PDA

View Full Version : Do woofer box size affect frequency response around 500Hz?



Jakob
02-06-2011, 11:16 AM
Hi all!

I'm using a pair of ME150's in what I think is too small boxes. The are approx. 2,5 cu ft each. I use 4X18" as bottoms so I'm not that concerned the ME150's maybe doesn't do as good as they can below let's say 80Hz, but can this small box size be to blame for the hump I see around 500 Hz (the rise begins at 250 Hz, peaks at +5dB at 500 Hz and descends to 0 dB at around 800 Hz)? I've measured both left and right speaker and at different distances, and the hump is always present.

If this just is some kind of rising response in the woofer, is there something one can do with the network to compensate for that and bring that region down a few dB?

I am using Giskard's equivalent N300 network wich I know wasn't made for this particular woofer. However, I really like the sound (except the hump) and would like to continue using this combo if possible.

Any suggestions welcome!

Kind regards,


Jakob

JBL 4645
02-06-2011, 02:38 PM
Have you measured within 1meter from the speaker it might well be the room that’s doing this, to the frequency response.

Jakob
02-07-2011, 12:26 PM
Thank you for your reply!

Yes, I did several measures. At approx. 2 feet, 3 feet and I also measured both left and right together at a distance of 12 feet. The hump was always present, with very small variation...

Mr. Widget
02-07-2011, 12:38 PM
You could be hearing the effects of standing waves due to the particular box dimensions. In general a box being too small will raise the system Q and affect low end output, but the box dimensions themselves as well as the construction of the box can influence higher frequencies.


Widget

grumpy
02-07-2011, 01:39 PM
adequate stuffing?

ratitifb
02-07-2011, 01:39 PM
but the box dimensions themselves as well as the construction of the box can influence higher frequencies.yes and don't forget that wavelenght is arround 0.7m only at 500Hz ;)

grumpy
02-07-2011, 02:21 PM
http://www.audioheritage.org/vbulletin/showthread.php?963-Me150hs&p=7990&viewfull=1#post7990

http://www.audioheritage.org/vbulletin/showthread.php?11347-ME150H

...using the T/S provided in either thread doesn't make for an obvious -modeled-
500Hz hump... and 2.5ft3-sealed could be considered on the largish side for that
driver, depending on your target system Q. I don't -think- I'm blowing smoke... :)

You have measured the hump with only the ME150 playing (and no passive or
electronic crossover affecting the response)?

I suppose if there's a rising response as well, that could be a factor... is there
a plotted response for this driver?

The response here might be a familial clue:
http://www.audioheritage.org/vbulletin/showthread.php?12061-ME120H

grumpy
02-08-2011, 07:45 AM
passively, grouping a 24uF, 4.2mH, and 4ohm (or less) resistor in parallel, then
putting that group in series with the woofer would be a very rough
(educational-use only :)) stab at a "fix" (lossy band-stop filter centered at 500Hz),
but getting good measurements and actively adjusting the line-level
frequency response would also be educational...

49810

Earl K
02-08-2011, 10:24 AM
I really don't think your 500 hz bump is endemic to this models typical response .

Here's a quick study for one of my ME150H(s) in a sealed 2.7 cu ft enclosure ( actually the enclosure is tuned to @ 55 hz , but it's been turned "aperiotic" by stuffing foam into the 2 ducts ) .

This enclosure has trapezoidal side walls ( reducing many standing wave ) along with having High Density 1" fiberglass applied on most surfaces .

http://www.audioheritage.org/photopost/data//500/me150h_2_7_cu_sealed-aperiotic_55_hz_port.png

I must imagine the official JBL curve is smoother ( in their standard 10 cu ft sealed enclosure / then baffle mounted into a wall/floor )

Turquoise curve is of the top woofer ( of an active MTM setup ) / measuring distance @ 1.5 inches from the cone near the surround / with an @ 600 hz, 24 db, LR, active crossover point in place .

Purple curve is just the top woofer working / measuring distance @ 1.5 inches from its' cone near the surround / with an @ 600 hz, 24 db, LR, active crossover point in place .

Red curve is just the top woofer working / measuring distance @ .75 inches from its' dustcap midpoint . / active crossover point now @ 6000 hz.

What's going on above 700 hz ? / I don't know what is real - spec . or not / it seems all real enough to me / and with my steep, lowish crossover point, it's quite a moot point .

:)

Jakob
02-10-2011, 12:01 PM
Thank You all for Your input.
It seems You are all right and there is no peak in the ME150's.
I've moved one of the speakers away from walls and done some more measurements and it seems it isn't just a peak around 500 Hz. The response of the system is "flat" up to 1.7kHz and then drops and is 10dB down at 4kHz. After that it's "flat" again...

In the attached file, the blue graph is the H9800+2451SL and 25405, the yellow is all drivers together.

I guess it all depends on the fact that the network wasn't made for these drivers and horns. What I really would need is to bring up the response above 2kHz. I've tried to compensate with the L-pads but that just messes up the response above 2kHz :biting:.

I guess there is padding added in the network to bring the original mid and tweeterhorns down in dB to match the bassdrivers level. Maybe if I could change this padding?

Grateful for hints!

PS: they sound WAAAAYYYY better than they measure :D

Jakob
02-12-2011, 02:19 AM
Please help me out if I´m thinking crazy here:

The N300 was made for the 2231 woofer wich had a sensitivty of approx. 93 dB. My ME150's has approx 95-96 dB.
When looking at the network layout here:
http://audioheritage.org/vbulletin/showthread.php?166-4333-S300-L300-equivalent-bandbass-circuit&pp=15&highlight=l300%20charge

one can se a 20 ohm resistor in parallel to both the 2405 and the 2420/2307. Could changing this value to a bigger one give an overall higher output from the mid and high frequency drivers?

Thank you for any hints...

yggdrasil
02-13-2011, 01:52 AM
one can se a 20 ohm resistor in parallel to both the 2405 and the 2420/2307. Could changing this value to a bigger one give an overall higher output from the mid and high frequency drivers?
Yes, but it will also change the crossover frequencies.

Jakob
02-14-2011, 10:48 AM
Aha... thats too bad... In what way does this resistor affect the crossover frequencies?

Tried to remove the resistor all together on the 2451 and that resulted in an approx. 2dB higher response...

Ruediger
02-14-2011, 11:11 AM
Aha... thats too bad... In what way does this resistor affect the crossover frequencies?

The network wants to see a constant resistive load. The L-pad behaves like an 8-Ohm resistor, if the speaker connected to it is an 8-Ohm resistive load.

The drivers have an impedance of about 16 Ohm. When a 20-Ohm resistor is connected in parallel, the impedance of that combination is roughly 8 Ohms. -> L-pad happy -> network happy.

When You remove the resistor the network will be underdamped, which results in a lift at the crossover frequency.

Ruediger

lgvenable
02-14-2011, 06:39 PM
Just a recommendation; here's the jbl pro enclosure guide

http://www.jblpro.com/catalog/support/getfile.aspx?docid=219&doctype=3