This has been a consistent issue I've faced. It appears as if everyone I watch building speakers can make boxes near recommended and fairly flat, but whenever I do it the excursion is through the roof. Here are some examples:
1. Driver at 0.2ft3 to not go into over excursion can only handle 4w of power.
2. Parts express recommends 0.2ft3 for sealed enclosure and 0.3ft3 for vented. If I use the same 4w for vented, it has pretty bad over excursion.
3. These drivers should be able to handle 30w, but no enclosure seems like it will work because if you get big for any type of response, even as recommended, it will get no where close. There is no amplifier in the world small enough, so therefore I must be doing something wrong here.
Can anybody assist? Let me know where I'm going wrong?
https://www.parts-express.com/pedocs/specs/294-2760--prv-audio-4mr60-4-specifications.pdf
1. Driver at 0.2ft3 to not go into over excursion can only handle 4w of power.
2. Parts express recommends 0.2ft3 for sealed enclosure and 0.3ft3 for vented. If I use the same 4w for vented, it has pretty bad over excursion.
3. These drivers should be able to handle 30w, but no enclosure seems like it will work because if you get big for any type of response, even as recommended, it will get no where close. There is no amplifier in the world small enough, so therefore I must be doing something wrong here.
Can anybody assist? Let me know where I'm going wrong?
https://www.parts-express.com/pedocs/specs/294-2760--prv-audio-4mr60-4-specifications.pdf
Well for one thing it is a midrange speaker.
The resonant frequency is 89 Hz so x2 is 178 Hz
As with any speaker looking at the impedance curve that is where the useable bandwidth starts.
It will never make bass, so venting around 89 Hz rather 10 Hz above or below would be obnoxious.
Usable bandwidth starts at 178 Hz. And assuming a typical woofer being crossed around 250 to 300 Hz
It is fine for such application. = Midrange
The good news is xmax is 2.8 which is a mathematical equation based on the gap height.
According to the datasheet.
Typical midrange might be only 1 to 2mm.
Looking at Qts is .68 which indicates a somewhat weak magnet, but is somewhat typical for many midrange drivers.
If they actually calculated 10% distortion at 70% BL then xmax is likely more like 1.8 to 2mm
A gap height of 3.5mm the speaker coil can only move so far forward or so far rearward the magnet starts losing
its ability. So distortion will rise. So typically this is where " xmax" is measured 70% magnet BL.
Anyways it is a midrange and venting and trying to get bass is waste of time in vented.
For cone control you want sealed and then with a highpass filter, it will be possible to exceed the 30 watt thermal rating.
30 watts is a thermal rating over a timeline. Basically any speaker aside from a few exceptions.
When they hit Xmax or 10% distortion when the magnet has little control.
Is far far far under the thermal rating.
If you look at the datasheet frequency response at ( 1 watt) you will see in the red circle the top end
already has a obnoxious amount of cone breakup.
So the actual useable bandwidth on a crossover is roughly 2x the Fs up till the cone breakup which starts rather early.
2K might be ok, this is typical for some tweeters. Very large tweeter to cross that low.
Realistic below the breakup is more like 1400 to 1800 Hz.
The whole point in using a small 4" midrange would be good off axis way up at 3k to crossover a small tweeter.
But to much breakup to make that feasible.
PRV does make a few decent drivers, this one, not so much. least they are honest with datasheets. Why your using a 4 ohm driver to start with not sure.
It is horrible impedance. You dont get " more power" with 4 ohm you use more power. The impedance is lower.
I would look at drivers with 8 ohms impedance. less cone break up and with Qts .5 or lower.
.5 to .6 Qts is rather weak magnet for the suspension type. .4 to .3 Qts indicates a powerful magnet with control.
For " bass" a driver needs minimum Fs of 40 Hz of lower. Otherwise it is midrange or mid bass.
" Subwoofers" are crossed around 80 to 100 Hz so any midbass would need a Fs divided by 2
So 40 to 50 Hz Fs to cross to a sub. Otherwise you are crossing to a woofer, then to a sub because Fs is high.
The resonant frequency is 89 Hz so x2 is 178 Hz
As with any speaker looking at the impedance curve that is where the useable bandwidth starts.
