Measuring compression drivers. Questions.

[kelossus]

Hi

This weekend I will be installing a "lightly used" pair of 2441 diaphragms into my 4350 system. Once installed I would like to measure the CD's to ensure the diaphragms and the drivers are good to go.

Am I able to do this with the CD's hooked up to the passive network? I would not connect any of the drivers, just the 2441's. I understand the network may do something to the FR but as long both sweeps are similar can I take that as an indication that all is well? Is REW the right software to use to do the sweeps?

The diaphragms have a few tiny dents and one larger crease that was apparently from the previous install. I am hoping that these imperfections don't cause any anomalies. These diaphragms are hard to get and the drivers are quite difficult to remove from the 4350 cabinet. I only want to do this once.

[Robh3606]

I would run an impedance sweep on them no network and also do a low power sine sweep to listen for any issues. Thy should match up pretty close. Just put a big cap in series to protect against DC. I will try to find the JBL installation instructions and post. It should look something like this. Black curve is a 4" on a waveguide as an example. Figure a couple of peaks on the low end and smooth above.



Rob:)

[kelossus]

Just a follow up. I ended measuring the drivers still within the passive networks. I just wanted to confirm that both the diaphragms measured fairly closely. Sweeps were taken within the drivers operating range within the network, 1k-10k.

I had never opened up my 2440's so I was not aware what diaphragms were installed. I knew they had been cracked open in the 90's. Much to my pleasure I found a mint pair of 375 diaphragms. These speakers had not been used in nearly 20 years so I can assume the 375 diaphragms have had little to no use

One of the 2441 diaphragms does not look good at all. I switched diaphragms between drivers to rule out the driver itself. The bad diaphragm measured consistent no matter what driver it was in. It appears to have a fairly large dip easily visisble in the sweeps.

Measurement distances were not perfectly controlled but I took multiple sweeps and the results always looked the same. All sweeps were taken with the drivers attached to the 2311 horn outside the box sitting on a chair. 1/48 smoothing applied to sweeps.

Any thoughts on the sweeps? Is one of the diaphragms just worn out? They do have a few marks on them but oddly enough the one with more substantial creasing/marks measures the best.

2440 BOTH
84309
2441 LEFT
84312
2441 RIGHT
84310
2441 BOTH
84313

[kelossus]

Also on the sweeps which show both drivers I used the "Separate" Function in REW. There is not a drop in output between the diaphragms.

[Mr. Widget]

Looking at those curves, since the vertical scale is set at 20dB per each delineation, any noticeable difference is actually quite extreme. If the software allows you to run it with 5dB per delineation you will be able to compare them more accurately. Any difference of 1-2 dB between the two diaphragms will be quite audible and will very likely be out of spec.


Widget

[kelossus]

Could the uneven response of the diaphragm be a result of the coil rubbing somewhere?

[Mr. Widget]

Could the uneven response of the diaphragm be a result of the coil rubbing somewhere?Could be. It could also be due to improper centering/alignment, or a partially damaged diaphragm due to wear/age.


Widget

[kelossus]

Could be. It could also be due to improper centering/alignment, or a partially damaged diaphragm due to wear/age.


Widget

Is the response of the "good" 2441 diaphragm what you can expect of this driver functioning properly?

Just checked and I can't change the delineation of the sweeps.

[Earl K]

Is the response of the "good" 2441 diaphragm what you can expect of this driver functioning properly?

Just checked and I can't change the delineation of the sweeps.

The response of your "good" 2441 diaphragm looks to be normal.

Its rising response will flatten out somewhat when you put the 2308 ( slant-plate ) in front of the 2311 horn.

You might want to upload your REW file here so others can look at it for you.

Simply "Zip" it first so that the forum software allows it as a proper file type to be uploaded.

:)

PS; There are a couple of places to change the vertical db scaling ( one is within a small icon that has 2 sets of arrows overlaying a rectangular box ).

PS2; The "bad" response is so off it has me thinking its not a bona-fide JBL diaphragm ( you might want to post a few pics so that others can help with that determination ).

[kelossus]

The response of your "good" 2441 diaphragm looks to be normal.

Its rising response will flatten out somewhat when you put the 2308 ( slant-plate ) in front of the 2311 horn.

You might want to upload your REW file here so others can look at it for you.

Simply "Zip" it first so that the forum software allows it as a proper file type to be uploaded.

