My Impedance mesauremant "FS" and manufacturer's "FS" differs significantly

Hi folks

Just have spend all weekend trying to build a functional impedance-Jig, and "make" it work correct on my computer - Receiver - REW connections.
Was real hard to make it all work, must for all the settings on the computer while REW would have exactly the right settings to, and several times i made modifications om the Jig it self.....Wrongly....Because it was "right" from the beggining.

So after 3 calibrations in REW ( open line, short lineand a reference 5,6 ohms) it all seems correct, and i begain to mesaure my 4 drivers mounted in the diy cabinette.
Both the JBL 2235h and JBL E-145 have about the same shape as the manufacturer's mesaurements, but the ohm-peak is quite different.

JBL on the 2235h its peak is at 35 hz, but my mesaured in cabinette ( 126 liter ported) is peaking at 49 hz.....Is this "as i should be" ?
And the E-145 ohm-peak at 101 hz in my mesaurement mounted in cabinette ( ca 30 liter closed), JBL´s mesaurement says ohm-peak at 50 hz.

All this make my wondering......

Best regards John
 

Attachments

  • Jbl_My own on 2 2235h.png
    Jbl_My own on 2 2235h.png
    199 KB · Views: 82
  • JBL E145-8 impedance.jpg
    JBL E145-8 impedance.jpg
    269.4 KB · Views: 85
  • JBL E-145 hemma.png
    JBL E-145 hemma.png
    25.4 KB · Views: 80
Yes. The manufacturer data will be nominal IB / freespace, so if you're measuring in your own modest-sized sealed boxes then you can expect the system resonance (which is what you're measuring under the above conditions rather than the driver's own) to rise, sometimes quite significantly depending on the Vb / Fs / Vas / Qt ratio and what damping, leakage is present. You may also find some shifts depending on whether you're measuring at or near the same voltage as JBL were.
 
Yes. The manufacturer data will be nominal IB / freespace, so if you're measuring in your own modest-sized sealed boxes then you can expect the system resonance (which is what you're measuring under the above conditions rather than the driver's own) to rise, sometimes quite significantly depending on the Vb / Fs / Vas / Qt ratio and what damping, leakage is present. You may also find some shifts depending on whether you're measuring at or near the same voltage as JBL were.

Thank you for answer Scottmoose, so my impedance mesaurements seems "normal" ?.

Can you read out more things with this mesaurements?
One thing that im thinking about, is if the ca 30 liter cloesed for the E-145 "needs" a little port so it can "breathe easier".

It is "little waek" now between 180-300 hz ( 4-5 dB down in simulations)
EDIT....I forgott to say the E-145 "works" from 150-900 hz.


Regards John
 

Attachments

  • 1 aug.png
    1 aug.png
    195.3 KB · Views: 53
Last edited:
For sealed box speakers, yes.

Eyeballing your current values, assuming the drivers are in good shape & JBL's data is accurate (can't comment on the former, probable on the latter), there's an obvious resonance on the E-145 box at about 480Hz, probably some kind of internal standing wave. Similar story for the 2235; looks to be some kind of mode coming in around 290Hz. So if these are the actual speaker cabinets rather than test boxes, you could probably do to get some mode damping in there in outright terms, although if you can't hear anything objectionable, I'd probably be inclined to leave well alone. I don't have time to go looking for the driver data at the moment (currently on my laptop) so I can't make any suggestions at the moment regarding the alignments -will have another look later unless somebody beats me to it, which hopefully they will. 😉
 
  • Like
Reactions: jawen
To the 1st approximation:

E145 max flat BR: Vb = 20*274.7*0.25^3.3 = 56.64 L, Fb = 0.42*35*0.25^0.96 = 55.63 Hz

Max flat sealed/0.707 Qt: a = (0.707/0.25)^2-1 = 6.998, Vb = 274.7/6.998 = 39.254 L, Fc = (0.707*35)/0.25 = 98.98 Hz

At ~30 L: a = 274.7/30 = 9.157, Qtc = (9.157+1)^0.5*0.25 = 0.797, Fc = (0.797*35)/0.25 = 111.58 Hz, so 101 Hz measured implies either the driver specs are a bit different and/or the box is a bit larger and/or there's a minor leak.

Need vented cab details to check it............
 

Attachments

Can you prevent D3 from rising as it does below 300Hz so you can cross it more evenly?

All my "outside" mesaurement from 13 Maj disappeared when REW suddenly "shuts down" by itself, and all work was gone. ( happend 15 times now)
So i only have 1 quite bad SPL mesaurement left from the E-145 , witch i use in this simulation ( will try to make a "new ones" inside tomorrow, best i can do)

See now that the used mesauerement on the E1345, have big dipp from 150-450 hz ( but in reality mesaurement it doesent)

So now in the simulationsprograms, i can´t do much better than this.