It will never make bass, so venting around 89 Hz rather 10 Hz above or below would be obnoxious.
Usable bandwidth starts at 178 Hz. And assuming a typical woofer being crossed around 250 to 300 Hz
It is fine for such application. = Midrange
The good news is xmax is 2.8 which is a mathematical equation based on the gap height.
According to the datasheet.
Typical midrange might be only 1 to 2mm.
Looking at Qts is .68 which indicates a somewhat weak magnet, but is somewhat typical for many midrange drivers.
If they actually calculated 10% distortion at 70% BL then xmax is likely more like 1.8 to 2mm
A gap height of 3.5mm the speaker coil can only move so far forward or so far rearward the magnet starts losing
its ability. So distortion will rise. So typically this is where " xmax" is measured 70% magnet BL.
Anyways it is a midrange and venting and trying to get bass is waste of time in vented.
For cone control you want sealed and then with a highpass filter, it will be possible to exceed the 30 watt thermal rating.
30 watts is a thermal rating over a timeline. Basically any speaker aside from a few exceptions.
When they hit Xmax or 10% distortion when the magnet has little control.
Is far far far under the thermal rating.
If you look at the datasheet frequency response at ( 1 watt) you will see in the red circle the top end
already has a obnoxious amount of cone breakup.
So the actual useable bandwidth on a crossover is roughly 2x the Fs up till the cone breakup which starts rather early.
2K might be ok, this is typical for some tweeters. Very large tweeter to cross that low.
Realistic below the breakup is more like 1400 to 1800 Hz.
The whole point in using a small 4" midrange would be good off axis way up at 3k to crossover a small tweeter.
But to much breakup to make that feasible.
PRV does make a few decent drivers, this one, not so much. least they are honest with datasheets. Why your using a 4 ohm driver to start with not sure.
It is horrible impedance. You dont get " more power" with 4 ohm you use more power. The impedance is lower.
I would look at drivers with 8 ohms impedance. less cone break up and with Qts .5 or lower.
.5 to .6 Qts is rather weak magnet for the suspension type. .4 to .3 Qts indicates a powerful magnet with control.
For " bass" a driver needs minimum Fs of 40 Hz of lower. Otherwise it is midrange or mid bass.
" Subwoofers" are crossed around 80 to 100 Hz so any midbass would need a Fs divided by 2
So 40 to 50 Hz Fs to cross to a sub. Otherwise you are crossing to a woofer, then to a sub because Fs is high.
Looking at sealed midrange ideally we want Qtc of .707
Being the Qts is so high .68 ( weak magnet for suspension type)
To even achieve .707 Qtc it wants 77 liters LOL or 2.7 cubic feet.
Moving up to .8 and .9 Qtc it wants 4 to 9 liters.
4 liters rather high Q of 9 but it is a mid.
As you can see looking at no filter in 4 liters sealed ( yellow) even then 30 watts would drive way past xmax.
Because it is a midrange not designed for bass.
Again 89 Hz Fs x2 puts your bandwidth at 178 Hz and up.
If you apply a standard 2nd order Butterworth highpass filter 180 Hz could be up to 250 to 300 Hz depending on woofer.
Then Xmax is not exceeded at 30 watts thermal limit. It can be driven past the thermal limit to 50 watts.
Which is 10% distortion. Nobody wants to hear that. SPL goals should be reached way under Xmax.
Otherwise High Qts speakers that need very very large sealed enclosures. Is common for car stereo speakers.
Weak magnets work fine in average to large doors. Assuming a Car door is 1.8 to 3 cubic feet.
Then the .707 Qtc is met with such a large leaky enclosure such as a car door.
Either way the speaker would distort rather quickly without a 180 Hz high pass.
89 Hz does not and will never make bass.
Likewise weak magnets around .5 to .6 Qts dont have enough control regardless to make fast accurate bass.
As noted as with any enclosure, Fs will rise. All speakers do this.
89 Hz free air impedance turns into 119 Hz.
This again tells you what the suspension and magnet can really actually do.
Once you get past that impedance peak. This is a message from the speaker it is saying #### you I dont want to do that.