:)

PS; There are a couple of places to change the vertical db scaling ( one is within a small icon that has 2 sets of arrows overlaying a rectangular box ).

PS2; The "bad" response is so off it has me thinking its not a bona-fide JBL diaphragm ( you might want to post a few pics so that others can help with that determination ).

Thanks for the reply mate.

I noticed on the 375 diaphragms that came out of the drivers they had shims underneath the plastic ring. Could the lack of shims on the "bad" diaphragm cause this kind of uneven response? Also strange that the 2441 diaphragm that measures well did not require any shims to give a decent sweep.

Diaphragms are both JBL's. I will attach pictures of them below as well as the REW files compressed as you suggested.

2441 REW SWEEPS @ https://ufile.io/w1dvztyg


84315843168431784318

[Earl K]

http://www.audioheritage.org/vbulletin/attachment.php?attachmentid=84317&d=1559565537

This photo of yours ( of 2x 2441 diaphragms ) shows that each diaphragm indeed has 8 shims evenly spaced around the mounting ring.

I suppose ( if you want ) you can add additional ( removable shims ) to see if you can improve the response of your one ( poor measuring ) diaphragm ( its far from a sure-thing that you'll get a response improvement ).



:blink:

PS; Here's a pic of your 2 drivers responses ( with some proper scaling ) > use the yellow-highlighted box to set scaling in REW.

84319

[kelossus]

I suppose ( if you want ) you can add additional ( removable shims ) to see if you can improve the response of your one ( poor measuring ) diaphragm ( its far from a sure-thing that you'll get a response improvement ).



That is quite upsetting to hear. Do you think the poor response is indicative of the diaphragm being worn out or damaged? I was hoping my lack of knowledge installing the diaphragms was to blame.

As I mentioned I did not strictly control the microphone distances when measuring the diaphragms so over lapping the two curves may show a difference in output. The microphone remained static but the drivers were placed on a chair pretty closely in the same position.

Heartbreaking........I managed to buy Speaker Exchanges last 2441 diaphragm which should arrive next week. I don't like the idea of installing just a single new diaphragm. Makes more sense to do it in pairs but if it measures close enough to my "good" diaphragm I don't have a choice.

Otherwise I buy a pair of Radian diaphragms and sit on the brand new 2441 diaphragm until I can find another. This is turning into a very costly lesson.

[kelossus]

Also how do you determine if you should add/remove a shim?

[Earl K]

That is quite upsetting to hear. Do you think the poor response is indicative of the diaphragm being worn out or damaged? I was hoping my lack of knowledge installing the diaphragms was to blame.

<<<SNIP>>>

It's impossible to answer your question since ( IME + I'm in Pro-Audio ) a good 25% of JBL's over-the-counter replacement diaphragms had ( in the past before production was shifted to Mexico ) problems with Quality Control.
Now that their QC is apparently worse I have to wonder what the percentage is ( for us out in the booney hinter-land of Canada ).

IMHO, JBL got away with lax QC because very few had the means to test what they had just purchased ( I also think they shipped more of their "dicey" components out-of-country ).

Therefore, I have no idea whether you purchased a worn-out diaphragm or one that just didn't meet proper Quality Control.

IM (personal) E, distortion ( with a sine-wave ) has to rise above something like 3% before it's obvious to my ears that there's a real problem.

The 3rd Harmonic on your bad diaphragm registers as well below 1% ( somewhat implying that there is not a fitment problem > iow; where the VC is rubbing and generating odd numbered harmonics ).

Using a tone sweep ( again, IME ) was easier to pin-point a bad diaphragm ( assuming one has a good one to compare to ).

In all fairness to eBay sellers ( or any other ) , few-to-none measure the response of what they sell before putting it on the market ( most simply take a DCR reading and call it good ) .

I'd recommend ( since you have measuring capability ) that you play around with the diaphragms fitment to see if you can adjust the FR of that diaphragm ( by re-torquing the 8 screws wth various amounts of pressure and changing-up the sequential pattern used for the 8 screws ) .
- Do you know how to torque down a head-gasket ( or even a car-tire for that matter )?

:)

PS; You might find that the components dedicated for install into systems ( ie; speaker boxes or monitors ) may be better matched ( iow; hand-picked & tested for QC by JBL ) .

[Earl K]

Also how do you determine if you should add/remove a shim?

By measuring FR response ( before & after ).

One can make and use paper shims .

