Did my JBL2445 drives fail or not? If yes, how?

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Member
Joined 2008
Paid Member
I have a problem. The FR of my drivers (both) has changed. I checked everything, even the UMIK, but it really seems something happened to the drivers.

The FR plot in blue shows normal (before, with XO) and the red the current state (without XO).

In the impedance plots, the black one was taken at the same time as the normal FR at the right driver using headphone output and a 100 ohm sensing resistor.

The red trace is the current imp curve taken with the same equipment, the yellow one is the left driver.

The funny thing is, both drivers behave very similar (the FR is same). I measured the right driver a lot and I might have sent too low frequency to it - but definitely less than one watt, it must have been more likely around 100 mW maximum. The other thing is, the drivers are connected to a SE tube amp - directly to the 8 ohm tap with active XO. Maybe the damage comes from there somehow?

I could understand a failure of the right driver - that one was definitely under some LF stress - but at a very low volume. But not an identical failure on both at the same time.

The question is, have you ever seen anything similar? What is the cause? Can it be fixed without replacing the diaphragm? How to prevent this in future? Could the diaphragm crack? The drivers are old, so it could be a failure due to age?
 

Attachments

  • DamagedFR.jpg
    DamagedFR.jpg
    229 KB · Views: 224
  • DamagedImp.jpg
    DamagedImp.jpg
    197.9 KB · Views: 220
You aren't driving it from a headphone output for the FR test are you? That won't work, gotta go to something that can drive a low impedance load and not through a series resistor.

If that's not it (if you were using an amp), what cabinet is the driver in? Looks like a ported, how is the mic placed? The impedance graph indicates a port frequency of 400Hz, that's a no-go for a woofer.
 
Last edited:
you need to provide more information about your setup. Which amp did you use? naked drivers, or loaded on a horn?

not an expert, but at very low input levels it is unlikely that you damaged anything. What was the lowest frequency you ran through them?
 
Member
Joined 2008
Paid Member
These are in horn, nothing has changed between the measurements. it is a tube amp, so definitely no DC. I do not know if there could be a voltage spike due to transformer? They could get 20 Hz, but for a short time during the sweep with low levels. Inbetwen ste measurements before I noticed I listened to it for some 3 hours as background music, low level and with proper XO all the time. The horn is a JBL2386 copy and when I do not make a mistake, I sweep from 400 Hz. I have a Beyma CP385Nd in the system connected to a Behringer SS amp, but that one still works perfectly.

In the total system FR there is a shallow walley between ca 600 Hz and 2 kHz, which got unnoticed due to low level background listening.

I wonder if the change in FR and Z can be explained somehow. To confirm that the change is really in the drivers, I changed everything in the signal chain, including the measurement computer and left right side and measuring another speaker with the same amp. So you can consider the measurements are valid and no mistake.


I wonder why the FR has changed when nothing else was changed. I suspected the UMIK, bud measured curves of Fane 12-250TC are identical as the oldel ones, so the mic is OK.
 
Last edited:
Looking at your first graph red and blue traces while I can not offer an explanation.

It appears the red and blue traces are inverted ie the high blue has become low red and centered around 2.8 kHz the high red has now become low blue in other words the traces have reversed

It’s as if more power is sent above 2.8khz and less power below and your second graph trace

I think both traces are bit off in any case


http://www.jblpro.com/pub/components/2445J.pdf
 
Last edited:
Administrator
Joined 2004
Paid Member
I have 2440 drivers with JBL OEM replacement diaphragms and have found them to be very sensitive to measurement setup. You need to drive them with a proper low impedance voltage source and sample their VC current with a low value of sampling resistor.

I doubt there is anything wrong with the driver it likely has something to do with how you measure them.

Note that the effects of variations in horn loading over frequency range measured will be visible in your measurements.
 
Member
Joined 2008
Paid Member
My midrange is back

All the measurements were made without crossove, driver directly to amo. And yes, no horn compensation, there is a tweeter above.

For some reason, the horns measure correctly again. I do not know, what happened. As I wrote before, I tried two different computers, two amps, measured both horns. The change now is that I used another cable between the computer and amp. I cannot think of any reason how this could work. I will measure again with the old cable today. Since now it works also with the crossovers with them.

The other possibility is it was a UMIK glitch.

I checked one driver - there was one loose membrane screw and the foam inside turned into a sticking glue, so I replaced it temporarily with polyfil. I did measure the horns before doing this and found out later that the problem was gone before I touched them.
 
Member
Joined 2008
Paid Member
I originally thought that the decomposed foam was touching the diaphragm from the back - it definitely was, but it was not the cause of this. If I find out it was the (cheap) signal cable, I still do not understand, how that could suck out the mids only. If anything, I would expect attenuation of HF.

I also discovered a very small dent in the diaphragm, but no light through, so no hole there. One day I will disassemble the other one and remove the foam.

Any tips for damping the back chamber? The polyfil seems to work almost the same as the old foam.
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.