What is saying is anything up from 200 Hz is what I actually feel like doing.
Technically 119 Hz x2 = 238 Hz is my bandwidth start point.
Which is fine for typical generic 250 to 300 Hz crossover used by many 3 ways.
The upper bandwidth is limited as well since the cone breakup even at 1 watt is pretty bad
as seen in the data sheet around 1000 Hz. Basically a good 8" would be crossed that low.
4 " should exceed that by quite a bit. but does not.
Being the Qts is so high .68 ( weak magnet for suspension type)
To even achieve .707 Qtc it wants 77 liters LOL or 2.7 cubic feet.
Moving up to .8 and .9 Qtc it wants 4 to 9 liters.
4 liters rather high Q of 9 but it is a mid.
As you can see looking at no filter in 4 liters sealed ( yellow) even then 30 watts would drive way past xmax.
Because it is a midrange not designed for bass.
Again 89 Hz Fs x2 puts your bandwidth at 178 Hz and up.
If you apply a standard 2nd order Butterworth highpass filter 180 Hz could be up to 250 to 300 Hz depending on woofer.
Then Xmax is not exceeded at 30 watts thermal limit. It can be driven past the thermal limit to 50 watts.
Which is 10% distortion. Nobody wants to hear that. SPL goals should be reached way under Xmax.
Otherwise High Qts speakers that need very very large sealed enclosures. Is common for car stereo speakers.
Weak magnets work fine in average to large doors. Assuming a Car door is 1.8 to 3 cubic feet.
Then the .707 Qtc is met with such a large leaky enclosure such as a car door.
Either way the speaker would distort rather quickly without a 180 Hz high pass.
89 Hz does not and will never make bass.
Likewise weak magnets around .5 to .6 Qts dont have enough control regardless to make fast accurate bass.
As noted as with any enclosure, Fs will rise. All speakers do this.
89 Hz free air impedance turns into 119 Hz.
This again tells you what the suspension and magnet can really actually do.
Once you get past that impedance peak. This is a message from the speaker it is saying #### you I dont want to do that.
What is saying is anything up from 200 Hz is what I actually feel like doing.
Technically 119 Hz x2 = 238 Hz is my bandwidth start point.
Which is fine for typical generic 250 to 300 Hz crossover used by many 3 ways.
The upper bandwidth is limited as well since the cone breakup even at 1 watt is pretty bad
as seen in the data sheet around 1000 Hz. Basically a good 8" would be crossed that low.
4 " should exceed that by quite a bit. but does not.
Last edited:
Thanks for all the information everybody! I'm completely new to this world. I have designed a subwoofer box for the car, but haven't done wood working. I wanted to try my hand at building some boxes for smaller speakers, and this enclosure thing kept being an issue. For some reason I believed you could get below Fs by having things like transmission line style enclosures as they can be tuned lower, but the xmax issue kept popping up.
My real goal is to use BMR with ribbon tweeter and some bass (but might not even need the ribbon tweeter? I'm so new to this realm), again for some reason thought I could push a midrange lower than possible using things like passive radiators or long ports.
My real goal is to use BMR with ribbon tweeter and some bass (but might not even need the ribbon tweeter? I'm so new to this realm), again for some reason thought I could push a midrange lower than possible using things like passive radiators or long ports.
While we are on this topic of crossing over to different speakers. My plan would be to use a KABD board with dsp to mesh all the channels. Give me a better understanding of crossovers.
I can model each driver in Winisd, but how would I test out their combined transfer functions? What software should I use to make 2/3 way systems?
I can model each driver in Winisd, but how would I test out their combined transfer functions? What software should I use to make 2/3 way systems?
And what about this, I just did it to learn. I chose an el cheapo 8in subwoofer. Sealed box 1ft3, doesn't pass xmax. I then throw in a passive radiator, 1ft3, no over excursion. Why doesn't this do absolutely anything for bass extension? How are people getting more bass extension, I'm just missing something clearly. Any bigger box, things over exert. It seems to me so far like no matter what I do, these drivers basically are stuck doing one thing no matter what you chuck at them.
GRS 8SW4 is another high Qts driver.
Qts is .89 which indicates a extremely weak magnet.
