Whats the electronics equivalent?
4. Now adding a suitable passive radiator improve the low end further and further protect the fountek from over excursion?
Thanks.
Last edited:
- Status
- Not open for further replies.