JBL 2245H with crazy TS specs measured with Dayton DATS 3

[Morten Krogh]

Hi
This is my first post in here so please have me excused if I make rookie mistakes. First of all I would have posted more pictures of the two JBL 2245Hs. Is there a way to do that?
My question, that I hope some of you can help me with is: I have just bought two great looking 2245H, as far as I can see they are either unrestored late models or they have been reconed with original recone kits some time in the past. I have tested the with at tonegenerator and they perform perfect at all frequencies and amplitudes.
But when I test them with the Dayton DATS V3 I get some crazy TS data. Fs around 13hz; Qts 0,79 - 0,71; Qms 4,6- 6,3 ; Qes 0,95 - 0,8.
If I can somehow upload more files I could Upload the DATS V3 files and graphs.
The surrounds look and feels fresh.
I have never heard of TS specs being that much off.
Do anyone have ideas to what is going on here?

[JeffW]

looks like a recone with all the glue around the spider.

Mogami cone according to this thread

http://www.audioheritage.org/vbulletin/showthread.php?20292-2245H-recone-done-right

But maybe wrong motor? Voice coil? Who knows.

[Morten Krogh]

looks like a recone with all the glue around the spider.

Mogami cone according to this thread

http://www.audioheritage.org/vbulletin/showthread.php?20292-2245H-recone-done-right

But maybe wrong motor? Voice coil? Who knows.


Thanks JeffW
It do looks like a Mogami cone according to the tread. I can't find any Mogami 2245H kits for sale online, are they out of the market?
The rear of the cone show aquaplas on the entire surface, is there no chance that it could be a original JBL reconekit?
The motor looks right to me, at least the badge!
I think it takes a very soft suspension or a very heavy cone to send the Fres down to aprox. 13 Hz. The suspension feels quite soft/free, the cone does not "feel" or look very heavy. I have never touched a genuine 2245H so I don't know have it should "feel".
Could the DATS V3 in this instance, be fooled to do wrong measurements?
I have not measured the Vas and Mm because it takes more than 600g of weight to lower the Fres more than 25% as required by DATS V3 and at that point the suspension looks like is it about to bottom out!!

[Earl K]

`Some advice; You should always include in your first post some proof that you're experienced in making these sort of tests.

As far as we are left to speculate, you might also simply be a rookie at testing ( and unable to get DATs to deliver any reliable readings > that scenario wouldn't be the first time ).

:)

[Robh3606]

Make sure when you run your tests the vent isn't blocked. That low FS? could be a very tired spider. My 2216Nd have a lower than spec FS. They were subjected too the 100hr power test so well broken in the say the least.

Rob:)

[JeffW]

It do looks like a Mogami cone according to the tread. I can't find any Mogami 2245H kits for sale online, are they out of the market?
The rear of the cone show aquaplas on the entire surface, is there no chance that it could be a original JBL reconekit?


Mogami was the JBL supplier for some time from what I gather (I'm no 2245 expert, but have some and have read a bit about them) so it absolutely can be an original JBL part. JBL used a few different cone suppliers over the lifespan of the 2245.

[Ian Mackenzie]

Thanks JeffW
It do looks like a Mogami cone according to the tread. I can't find any Mogami 2245H kits for sale online, are they out of the market?
The rear of the cone show aquaplas on the entire surface, is there no chance that it could be a original JBL reconekit?
The motor looks right to me, at least the badge!
I think it takes a very soft suspension or a very heavy cone to send the Fres down to aprox. 13 Hz. The suspension feels quite soft/free, the cone does not "feel" or look very heavy. I have never touched a genuine 2245H so I don't know have it should "feel".
Could the DATS V3 in this instance, be fooled to do wrong measurements?
I have not measured the Vas and Mm because it takes more than 600g of weight to lower the Fres more than 25% as required by DATS V3 and at that point the suspension looks like is it about to bottom out!!

I have not looked closely at this test kit.

Is there a calibration process?

I would contact Dayton and explain what your doing with the kit for their feedback.

I noticed recently someone else on the forums got odd results with this kit.

What l suggest you do is conduct a free air resonance test with a 1000 ohm resister with a sine wave generator App and your amp. Google loudspeaker impedance test.

This will tell you straight out the FS of the driver.