These drivers are more suitable for sealed only.
With high Qts drivers, the volume the suspension would like to see to obtain sealed Qtc of .707
would be extremely extremely large.
You are usually stuck with a very High Q alignment which yields what is expected in your graph a 3 dB peak.
That 3 dB peak will be very close to the driver Fs this case 48 Hz in box always higher.
That is the end of the road.
Passive radiators dont make more bass, they behave similar to a port.
Drivers with very very high Qts often wont make any more bass with a vented/ passive alignment.
You put them in a sealed alignment with tolerable Q and that is it.
"Vented" alignments are based on the same old volume the suspension and magnet call for.
A vented alignment just assumes the same old .7 to .8 Qtc and then adds a vent tuned to Fs.
The bass will be slightly more than sealed, trade off is poor transient above and below vent frequency.
And unloading.
There seems to be a magically misunderstanding that some magic port or magic passive radiator will make " more bass"
A speaker is what a speaker is. Bass goes no lower than the driver Fs = the end.
The magnet and suspension want a specific volume for .7 to .8 Qtc = the end.
The vent is tuned slightly above or below the Fs of the driver = the end. no more than 3 to 7 Hz
Whatever bass you get you get, there is no magical vent or passive radiator.
If you want to see a bigger change between vented and sealed alignments.
Simulate a woofer with a strong magnet compared to its suspension type . This would be .3 to .4 Qts
Anything above .5 is starting to indicate a weak magnet, for the weight of the cone and suspension.
The driver Fs basically tells you where the expected response will be.
So if you want a 30 Hz sub then use a driver with 30 Hz Fs or lower. And the magnet/suspension
should be at least .4 Qts anything above .5 indicates a weak magnet.
Cheapo Depot sub that actually hits low is GRS 12SW-4HE
that is a actual 4 layer coil with Vented spider.
Having a very heavy cone and massive magnet to move it the efficiency is expected low 84 dB
Compared to a lighter more normal Woofer with 90 dB efficiency like a Dayton classic DC-300 8
is has Qts of .33 indicating a very powerful magnet for the cone weight and suspension.
It will go in a much smaller box.
Calculate a speakers needed volume for Qtc of ,7 to ,8
The vented bass, will have slightly more bass than sealed.
Trade off is unloading and poor transient near vent tuning.
Qts is .89 which indicates a extremely weak magnet.
These drivers are more suitable for sealed only.
With high Qts drivers, the volume the suspension would like to see to obtain sealed Qtc of .707
would be extremely extremely large.
You are usually stuck with a very High Q alignment which yields what is expected in your graph a 3 dB peak.
That 3 dB peak will be very close to the driver Fs this case 48 Hz in box always higher.
That is the end of the road.
Passive radiators dont make more bass, they behave similar to a port.
Drivers with very very high Qts often wont make any more bass with a vented/ passive alignment.
You put them in a sealed alignment with tolerable Q and that is it.
"Vented" alignments are based on the same old volume the suspension and magnet call for.
A vented alignment just assumes the same old .7 to .8 Qtc and then adds a vent tuned to Fs.
The bass will be slightly more than sealed, trade off is poor transient above and below vent frequency.
And unloading.
There seems to be a magically misunderstanding that some magic port or magic passive radiator will make " more bass"
A speaker is what a speaker is. Bass goes no lower than the driver Fs = the end.
The magnet and suspension want a specific volume for .7 to .8 Qtc = the end.
The vent is tuned slightly above or below the Fs of the driver = the end. no more than 3 to 7 Hz
Whatever bass you get you get, there is no magical vent or passive radiator.
If you want to see a bigger change between vented and sealed alignments.
Simulate a woofer with a strong magnet compared to its suspension type . This would be .3 to .4 Qts
Anything above .5 is starting to indicate a weak magnet, for the weight of the cone and suspension.
The driver Fs basically tells you where the expected response will be.
So if you want a 30 Hz sub then use a driver with 30 Hz Fs or lower. And the magnet/suspension
should be at least .4 Qts anything above .5 indicates a weak magnet.
Cheapo Depot sub that actually hits low is GRS 12SW-4HE
that is a actual 4 layer coil with Vented spider.
