My 2451 + D8R2450SL diaphragm installation method

[jmpsmash]

Installing JBL compression driver diaphragm

Just want to share my experience and technique in installing a set of D8R2452SL into my JBL 2451J compression driver.

I got a pair of 2451J from ebay a little while ago and after doing some reading here I got tempted in replacing the original factory diaphragms with D8R2450SL with aquapas coating (which I got from speakerexchange).

Before removing the original diaphragm I measured, EQed and paired the 2451J with a Purifi 8” base cabinet. TBH, they measured and sounded really good already. Usable from 750Hz upwards and H2 at 65dB to 60dB. Some resonance starts showing from 7kHz onwards on H2 plot and breaking down starts to show on FR around 13kHz.




Even then, after reading reports of the SL diaphragm sounding, I wanted to try it.

I have read many threads and there are some details on how to install diaphragm. I have previously done it on a different driver so I am familiar with the general process but only by eyeball and ears.

So I installed it, trying my best to eyeball putting the coil into the gap and playing different note to make sure there is no rubbing. And while the FR looks great, the distortion looks rather bad. Not only is H2 higher than the factory diag by 10dB, H3 is off the chart, way higher than H2. I unscrewed the diag and made a 2nd attempt and it was partly better but H3 is still out of control.




So I was confused, either this is the best out of this SL diag, or eyeballing doesn’t work.

I figured I need to have more precision than eyeball the coil alignment and then I remember I bought a depth gauge from aliexpress recently for another purpose. It can be used to measure the range for alignment and precisely place the voice coil. It has 0.01mm precision which is absolutely necessary it turns out. I had to hack it by ziptie it onto the driver body but it was secure enough.

The range of alignment when trying to rock the diag in the gap is really small. Max is 0.1mm and sometimes just 0.07mm. So I need to center it within that range.

Another difficulty is to make all parts of the coil aligned. This I did by aligning the coil by quadrants. Start with aligning and fixing the alignment by very softly tighten one screw, measure the range for the coil movement, and tap it to align to the middle of the range, and then move to the next quadrant. Usually by the 3rd quadrant the motion is already fixed.

The result is a big improvement, H3 is now well controlled but H2 is still higher than the factory diag at around 55dB to 50dB. Not as good but I guess HD is not the absolute judgement of sound quality. The SL dia is better in upper treble, the FR is flatter/smoother, and also there is no sign of rising distortion.

In real music testing, it does sound smoother, not as harsh/piercing as the original diag, but perhaps very slightly less details/texture as a tradeoff.

L/R matching is also very good. Both FR and HD are almost matching.

So far, I am happy with the swap.

[jmpsmash]

Here is the final FR and HD, and the little measurement jig.

[Mr. Widget]

Very cool... thanks for taking the time to share this valuable information!


Widget

[hjames]

Excellent work, thank you for showing this technique!
I have ordered a depth gauge from Amazon to try this technique on my right horn driver which has some sonic irregularities compared to the left wood horn/driver (2445J drivers).

I also need to chase down some felt to go into the end-caps - (Pretty sure I removed the failing aged foam when I installed new diaphragms Oct '22).
I installed D8R2452SL diaphragms per Subwoof, and used test tones on my cellphone with the T-amp to center them.

Installing JBL compression driver diaphragm

Just want to share my experience and technique in installing a set of D8R2452SL into my JBL 2451J compression driver.

I got a pair of 2451J from ebay a little while ago and after doing some reading here I got tempted in replacing the original factory diaphragms with D8R2450SL with aquapas coating (which I got from speakerexchange).

Even then, after reading reports of the SL diaphragm sounding, I wanted to try it.

So I was confused, either this is the best out of this SL diag, or eyeballing doesn’t work.

I figured I need to have more precision than eyeball the coil alignment and then I remember I bought a depth gauge from aliexpress recently for another purpose. It can be used to measure the range for alignment and precisely place the voice coil. It has 0.01mm precision which is absolutely necessary it turns out. I had to hack it by ziptie it onto the driver body but it was secure enough.

The range of alignment when trying to rock the diag in the gap is really small. Max is 0.1mm and sometimes just 0.07mm. So I need to center it within that range.

Another difficulty is to make all parts of the coil aligned. This I did by aligning the coil by quadrants. Start with aligning and fixing the alignment by very softly tighten one screw, measure the range for the coil movement, and tap it to align to the middle of the range, and then move to the next quadrant. Usually by the 3rd quadrant the motion is already fixed.

The result is a big improvement, H3 is now well controlled but H2 is still higher than the factory diag at around 55dB to 50dB. Not as good but I guess HD is not the absolute judgement of sound quality. The SL dia is better in upper treble, the FR is flatter/smoother, and also there is no sign of rising distortion.

In real music testing, it does sound smoother, not as harsh/piercing as the original diag, but perhaps very slightly less details/texture as a tradeoff.