[ds23man]

Do not block the vent, fooled me in the past with a 2235...

[RMC]

Had a quick look at DATS 3 Quick Start Guide.

Item 4 is "Calibrate the test leads"

Item 6 mentions "For best results clamp the driver in place and allow adequate clearance for any rear pole piece vents.

Both calibration and rear vent clearance were already mentioned here. So one should make sure this is done. Also, there are already two small software updates for DATS 3 that one should not forget to install.

Putting pieces of wood, such as 2X3, 2X4, 3X3 or 4X4", under the driver in a way its lifted so the back vent can breathe easily may do the trick, or in absence of wood blocks the woofer's back edge where gasket is can be resting on the edge of two sturdy tables having same height and at the required distance. By analogy like "in between two chairs".

Item 8

Measuring Vas from one of the four methods mentioned in DATS is the bugger, even with the simpler added mass method because of the size of it. The following may be a work around to this Vas assessment.

"Once you have measured the speaker’s free air parameters the "Measure V(as)" button becomes enabled." Bypass this Vas button step. My inderstanding (right or wrong) is you already have at this time on DATS screen some free air parameters. Use these without Vas.

In Win ISD, driver editor, parameters tab, there is an "Auto calculate unknowns" feature to calculate missing TS parameters.

Input in Win ISD the known numbers you have from DATS incomplete TS list and ISD should calculate the missing one(s), in the present case Vas. Hopefully this will work to get you a Vas figure to model the driver in a box using Win ISD. Regards,

Richard

[Ian Mackenzie]

Richard is on the right track.

The measured Qes is crazy.

The TL parameters only change relatively when measured with the voice coil warmed up.

I would go through all the points in Richard’s point and run another series of measurements.

I might pull out one of my stock 2245’s during the week and run some measurements. I have a Dayton reseller in the area and might see if he is interested in validating the Dayton measurements.

[Morten Krogh]

Thank you to all of you for the help so far.

I did all the Dayton DATS V3 calibrations when I started using it some time ago and now again just to be sure it was done and OK
.
I have taken care off the vent is breathing freely
I have measured a lot of other drivers, some are measure a bit lower i Fres some a bit higher than the manufactors specs. The only ones that are far away are the two 2245Hs.
I will also so much appreciate if Ian will compare and validate DATS V3 data for one of his 2245H
I will attach pdf. and DATS V3 files with me latest recheck of the measurements. Hope you can open them and maybe import them into your DATS V3 box.

8594785948


I was so happy that I finally found two of them that was good looking and close to original in appearance. I bought them from a widow that has a lot of nice JBL stuff from her husband. But se have no idea what the state and historie is for some the parts, including my mysterious 2245H that can proof to be worthless :-(
She wants to sell it all but dont know how or where and I dont have the time to buy it and sell it on ebay. I would like to buy it all and keep it, but I have to stop buying and start using instead. :-)
Maybe I could post the parts for her in here and help her sell them. The items could be interesting for you guys?

[Earl K]

Morten,

You posted a pdf indicating the measured Re values were/are in the neighborhood of 22 ohms ( post 3 ) .

This should be a MAJOR red flag that either; those 2245H woofers Voice-Coils are toast ( and you've made a bad purchase ) or your measurement system or methodology are faulty.

Normally when presented with a figure so far from official spec. ( someone more experienced in these matters ) would simply go get their MultiMeter ( then set it to measure DC resistances > DCR ) to get a second opinion ( before posting their doubts to social medfia ).

Your failure to make ( or mention making ) a dedicated DCR test using an alternate method ( has me shaking my head ).

:)

[RMC]

To Earl:

With all due respect, seems to me you're pretty tough here on the new member who requested help.

To Morten:

Its always a good idea to bring a multimeter when going to purchase speakers. This way you can check driver DC resistance on the spot, the number measured can tell you something is good or bad.

A cheap digital multimeter can do but you must first determine its residual resistance to have a more accurate reading. Set the meter on DCR mode, then hold the two test leads together for a minute until the number shown stabilizes, take note of that number, its the meter's residual that you must subtract from the speaker's DCR value measured. This way you get a pretty accurate figure. A lot of people forget to do this.