Having a very heavy cone and massive magnet to move it the efficiency is expected low 84 dB
Compared to a lighter more normal Woofer with 90 dB efficiency like a Dayton classic DC-300 8
is has Qts of .33 indicating a very powerful magnet for the cone weight and suspension.
It will go in a much smaller box.
Calculate a speakers needed volume for Qtc of ,7 to ,8
The vented bass, will have slightly more bass than sealed.
Trade off is unloading and poor transient near vent tuning.
Last edited:
WinIsd is for basic Alignments or Filter Alignments.
All your doing is entering the speakers mechanical and electrical properties. ( T/S parameters)
And that will tell you the ideal volume the speaker wants.
That is it. The suspension is either so tight or so loose and has whatever magnet to manage that.
If the box is too small, the speaker is boomy and has to work harder since the suspension is either easy or hard to move.
If the box is to big, it is to easy to move so you loose control. Distortion happens faster.
So it was found if the speakers cutoff point and Q is vaguely aligned to common 3rd or 4th order Butterworth or Chebyshev filters
Then the suspension and whatever magnet is has will be happy.
Basically it. people that fart around for hours trying to make magic alignments or convivence themselves that a box that is too small is " ok"
can waste all the time they want.
There is a long set of rules to determine what type of speakers are suitable for sealed and if ported what is ideal alignment.
You can spend years going through that theory and memorize every little detail.
Or as noticed, all these rules are already included in WinIsd.
After you entered the drivers data. When you open a project it will instantly determine if bandwidth product is good for sealed or vented.
Then depending on its parameters will instantly find the correct vented alignment.
Press button, there you go. The box is this big, and the port is 3 to 7 Hz above or below Fs. that simple.
If manufacture does not provide Frd and Zma then you use a tracer to extract the data from the data sheet graph.
Tracer also included in virtuix cad. and no filter calculators needed, it has filter design blocks
To simulate a speakers actual Frequency response in full space on a baffle. You would use Virtuix Cad
You need frequency response data and impedance data for each driver FRD and ZMA
You either measure the driver with microphone on the baffle. Or you use manufacture data which is half space on a AES test baffle.
So you simulate the AES half space data on the baffle and convert to full space. Which will also generate off axis response.
Then you use the crossover designer in Virtuix cad for either passive or active filters. You have to assign X, Y Z data or actual positions of each driver so
phase is correct.
long long long learning process, but at least with a baffle simulator you can see and actually understand how baffle size and driver position
dramatically changes the response, and can toss away peoples magic theories and see what actually happens.
It will make common designs more understood as in why manufactures use typical arrangements, and then also toss into the trash peoples mystical magical nonsense theories.
As noted many people start with Dayton drivers because they provide FRD and ZMA files for all their drivers.
These again are standard 1 watt 1 meter measurements on a AES test baffle, half space
You convert those to full space on your baffle design in the software.
Assign the positions and distances they are actually mounted, then design the crossover.
Good luck you started simulations with some oddities.
High Qts speakers get weird results in sim.
It is up to the designer to use the software to determine a manageable Qtc or box size.
They will always be very high Q and will have weak magnets not ideal for ported.
Very normal average speaker should have Qts around .4 to be friendly to sealed and ported.
.5 about the limit.
Qts around .6 and especially .8 will be hard to sim with " auto alignments"
Dont high tune and if the driver has a real pair of balls it will actually tune below Fs.
Much better for transients. Ported wont add much bass.
Real speakers are " assisted " alignments, meaning you actual get 3 to 6 dB of bass by turning up the " bass" knob.
Depending on the Fs of the speaker will actually determine how meaningful that bass will be.
And how powerful that magnet actually is, connected to a cone of any meaningful strength and suspension.
High Qts speakers get weird results in sim.
It is up to the designer to use the software to determine a manageable Qtc or box size.
They will always be very high Q and will have weak magnets not ideal for ported.
Very normal average speaker should have Qts around .4 to be friendly to sealed and ported.
.5 about the limit.
Qts around .6 and especially .8 will be hard to sim with " auto alignments"
Dont high tune and if the driver has a real pair of balls it will actually tune below Fs.