:)

[Robh3606]

I have to ask you have a good pair or Aluminum 2440 diaphragms so why are you changing out to 2441's? You have 2405's up top no? The biggest difference between the 2440 vs 2441 is the increased HF extension. I don't see what you are gaining by the switch.

Rob:)

[kelossus]

I'd recommend ( since you have measuring capability ) that you play around with the diaphragms fitment to see if you can adjust the FR of that diaphragm ( by re-torquing the 8 screws wth various amounts of pressure and changing-up the sequential pattern used for the 8 screws ) .
- Do you know how to torque down a head-gasket ( or even a car-tire for that matter )?


Thanks for the time you are taking to reply. Its appreciated.

I will have a go at shimming and apply varying amounts of torque to the screws. Strange that changing the sequential pattern for the screws would make a difference. I did notice some of the screws are magnetized and some are not?

Head Gasket? No. Car Tire? Yes. Odd question, I'm not completely useless just have no experience in changing diaphragms and want to have a go.

[kelossus]

I have to ask you have a good pair or Aluminum 2440 diaphragms so why are you changing out to 2441's? You have 2405's up top no? The biggest difference between the 2440 vs 2441 is the increased HF extension. I don't see what you are gaining by the switch.

Rob:)

I should have mentioned. I have built Giskards charge coupled networks for my 4350's. The 2440 doesn't reach the now required 10k.

[Robh3606]

OK makes sense now!

Rob:)

[Mr. Widget]

Thanks for the time you are taking to reply. Its appreciated.

I will have a go at shimming and apply varying amounts of torque to the screws. Strange that changing the sequential pattern for the screws would make a difference. I did notice some of the screws are magnetized and some are not?

Head Gasket? No. Car Tire? Yes. Odd question, I'm not completely useless just have no experience in changing diaphragms and want to have a go.Having done a few myself and taking factory tours at JBL (pre departure from Northridge) and at Meyer Sound, it is obvious that properly installing a diaphragm is not trivial. I doubt 80-90% of the drivers that have been serviced by someone not at the factory ever gets the originally specified performance.

If your diaphragms are sound they should be MUCH closer that yours currently are. That said, even with a pair of perfect diaphragms, without shimming and accurately measuring and comparing with a known target there is no way to get theses precision devices set up correctly.


Widget

[kelossus]

Having done a few myself and taking factory tours at JBL (pre departure from Northridge) and at Meyer Sound, it is obvious that properly installing a diaphragm is not trivial. I doubt 80-90% of the drivers that have been serviced by someone not at the factory ever gets the originally specified performance.

If your diaphragms are sound they should be MUCH closer that yours currently are. That said, even with a pair of perfect diaphragms, without shimming and accurately measuring and comparing with a known target there is no way to get theses precision devices set up correctly.


Widget

I am getting the impression that most believe it is the diaphragm that is faulty more then the installation. I am happy enough to spend time playing with shims with the hopes that proper alignment inside the gap may fix things.

Unsure if a pair of used 2440 diaphragms would be worth anything but considering an attempt at a sale to fund a pair of radians. I should have just bought the Radian's in the first place and waited for a NOS pair of 2441 diaphragms to surface.

[kelossus]

Had a bit of a play and this is where I'm at. I did try to keep the microphone to driver distance the same but they may have been out by a few cm here and there. I think this result is good enough? Would you be happy to call it a day with these sweeps?

Strangely enough I could not get a consistent distortion curve. For whatever reason these last sweeps I took today look worse then my initial sweeps. Even on the driver that hasn't been touched.

84323

[Earl K]

Had a bit of a play and this is where I'm at. I did try to keep the microphone to driver distance the same but they may have been out by a few cm here and there. I think this result is good enough? Would you be happy to call it a day with these sweeps?

Strangely enough I could not get a consistent distortion curve. For whatever reason these last sweeps I took today look worse then my initial sweeps. Even on the driver that hasn't been touched.

84323


I'd recommend that you now stop fiddlin with diaphragm fitment ( those 2 are as matched as anyone else might achieve ). So, Congratulations!

Distortion specs won't be reliable until measurements are made above a certain level of background room noise ( iow; the "room noise" makes the results unreliable when the drivers distortion is below that level ).

:)

PS; typically one displays the frequencies in a logarithmic scale ( you've chosen linear ).

[1audiohack]

I would also say you're done.