Regards John
 

Attachments

  • E145 no xover maj-23.png
    E145 no xover maj-23.png
    26.2 KB · Views: 31
Ah yes. The simplest thing is to ignore that region, but it would be more productive if you merged a near field measurement with it. You measure the bass close to the cone. You use that part below the baffle step.
 
To the 1st approximation:

E145 max flat BR: Vb = 20*274.7*0.25^3.3 = 56.64 L, Fb = 0.42*35*0.25^0.96 = 55.63 Hz

Max flat sealed/0.707 Qt: a = (0.707/0.25)^2-1 = 6.998, Vb = 274.7/6.998 = 39.254 L, Fc = (0.707*35)/0.25 = 98.98 Hz

At ~30 L: a = 274.7/30 = 9.157, Qtc = (9.157+1)^0.5*0.25 = 0.797, Fc = (0.797*35)/0.25 = 111.58 Hz, so 101 Hz measured implies either the driver specs are a bit different and/or the box is a bit larger and/or there's a minor leak.

Need vented cab details to check it............

Wow GM.....Lots of "for me" advanced calculations!

And this was previously a DIY 3-way speaker, which I rebuilt to a 4 way from the inside (through the 15 inch hole where the 2235h sits at the side)
So E145 boxshape is hard to measure exact liter volume on, my guess was ca 30 liter.

Before the 15 inch in the front baffle has all the available boxvolym, but now i build a separate box for front baffle 15 inch ( and use the E145), and have mount a 2235h on the side of the speaker, so was trying to get as much volym as possible for the 2235h.

Can you please calculate E145 box-liter, based on the 101 hz I measured the impedance at?
Maby its 35-38 liter or something?, and maby it will wok a little better with a small port ?

Regards John
 

Attachments

  • IMG_20230421_130712.jpg
    IMG_20230421_130712.jpg
    360.7 KB · Views: 42
  • IMG_20230424_160926.jpg
    IMG_20230424_160926.jpg
    277.7 KB · Views: 43
  • IMG_20230421_135126.jpg
    IMG_20230421_135126.jpg
    378.9 KB · Views: 43
  • IMG_20230512_154218.jpg
    IMG_20230512_154218.jpg
    226.6 KB · Views: 44
Last edited:
Ah yes. The simplest thing is to ignore that region, but it would be more productive if you merged a near field measurement with it. You measure the bass close to the cone. You use that part below the baffle step.


Hahaha yes, i've been trying to ignore that region, but my brain keeps messing with that dip anyway.
When you wrote "bass", do you still mean the E145 ?

I have mesaurement from the 2235h from 2, 15, 30, 50 and 100 cm allready 👍

Regards John
 
The cabinet nearfield will invalidate the results lower in frequency than the cone nearfield, which is why I used the word baffle step.

Okey Allen and GM, now ivé spend a few hours to SPL-mesaure the JBL E-145/8 ohm again in the room.
Still big dipp between 200-350 hz from 1m, but much goes "away" when mesaured from 0,5 m and from 0,25m its almost "gone".

Had pillows and a few big towels between the speaker and mic for " a little" better mesaurement terms.

Regards John
 

Attachments

  • E145 from 25_50_ 100 cm mic center1.png
    E145 from 25_50_ 100 cm mic center1.png
    59.5 KB · Views: 32
Well, to get it small (how small a pipe & max length) it depends on the XO point/slope to the 2235H.

Thank yu so much for the calculation GM 🙂
In that case maby a pipe max 15 cm long and as thin as possible (without to big compromisses)

This is a Picasso painting from me on the diy speaker 😉
I have no musts in terms of xover solutions, i let the project "allready fixed parameters" control me. ( drivers, volym, shape etc)

My troughts was to have around 150 hz xcross 12dB from the 2235h to the E-145, mostly for avoid directional bass sound from the 2235h.
And also 12dB slope on E-145 ( both lowpass and highpass) and meating the cd (Aga driver) around 850-900 hz.

But nothing is "written in stone"
These mesaurements are without xover on the 2235h and E-145.

Regards John
 

Attachments

  • Jbl projektet.png
    Jbl projektet.png
    28 KB · Views: 38
  • JBL 2235h Red 50, Blue 30 and Darkgreen 5 cm ute 13 maj.png
    JBL 2235h Red 50, Blue 30 and Darkgreen 5 cm ute 13 maj.png
    35.4 KB · Views: 29
  • 2235h Ports blue upper red floor-port.png
    2235h Ports blue upper red floor-port.png
    45.5 KB · Views: 29
  • E145 from 25_50_ 100 cm mic center.png
    E145 from 25_50_ 100 cm mic center.png
    63.2 KB · Views: 32