Much better for transients. Ported wont add much bass.
Real speakers are " assisted " alignments, meaning you actual get 3 to 6 dB of bass by turning up the " bass" knob.
Depending on the Fs of the speaker will actually determine how meaningful that bass will be.
And how powerful that magnet actually is, connected to a cone of any meaningful strength and suspension.
That info on the Qts was really helpful. The GRS 8SW-4HE 8in is actually pretty solid for the price, can run it with the KABD 100w 4 channel amp and dsp to make a little full range boom box practice piece. Has a lower Qts and when put in Winisd with 100w full tilt it can handle the excursion and tune pretty well to 22hz F3 which is pretty darn impressive. Is there going to be quality issues with such a cheap sub? Ya probably, but for a beginner to practice with I think pretty reasonable.
I haven't used Virtuix Cad so I'll check it out.
My idea is a ported or passive radiator setup, I'd like passive radiator for the boom box, but need to practice with wood port building for my car subwoofer build in a few months so will probably go that route.
Subwoofer:
https://www.parts-express.com/GRS-8...-Excursion-Subwoofer-4-Ohm-292-816?quantity=1
Full Range:
https://www.parts-express.com/Tecto...-Full-Range-Speaker-4-Ohm-297-2164?quantity=1
Chucked in a 1.5-1.7ft3 box or so.
Throw subwoofer on one end, have an enclosed area for a couple of the BMR drivers. KABD with battery module, use the dsp to do crossover 150hz or so, eq everything to be pretty flat (will be my first time, so trying to keep it simple).
I haven't used Virtuix Cad so I'll check it out.
My idea is a ported or passive radiator setup, I'd like passive radiator for the boom box, but need to practice with wood port building for my car subwoofer build in a few months so will probably go that route.
Subwoofer:
https://www.parts-express.com/GRS-8...-Excursion-Subwoofer-4-Ohm-292-816?quantity=1
Full Range:
https://www.parts-express.com/Tecto...-Full-Range-Speaker-4-Ohm-297-2164?quantity=1
Chucked in a 1.5-1.7ft3 box or so.
Throw subwoofer on one end, have an enclosed area for a couple of the BMR drivers. KABD with battery module, use the dsp to do crossover 150hz or so, eq everything to be pretty flat (will be my first time, so trying to keep it simple).
Hi,
Let's observe some points.
1) Continuous power on a speaker depends much on which type of signal you apply to take the measurement.
Along the years, many standards were established.
But, in general, a pink noise is the signal chosen - it's a broadband random noise based on white noise (all frequencies) but low pass filtered to get closer to a music power distribution along the frequency spectrum. So you do not submit the speaker to a single frequency at maximum power.
2) For cone excursion, WinISD considers a single frequency sine wave applied with the maximum power set in the Signal Source tab.
It theoretically sweeps the speaker at maximum frequency within the frequency range you specify - I usually set 10Hz to 2kHz for woofer simulation.
This is by far unrealistic in practical use - it's just a tool for the designer.
3) In the case of your example, the manufacture clearly states that full power handling of 30W is only applicable when you high pass the signal to be applied to the speaker. The frequency cut is recommended to be 190Hz using 2nd filter. This is the standard for mid-range, which are not built to handle much power in low frequencies. And it's pink noise, not sine wave.
You can apply low frequencies to mid-ranges, but using much lower power than maximum - for example, if you want a simple and small PC speaker for close listening and very low average power (10W peak).
Conclusion: This speaker doesn't not support 30W in all frequencies and not in a single frequency as WinISD simulates.
It supports pink noise 30W power only when using a high pass filter set to fc= 190Hz and 12dB/oct roll off.
This can be simulated in WinISD by inserting a 190Hz high pass 2nd order filter - if you do so, the XMax is never reached even if setting signal to be 30W.
And this is the normal. For example, a typical 200W subwoofer do not support 200W sinewave @ 10Hz (xmax will limit) or even 200W pure sine wave at 300Hz, where the cone doesn't move much and cooling is not effective (thermal limit) - remember with only 40W we can reach more than 300C in a small workbench welder. Again, 200W is for pink noise broadband signal.
See details in the speaker datasheet.
Is that make sense?