Just bought a pair of 4FE42s because of their supreme specs and sensitivity, but I was a little disappointed with the impedance measurement:
As you can see, it's full of resonances in the 180-300Hz region, and another at 1.3kHz and 3.5kHz.
Both drivers measured similarly.
Is there anything I could try to do with the driver to reduce these?
As you can see, it's full of resonances in the 180-300Hz region, and another at 1.3kHz and 3.5kHz.
Both drivers measured similarly.
Is there anything I could try to do with the driver to reduce these?
Resonances in the 180-300 Hz are sure sign that you are doing something wrong during measurement (as other members pointed out).
Resonances at 1.3 and 3.5 kHz may be real, because there are some disturbances in the measured frequency response close to those frequencies:
The most reliable method for measuring resonances is "waterfall diagram".
The resonance is most likely not a measurement error, it shows up in the datasheet too, but it seems small because the datasheet has a linear scale whereas your measurement has a logarithmic scale. It would be interesting to see the impulse response and CSD.
Interestingly I did find this report of someone who used LTS-50 to treat the cone:
https://www.audiosciencereview.com/forum/index.php?threads/full-range-or-coaxial.51513/post-1880629
If you go that route please do some before and after measurements!
Interestingly I did find this report of someone who used LTS-50 to treat the cone:
https://www.audiosciencereview.com/forum/index.php?threads/full-range-or-coaxial.51513/post-1880629
If you go that route please do some before and after measurements!
I tried many ways. I measured it both held on a rigid surface, on a soft surface and just held in my hand.
They were a tiny bit different but this one was measured in the hand and was the best one.
I did also recalibrate the DATS with 1k and the leads and I measured a Faitalpro 3FE32 and it looks fine:
Here's the other 4FE42:
I just realized it is a stamped steel basket, so I might try dampen it with some butyl damping mat.
I don't have the facilities to make a wooden baffle currently, but I'll try and 3d print one, or do I need to, for impulse, and FR measurement?
Otherwise I'll have to try and get some LTS-50 and of course try to do some measurements.
I just realized it is a stamped steel basket, so I might try dampen it with some butyl damping mat.
I don't have the facilities to make a wooden baffle currently, but I'll try and 3d print one, or do I need to, for impulse, and FR measurement?
Otherwise I'll have to try and get some LTS-50 and of course try to do some measurements.
Lower/middle frequency resonances which happen when the cone is pistonic can be fixed easily enough with EQ.
REW will generate frequency response, impulse response and spectral decay from a single measurement (it's a bear to get the settings right for spectral decay though).
If you have polystyrene insulating board where you live, you can take a big sheet and cut out a hole and inset with a sharp box knife. A styrofoam food box could also work. The stuff that is a continuous foam is easier to work with than the stuff that crumbles into balls.
If you have polystyrene insulating board where you live, you can take a big sheet and cut out a hole and inset with a sharp box knife. A styrofoam food box could also work. The stuff that is a continuous foam is easier to work with than the stuff that crumbles into balls.
How so?
I haven't measured yet, but I'd assume this would show up as a little more jagged looking on the FR or not at all and would just show up as higher distortion.
Yes, I can EQ the whole area down, but it's just the are where this driver plays just about well enough to be played down that low, at loud levels.
I'll have to try and look for some polystyrene boards. I'd still have to 3d print a front baffle to get rid of those awful ears.
Yeah I'd be happier without them, but I don't see them as as big of a problem than the lower ones.
If you don't see a change in the response, then you don't have a resonance. If you do, you can straighten it. There's nothing more to it when the cone is pistonic.
If you tap the steel basket of the driver in the measuring conditions - what freq sound it makes? It may be one of the resonances you measure. The same for membrane itself. The same for spider if you can reach it.
It is a Pro driver... It has other requirements than hi-fi, like: cheap and loud, and just then quality. Maybe weather and abuse resistance. That 94dB sensitivity from such size and at such reasonable price will have drawbacks.
EDIT: I do not believe in 94dB sensitivity, but even 90-91dB is impressive
It is a Pro driver... It has other requirements than hi-fi, like: cheap and loud, and just then quality. Maybe weather and abuse resistance. That 94dB sensitivity from such size and at such reasonable price will have drawbacks.
EDIT: I do not believe in 94dB sensitivity, but even 90-91dB is impressive
Last edited:
This driver may benefit from some further tutoring to shine in it's full potential. And here some iterative, trial-and-error-DIY could help.
In gereral, I think a driver and it's basket should behave neatly as is, in it's "naked" own, in extremis only suspended at some point of the magnet assembly. I used some Excel drivers this way, suspended by two wires in in free air only. Without any front baffle, and therefore without any panel resonances. In order to damp the basket's resonances, I additionally epoxied a big and thick lead washer onto the back of the front basket structure meant to be fixed on a baffle. This indeed was very effective. For a Saudi ordered speaker, I would have choosen thick gold washers instead which probably would have served even better.
I initially saw this kind of approach in one of S.Linkwitz Orion iterations for the W22 driver, with the W22 mechanically nearly not touching the front panel but instead being fixed in place at it's magnet from the rear.