For example, just to illustrate what one may expect, my $30. digital one has a residual of 0.6-0.7 ohm, the $80. one 0.0-0.1 ohm, the $150. unit just zero. Regards,

[HCSGuy]

First, I do not have DATS v3 - I have v2, so I'm afraid I cannot help with anything in that regard. Second, I have not seen a significant change of Fs from blocking the rear vent. Because it acts as a mechanical resistance to resonance, it tends to damp the Fs to a lower, broader number, but it is usually still at the same frequency - instead of getting a nice, sharp peak, you'll see a lower, broader peak with the blocked vent. I am with others who would question the Re number - if in fact the DC resistance is 22ohms, then there is something really wrong here. I have had bad measurements in the past using DATS due to incorrect audio settings in Windows, but those always showed up as errors in the impedance calibration function of DATS - if you can do the impedance calibration with the resistor, and the test leads calibration, and they both show nice, straight lines, then DATS is probably operating correctly. Make sure your output level is reasonable on your computer - if you test with really low volume, like <5, things get weird as well. I have not noticed a significant change in the impedance sweep with volume levels from 20 to 100, so overdriving doesn't seem to be an issue. So, unless you are measuring in outer space, I don't have a good solution for you. Here's a sample 2245 graph from a beater woofer I had sitting on a shelf - Fs on this one is really high - 37hz or so, but I suspect it is the foam - it's about 18yrs old and is pretty hard, but the Re is within reason, so please check yours.85953

[Ian Mackenzie]

Thank you to all of you for the help so far.

I did all the Dayton DATS V3 calibrations when I started using it some time ago and now again just to be sure it was done and OK
.
I have taken care off the vent is breathing freely
I have measured a lot of other drivers, some are measure a bit lower i Fres some a bit higher than the manufactors specs. The only ones that are far away are the two 2245Hs.
I will also so much appreciate if Ian will compare and validate DATS V3 data for one of his 2245H
I will attach pdf. and DATS V3 files with me latest recheck of the measurements. Hope you can open them and maybe import them into your DATS V3 box.

8594785948


I was so happy that I finally found two of them that was good looking and close to original in appearance. I bought them from a widow that has a lot of nice JBL stuff from her husband. But se have no idea what the state and historie is for some the parts, including my mysterious 2245H that can proof to be worthless :-(
She wants to sell it all but dont know how or where and I dont have the time to buy it and sell it on ebay. I would like to buy it all and keep it, but I have to stop buying and start using instead. :-)
Maybe I could post the parts for her in here and help her sell them. The items could be interesting for you guys?



An observation.

Looking at the impedance curve it dips to a low value before resonance. That value is about the DCR value of the voice coil.
The DCR the value in the side table is mysterious. Normally a damaged driver will show a lower dcr value if there is a shorted turn on the voice coil or will be open circuit of it is blown.

A question?
Do both drivers measure the same on your Datsv3?

My recommendation is buy a low cost multi meter (less than US$50) as soon as you can and measure the dcr at the terminals of both drivers.If the value is close to the spec then something is screwed up with the Datsv3 or how you are using it.
If the value is the 22 ohms then you drivers are damaged. The problem is solved.

What I will do.

I will measure my 2245's with my Linear X LMS system which is a professional loudspeaker measurement system in the next couple of days and advise the results.

Edit
I measured one of the drivers. The Fs was about 25 hertz. This driver was re foamed at some point. The cone is an original JBL cone kit. I mounted the driver in a vertical position in a heavy vice.

I have not used the driver for some months and I did no pre conditioning routine.

Referring to the LMS manual they advice impedance curves are very sensitive to the environment and how the driver a supported. The driver needs to be very rigidly clamped on both sides between two tables with G clamps.

When time permits I will run the LMS driver parameter program and generate a full set of driver data with either delta mass or delta compliance methods using a test box.

One thing to point out is that if one parameter is not quite right it means another or more parameters will have also been effected. The result is that these difference can offer each other and you end up with a similar situation as far as tuning goes.




I will see if I can get hold of a Datsv3 to re rest my 2245's.

I think its great that you have posted this information and one way or another it can be validated.
However you need to help yourself and measure the DCR of your drivers with a multimeter.