What made it happen for you?

Barry.

[kelossus]

I shimmed the diaphragm with some sticky tape. Also there was a bit of debris in phase plug slit.

[Mr. Widget]

Congratulations, your results look very good!


Widget

[kelossus]

Nearly ready to go.

84342

[kelossus]

Up and running. Everything works as it should. Will do some system measurements tomorrow and dial them in via the l-pads.

I do believe in capacitor burn-in and they sounded a little closed in for the first hour. Starting to come to life now. Haven't put the woofers in yet in case I need to change something.


84343

[kelossus]

Going through old posts I can see a few members mention balancing the HF and LF sections by measuring with an SPL meter.

Can someone please detail what the process is?

[kelossus]

Bloody hell I have no idea whats going on.

Noticed one of the diaphragms buzzing this afternoon. Measured the driver and its back to shit again? WTF!!!!

Can anyone advise what exactly the tolerance markings on the top plate mean? My compression driver top plate has +2 +2 +2 +1 and my Diaphragm is a -2. My diaphragm comfortably clears the phase plug and further shimming completely ruins the already shit measurements. I feel like the coil might be rubbing somewhere in the gap. Is there a way to sweep these drivers without having to put the back cap on?

[Robh3606]

Is there a way to sweep these drivers without having to put the back cap on?

You are supposed to do a sweep open as part of the install before the cover is on. There is no reason you can't screw on test leads where the regular leads would go. The frequency sweep range and voltage on the PDF Aural Test

Rob:)

A.
PREPARATION
1. Check unit for damage (cracked throat, shifted top plate, etc.) Remove front
cover and old diaphragm assembly. On ring radiators, remove inner cone and
outer horns and unsolder old diaphragm.
2. Using gap gauge, check gap for size and alignment.
3. On 2-inch throat drivers, tap on top phasing plug. The unit should make a solid
sound. A “hollow” or “clicking” sound indicates a loose phasing plug.
B.
INSTALLATION
1. Fold piece of masking tape lengthwise around cardboard digger with adhesive
side exposed. Insert in gap and rote to clean out debris. Continue cleaning until
tape is clean when withdrawn. Inspect visually with magnifying light.
2. Inspect new diaphragm for damage. Line up holes in diaphragm to mounting
pins on the driver. Carefully lower assembly into place. Replace diaphragm
mounting screws. On ring radiators, solder leads to terminals.
3. On drivers, hook unit up to oscillator set at correct input voltage indicated on the
appropriate technical manual and sweep according to technical manual. Listen for
buzzing. Use small mallet to tap lightly on side of diaphragm frame until it is centered
in gap and buzzing ceases. Tighten down diaphragm mounting screws.
4. Reconnect terminal leads onto diaphragm assembly and replace cover. On ring
radiators, replace inner and outer horn.
5. Perform final sound test.

[Earl K]

Bloody hell I have no idea whats going on.

Noticed one of the diaphragms buzzing this afternoon. Measured the driver and its back to shit again? WTF!!!!

Can anyone advise what exactly the tolerance markings on the top plate mean? My compression driver top plate has +2 +2 +2 +1 and my Diaphragm is a -2. My diaphragm comfortably clears the phase plug and further shimming completely ruins the already shit measurements. I feel like the coil might be rubbing somewhere in the gap. Is there a way to sweep these drivers without having to put the back cap on?

It's been mentioned ( & it makes sense to me ) that those numbers indicate just how much ( likely in microns or some very small unit ) that quadrant of the phase plug is out of spec. to the exact design measurement.

Typically ( from what I've seen ), most of these numbers ( on the alnico 4" drivers ) are in the positive zone.

Consider the numbers as a guide suggesting where ( in the quadrant ) you might put an extra shim.

Regarding your latest distortion.

Since you found metal within a slot of a phase plug that strongly suggests to me that the VC gaps may also still need a better cleaning.

These metallic bits will migrate about ( which somewhat helps explain the sudden onset of your drivers buzzy symptoms ).

:)

[kelossus]

Thank you very much for the responses guys. I attacked it this morning after with a fresh mind and had a slight improvement.

I cleaned the gap again, shuffled the shims and I found only one quadrant needs to be shimmed. Once shimmed I installed 3 of the mounting screws and ran a 500hz-1200hz sweep. Edgewound on an earlier thread specified 3.5v for the 16ohm phragms but I didn't take it over 2.5v, 3.5v would have been insanely loud. Couldn't hear any fouling so I assembled it all back together. I am now an expert at removing/installing the 2440 driver from the 4350 cabinet by myself though I hope to never do it again.