As for this 4FE32, I would try to epoxy (stiffening) or soft-glue (dampening) a 75x15x3mm aluminium ring at rear side of the stamped basket, but eventually not touching the magnet assembly. Additionally I would try to attach carbon profiles onto the outer sides of the 8 trapezoid distance holders of the basket, increasing their strength concentrically to the driver's center.
Having done so, one could imagine to further filing the borders of these already strenghended distance holders into non-linear, curved shapes, thus making them substantially slimmer in the middle of theirs hight.
In gereral, I think a driver and it's basket should behave neatly as is, in it's "naked" own, in extremis only suspended at some point of the magnet assembly. I used some Excel drivers this way, suspended by two wires in in free air only. Without any front baffle, and therefore without any panel resonances. In order to damp the basket's resonances, I additionally epoxied a big and thick lead washer onto the back of the front basket structure meant to be fixed on a baffle. This indeed was very effective. For a Saudi ordered speaker, I would have choosen thick gold washers instead which probably would have served even better.
I initially saw this kind of approach in one of S.Linkwitz Orion iterations for the W22 driver, with the W22 mechanically nearly not touching the front panel but instead being fixed in place at it's magnet from the rear.
As for this 4FE32, I would try to epoxy (stiffening) or soft-glue (dampening) a 75x15x3mm aluminium ring at rear side of the stamped basket, but eventually not touching the magnet assembly. Additionally I would try to attach carbon profiles onto the outer sides of the 8 trapezoid distance holders of the basket, increasing their strength concentrically to the driver's center.
Having done so, one could imagine to further filing the borders of these already strenghended distance holders into non-linear, curved shapes, thus making them substantially slimmer in the middle of theirs hight.
Last edited:
What lower resonances? I don't see anything below 1.3KHz. The thing at 170Hz is unlikely to be remotely related to the driver. The ripples at 200Hz in the first measurement are likely a result of mains hum.
1.3KHz coincides with the directivity breakpoint, it can possibly be compensated with a notch filter because the directivity is still low. 3KHz is very difficult because in order to boost that frequency you would have to dip the impedance by just as much, and the effect of an impedance dip on EPDR is disproportionate. Furthermore if you boost that dip you may also boost ringing that follows it in the spectral decay. So cone treatment is really the ideal solution here, and the results I posted earlier show that the resonance was resolved, which makes it unlikely to be a basket resonance as cone treatment does not affect the basket at all.
1.3KHz coincides with the directivity breakpoint, it can possibly be compensated with a notch filter because the directivity is still low. 3KHz is very difficult because in order to boost that frequency you would have to dip the impedance by just as much, and the effect of an impedance dip on EPDR is disproportionate. Furthermore if you boost that dip you may also boost ringing that follows it in the spectral decay. So cone treatment is really the ideal solution here, and the results I posted earlier show that the resonance was resolved, which makes it unlikely to be a basket resonance as cone treatment does not affect the basket at all.
Very unlikely that a basket has such a low resonance (200-300 Hz).
Also the several similar peaks seem to indicate a measurement issue/artefact rather than a (single) resonance.
Agree: it can be table, if driver is measured just laying on the table or the baffle itself, or driver rattling because of flimsy fixation
You are right. My suggested mods would not change anything, then.
Agreed, not a single resonance. But disagreed about the measurement artefacts hypothesis: Three independent measuremets posted in this thread suggest that there might a real problem for this driver in the range between 150Hz ... 500Hz. Because all of these three measurements show the same kind of jaggies all over this frequency range.
First, and maybe most relevant because it's from Faital itself - The FR measurement from the official Faital Datasheet. Here, it's stretched vertically by a factor of 4 in orrder to to better show all these tiny jaggies in the curve between 150Hz and 500Hz. Compare this uneven section with the relatively smooth one between 500Hz and 1kHz:
2nd: the well known, the initial impedance measurement of this thread. Graph here is stretched horizontally by a factor of 2 for better visibility ... jaggies between 150Hz and 300Hz, this time:
3rd: Last, not least, also the 2nd Impedance of nandappe show the same kind of jaggies. Graph is stretched horizontally by a factor of 2 for better visibility ... this time jaggies between 130Hz and 220Hz:
So, where do all these jaggies come from? As said, the basket might not be the culprit, agreed. Maybe some more, and better resolved measurements in the range between 100Hz ... 1kHz might help to understand.
Last edited:
I have 3FE25 laying around somewhere.
Can tap on the basket and the cone to see what approx. it may be. But not today and will measure with phone probably, as cannot set up proper measurement mic or impedance measurement ATM. From the manufacturer graph, there are also some wiggles at 1-2kHz and 3+kHz, but not on low freq. Maybe smth related.
Can tap on the basket and the cone to see what approx. it may be. But not today and will measure with phone probably, as cannot set up proper measurement mic or impedance measurement ATM. From the manufacturer graph, there are also some wiggles at 1-2kHz and 3+kHz, but not on low freq. Maybe smth related.