[grumpy]

Wanna hear my wild ass guess? Someone reused a cone/voicecoil/spider assembly (the latter looking a bit torn up in one of the pics), then refoamed it.
Would explain a worn/sloppy suspension (low Fs). I don't really think the DATS-3 is the problem; the high Re just --appears-- to be function of their algorithm that made some assumptions about how low (and how high a Q) the resonance of a driver would be. The curve shapes and 6 ohm minima between 100-200 Hz isn't crazy ... even cone resonance wiggles look kinda right. I hope I'm wrong about the driver.

[Ian Mackenzie]

That’s a possibility.

I think putting it is a test box would disclose the true parameters.

[Ian Mackenzie]

My DATSv3 arrived and l have started some testing on several Jbl drivers. I am using a Windows Lap top and it installed without a hitch. There are to calibration routines. That is fool proof that it’s working correctly.

With the actual testing with a 2235H sitting on my desk l got an FS very close to the factory spec.

My 2245 all measured with an FS of between 26 and 30 hertz.

The Vas value was about half the factory spec using added mass of 145 grams. I suspect that because the drivers have been re foamed that has impacted on the Fs and the VAS.

l dare say this is the case with most 2245 currently in use where the driver has been re foamed (Edgewound will be shaking his head...told you soooooo).

I will plug all the values into a simulator to compare to the factory spec drivers.

Intuitively l realised that the Piston Diameter measurement is sensitive to the VAS measurements. The measurement of the piston diameter should be at the mid point of the surround (according to a pop up error message l got).

I am going to build a clamp to support the drivers in free air and do more tests.

In the case of Morton’s drivers l think the driver measurements of the FS are most likely accurate.
Lowering the FS with added mass does increase the Qts. The question is why the drivers have a lower FS than spec.

Why also the higher than spec RE?

At a guess the surround has been over stretched beyond the useable limit.
Obviously something is not right.

My suggestion is to have them looked at by a technician. They may be able to re build the drivers.

Holding the driver firmly alters the figures a bit. I am going to make a clamp jig to support the drivers firmly with space under them.

[Robh3606]

My 2245 all measured with an FS of between 26 and 30 hertz.


l dare say this is the case with most 2245 currently in use where the driver has been re foamed (Edgewound will be shaking his head...told you soooooo).

Hello Ian

Wow that's something to remember. I have re-foamed a decent amount of drivers over the years including 2 pairs of 2235 and Fs were within a hertz or 2. Same with 2122's 2108 125 2214 LE-14 128's the only common was all foams from Rick Cobb. Guess it does matter who you get them from!

Rob:)

[Ian Mackenzie]

Hello Ian

Wow that's something to remember. I have re-foamed a decent amount of drivers over the years including 2 pairs of 2235 and Fs were within a hertz or 2. Same with 2122's 2108 125 2214 LE-14 128's the only common was all foams from Rick Cobb. Guess it does matter who you get them from!

Rob:)

Well thats right.

Mind you they work all the same.

I will model the parameters once I do the Vas in a test enclosure.

Ian

[Ian Mackenzie]

Morten,

You posted a pdf indicating the measured Re values were/are in the neighborhood of 22 ohms ( post 3 ) .

This should be a MAJOR red flag that either; those 2245H woofers Voice-Coils are toast ( and you've made a bad purchase ) or your measurement system or methodology are faulty.

Normally when presented with a figure so far from official spec. ( someone more experienced in these matters ) would simply go get their MultiMeter ( then set it to measure DC resistances > DCR ) to get a second opinion ( before posting their doubts to social medfia ).

Your failure to make ( or mention making ) a dedicated DCR test using an alternate method ( has me shaking my head ).

:)

Hi Earl,

I was re reading this thread before my last post.

This is a person who has come here for help and it’s only his third post.

Most LHS member were helpful and considerate with their posts.

Isn’t your notion that a visiting person looking for help should be familiar with testing a loudspeaker rhetorical? Why else would they visit this site?

How could you have done this better?

[FE3T]

This 2245H was measured with a S&L Woofertester 2
At the time measurement was done the driver was just reconed with one of the last oem kits sold in Norway by me.

Meaasured with the driver in horisontal position, screved too woden blocks positioning it 4" above the floor.
Test mass was a roll of duct tape weigthed on a kitchen scale.