These sweeps aren't as good as my initial ones but the 6% THD from yesterdays sweep is now gone. If I can get another brand new 2441 Diaphragm I will send these interstate to an expert and get them fitted and recharged. Sadly in Australia there a few who can service these JBL drivers and even fewer that would back their installation up with measurements.

BLUE TRACE = LEFT 2441, the one I had had issues with. RED TRACE = RIGHT 2441

WITH 2308 ATTACHED, 1 foot away from speaker

84442

WITHOUT 2308, 1 foot away from speaker
84443

[sebackman]

Hi there,

Sorry to be a party pooper but how do you measure and with what gear. Unless you can make certain your gear is calibrated and you are measuring both drivers with exactely the same position there is no knowing if the difference is real or just the result of poor placement. And pretty sweeps pointless if you are not using a echo free chamber or a MLS pulse. Sweeps are usefull in controlled environments but less so outside.

Are your filters identical?

Are you measuring on horn or just the naked drivers?

kind regards
//Rob

[kelossus]

Hi there,

Sorry to be a party pooper but how do you measure and with what gear. Unless you can make certain your gear is calibrated and you are measuring both drivers with exactely the same position there is no knowing if the difference is real or just the result of poor placement. And pretty sweeps pointless if you are not using a echo free chamber or a MLS pulse. Sweeps are usefull in controlled environments but less so outside.

Are your filters identical?

Are you measuring on horn or just the naked drivers?

kind regards
//Rob

I use a MiniDsp UMIK and REW. The mic comes with a calibration file specific to it. I think the consistency of my sweeps is an indicator they are somewhat useful.

Filters are identical.

Always measured on the 2311 horn

[Mr. Widget]

I use a MiniDsp UMIK and REW. The mic comes with a calibration file specific to it. I think the consistency of my sweeps is an indicator they are somewhat useful.

Filters are identical.

Always measured on the 2311 hornHave you seen this document? It covers measuring diaphragms and has a lot of interesting info.


Widget
.

[ivica]

Thank you very much for the responses guys. I attacked it this morning after with a fresh mind and had a slight improvement.

I cleaned the gap again, shuffled the shims and I found only one quadrant needs to be shimmed. Once shimmed I installed 3 of the mounting screws and ran a 500hz-1200hz sweep. Edgewound on an earlier thread specified 3.5v for the 16ohm phragms but I didn't take it over 2.5v, 3.5v would have been insanely loud. Couldn't hear any fouling so I assembled it all back together. I am now an expert at removing/installing the 2440 driver from the 4350 cabinet by myself though I hope to never do it again.


These sweeps aren't as good as my initial ones but the 6% THD from yesterdays sweep is now gone. If I can get another brand new 2441 Diaphragm I will send these interstate to an expert and get them fitted and recharged. Sadly in Australia there a few who can service these JBL drivers and even fewer that would back their installation up with measurements.

BLUE TRACE = LEFT 2441, the one I had had issues with. RED TRACE = RIGHT 2441

WITH 2308 ATTACHED, 1 foot away from speaker

WITHOUT 2308, 1 foot away from speaker


Hi kelossus,

I would suggest to do a measurements at least 1m (3+ feet) from the horn mouth, may be 2m would be better. A baffle about 50cm x 50cm has to be applied too, if more realistic response is of interest.

regards
ivica

[kelossus]

Hi kelossus,

I would suggest to do a measurements at least 1m (3+ feet) from the horn mouth, may be 2m would be better. A baffle about 50cm x 50cm has to be applied too, if more realistic response is of interest.

regards
ivica

My main goal was to sweep the horns in identical settings to confirm that they were both measuring close to one another. I understand this may not be the best way to get a realistic FR sweep but being in identical settings the sweeps should be close if all else is equal.

[ivica]

My main goal was to sweep the horns in identical settings to confirm that they were both measuring close to one another. I understand this may not be the best way to get a realistic FR sweep but being in identical settings the sweeps should be close if all else is equal.

Hi kelossus,

I think putting mic to near the horn a small differences in the distance, or mic angle can introduce larger error, so I would suggest You to put mic 1m apart from the horn (or lenses) mouth and the to do the measurements.....
regards
ivica