As seen there are some parametres that are suspicious. MMS are very high (More than tvice the real weigth of the cone Assembly), Vas are low, BL are very high, and Qms are high.

I vaguely remember somone claimed that the 2245H are very difficult too get dependable measurements of.
so i use factory specs for my simulations rather than trusting DATS or WT2.

The driver are currently in use in a 8cuft cabinet ported too 26hz, 6db boost at 26hz Q=2
20Hz 24db/oct HP. 50hz 18db/oct LP. Performs great.



Buffer[1] Vas Ver 7.01
Completed: Sat Jul 16 20:11:52 2016
Drive level 100.000% [3.417 mA]
Sine,LoZP(LV/LA)->Vas,19 pts
;------------------------------------------------------------------
Re = 5.0568 ohms
Fs = 19.2550 Hz
Zmax = 158.7390 ohms
Qes = 0.2330
Qms = 7.0824
Qts = 0.2256
Le = 1.3190 mH (at 1 kHz)
Diam = 400.0000 mm ( 15.7480 in )
Sd = 125663.7126 mm^2(194.7791 in^2)
Vas = 507.3981 L ( 17.9186 ft^3)
BL = 28.1596 N/A
Mms = 302.0576 g
Cms = 226.1853 uM/N
Kms = 4421.1528 N/M
Rms = 5.1598 R mechanical
Efficiency = 1.4605 %
Sensitivity= 93.6630 dB @1W/1m
Sensitivity= 95.6551 dB @2.83Vrms/1m
Krm = 9.195E-03 ohms Freq dependent resistance
Erm = 702.906E-03 Rem=Krm*(2*pi*f)^Erm
Kxm = 16.504E-03 Henries Freq dependent reactance
Exm = 711.469E-03 Xem=Kxm*(2*pi*f)^Exm
;------------------------------------------------------------------
Ftest = 12.005 Hz
Ftest/Fms = 0.6235
Test Mass used = 475.0000 g (Equal to 95.0 nickels)
Test Mass (Ft=Fms*0.90) = 70.853 g (Add -404.147g for Ft=17.329)
Test Mass (Ft=Fms*0.75) = 234.934 g (Add -240.066g for Ft=14.441)

[RMC]

Hi FE3T,

Thanks for posting your actual results.

RE "At the time measurement was done the driver was just reconed with one of the last oem kits sold in Norway by me."

Was there a good driver break-in period before measurements?

RE "so i use factory specs for my simulations rather than trusting DATS or WT2."

Why not use your actual ones, not confident? Have you tried it? I wonder if there would have been major box modeling differences had you used your actual measured parameters instead of those from JBL's table. Might have been interesting to see.

Recently, you're the second member I see posting actual data measured following original recones, and some of their numbers showing notable parameter differences compared to JBL T/S table, pic below.

This is NOT an evaluation of your work nor of your driver. I simply find this puzzling for the second time considering its ORIGINAL recones, not aftermarket stuff... Plus you indicate "Performs great."

Some TS shown are pretty much on par, while others are off. I had noted quickly Qts, Vas, BL, Mms, No. The most intriguing one being Mms since this one should not vary much...

The reason for my post is that recently I've been looking into this TS parameters shift issue and its possible impact or not in real life use, plus to what extent.

I found from credible sources interesting things that may explain why drivers having some shifted parameters can nevertheless perform acceptably. Hopefully I'll be able to report that soon, though its income tax return time here now, I'm loaded, it may have to wait a little. Best Regards,

Richard

86019

[Ian Mackenzie]

This 2245H was measured with a S&L Woofertester 2
At the time measurement was done the driver was just reconed with one of the last oem kits sold in Norway by me.

Meaasured with the driver in horisontal position, screved too woden blocks positioning it 4" above the floor.
Test mass was a roll of duct tape weigthed on a kitchen scale.

As seen there are some parametres that are suspicious. MMS are very high (More than tvice the real weigth of the cone Assembly), Vas are low, BL are very high, and Qms are high.

I vaguely remember somone claimed that the 2245H are very difficult too get dependable measurements of.
so i use factory specs for my simulations rather than trusting DATS or WT2.

The driver are currently in use in a 8cuft cabinet ported too 26hz, 6db boost at 26hz Q=2
20Hz 24db/oct HP. 50hz 18db/oct LP. Performs great.



Buffer[1] Vas Ver 7.01
Completed: Sat Jul 16 20:11:52 2016
Drive level 100.000% [3.417 mA]
Sine,LoZP(LV/LA)->Vas,19 pts
;------------------------------------------------------------------
Re = 5.0568 ohms
Fs = 19.2550 Hz
Zmax = 158.7390 ohms
Qes = 0.2330
Qms = 7.0824
Qts = 0.2256
Le = 1.3190 mH (at 1 kHz)
Diam = 400.0000 mm ( 15.7480 in )
Sd = 125663.7126 mm^2(194.7791 in^2)
Vas = 507.3981 L ( 17.9186 ft^3)
BL = 28.1596 N/A
Mms = 302.0576 g
Cms = 226.1853 uM/N
Kms = 4421.1528 N/M
Rms = 5.1598 R mechanical
Efficiency = 1.4605 %
Sensitivity= 93.6630 dB @1W/1m
Sensitivity= 95.6551 dB @2.83Vrms/1m
Krm = 9.195E-03 ohms Freq dependent resistance
Erm = 702.906E-03 Rem=Krm*(2*pi*f)^Erm
Kxm = 16.504E-03 Henries Freq dependent reactance
Exm = 711.469E-03 Xem=Kxm*(2*pi*f)^Exm
;------------------------------------------------------------------
Ftest = 12.005 Hz
Ftest/Fms = 0.6235
Test Mass used = 475.0000 g (Equal to 95.0 nickels)
Test Mass (Ft=Fms*0.90) = 70.853 g (Add -404.147g for Ft=17.329)
Test Mass (Ft=Fms*0.75) = 234.934 g (Add -240.066g for Ft=14.441)

Thank you for posting your data.

You raise from good points about testing of these drivers and when to rely on DATSV3 data?

The question then is how and when can the DATSv3 be reliably used?
The problem its there are unknown unknowns like is the driver at fault, the measurement method or the measurement device?

I am told those big speakers are tough to measure, particularly with the added mass method. The resonance peak should be high and sharp. If it looks heavily damped, the coil could have a shorted turn. A shorted turn won’t change the DC much but it will screw up the parameters big time.










I plan to look into this further

[Kay Pirinha]

I have the suspicion that there's something very bad with the software or the measurement setup Morton has been using. Look at his graphs and results in #11. Both drivers show a minimum impedance at some frequency between 100 and 200 Hz with a value of about 6 ohms in both cases, which appears to be reasonable for an 8 ohms speaker. As we know that a speaker's impedance never can fall below it's Re value, I'm heavily scratching my head how the software had calculated it to about 22 ohms? And if the software - or setup - fails in calculating such an easily to determine parameter, how can one trust it's other results?

Keep your head up, Morton!

Best regards!

[ds23man]

The only way to measure T&S accurate on drivers with a low FS is the added volume method. With added mass you will reach the limits of the measurement setup ( soundcard). We know that the provided T&S parameters by JBL are pretty accurate. To check the condition of a JBL driver only two parameters are needed: Dc resistance and Fs. To check the Fs you only need a sine generator and observe the max cone exursion at which frequency it occurs in free air.

[Kay Pirinha]

Another possibility (how it is done by me): Arrange a flying lead lab bench setup: Sinusoidal signal generator - power amplifier - lab bench power supply. Tune frequency and watch the amplifier's power consumption. The frequency that results in the amplifier's lowest power consumption is fs.

Best regards!

[Kay Pirinha]

Some more thoughts on the possibility of a blocked back vent: I don't follow the argument that a blocked vent only will affect Q, but not fs. As this adds some more stiffness due to the locked air between the - hopefully air tight - dust cap and the obstruction, I'd expect a major influence on fs - it will raise (not fall!) dramatically, as well as Vas will dramatically be decreased.

Best regards!

[ds23man]

Another possibility (how it is done by me): Arrange a flying lead lab bench setup: Sinusoidal signal generator - power amplifier - lab bench power supply. Tune frequency and watch the amplifier's power consumption. The frequency that results in the amplifier's lowest power consumption is fs.

Best regards!

Nice method! I call my test the dribble test, with some power the driver won't stay put at Fs.:D