L/R matching is also very good. Both FR and HD are almost matching.

So far, I am happy with the swap.

[jmpsmash]

After spending a couple of hours playing music on this last night. Today I measured it again and H2 went down to -60dB. That's interesting!

[engineerjoe]

It was covered here before. Does this diaphragm fit the 2450 and the 2446 drivers?I remember reading that it would improve the high end response curve on my older stock drivers.

[jmpsmash]

It certainly fits the 2450 as it is a D8R2450SL dia.

[Earl K]

It certainly fits the 2450 as it is a D8R2450SL dia.

In your first post you state that the replacement diaphragm you used was a D82452SL.

So, which is it??

[jmpsmash]

My bad. It's a typo. It is a D8R2450SL.

Not sure if there is a difference between the two. Others might know.

[sebackman]

Great info and graphs.

A few observations.

I like your "depth gauge" approach but it may be valuable to consider that the outside of the driver is a "cover" which is not always centered and probably glued to the core. I have seen several cores where the small space between the cover and the core is concentric (with the back cap removed). That may mean that even if it is centered compared to the out rim it will not be so in the gap.

For the quadrant approach I do the same but using a signal generator. I use a good analogue signal generator providing a clean sinus wave from about 300Hz to at least 10kHz.

I found that the 1,5" cores typically are sensitive to resonances in the 500-600Hz sector and 1300-1400 Hz area. If that is fine I rarely see other problems if the gaps are clean. It is almost always audible when sweeping a sine wave. The sound changes to reminding of the sound from a Square wave. It is also visible on a impedance measure with something like a DATS3 or similar.

The drives are sensitive to if the back cap is on or off and also must see free air for at least 0,5-1m in front. If you put it on a table with the back cap facing down it must be on a soft surface as a hard surface may affect the back cap. I use some 1 inch foam on a short pedestal up from table when measuring the driver on its back.

And always change to foam to felt when you have it open. The foam may look fine but if you push with your finger it often falls apart. It is easy as it sits with a large double-sided self adhesive tape. It comes out easy. I typically glue the woolen felt to back back with contact glue. The size of the felt is easy too. The outer diameter of the back cap is perfect as a template it will cover the sides on the inside. You just have to cut a small piece for the terminals on the inside of the back cap. I apply contact adhesive on the inside (including the angled sides) and just put in the felt circle in. Then I put a quarter or a washer in the center to pust down the felt on the resessed square in the middle of the inside of the back cap and the just apply suitable weight for a few hours to. I often use a round flat bottom breakfast bowl that I fill with water.

I typically run about 80-75 db without horn with no capacitor but a series resistor of about 20-50 ohms.

One thing to consider with the newer Mexico diaphragms is that the lacquer on the coil often has blobs and spill. I typically "feel" with the fingers on the coil and on the inside to detect blobs and often find that I have to adjust them with a knife or a file.

It may prove more or less impossible to get a "Mexico diaphragm" with a lacquer blob silent and that can be darn annoying when the cores may be clean and in good shape.

The phase plugs are glued together from sections so they can differ quite a bit from core to core. It sounds like you were lucky finding two good cores with similar paterns. Good news.

I did spend quite some time a few weeks ago to find two well matched drivers from a sample of quite a numer of cores and diaphragms. I did find a few good pairs but it was very time consuming and then several of the cores were NOS and never used.

Have a good weekend
//Rob

[engineerjoe]

great information!

[jmpsmash]

sebackman, Thanks for the wonderful tips! I will keep them in mind next time I change diaphragm, which I am contemplating swapping in another pair of D8R2450SL into my 2441. Given that the dia that came with the 2441 are very old, new genuine D16R2441 are unavailable, and how well the D842450SL look on the measurement. And that pair definitely need some new foam/pads!

I'd just like to point out about the measurement method. I wasn't using the driver body as an absolute reference. I wiggle the dia/coil to it's limits inside the gap, figure out the range of motion (usually around 0.07mm to 0.1mm), and then fix the dia at the middle of that limit. That I think is the key of the method and give the result some consistency and less guesswork.

As for the matching, I guess I got lucky. I got them from ebay I think a year ago and only managed to have some free time to work on it recently (caught covid and was isolating myself in my room for a week, so that was a blessing in disguise! )

Great info and graphs.

A few observations.

I like your "depth gauge" approach but it may be valuable to consider that the outside of the driver is a "cover" which is not always centered and probably glued to the core. I have seen several cores where the small space between the cover and the core is concentric (with the back cap removed). That may mean that even if it is centered compared to the out rim it will not be so in the gap.

For the quadrant approach I do the same but using a signal generator. I use a good analogue signal generator providing a clean sinus wave from about 300Hz to at least 10kHz.

I found that the 1,5" cores typically are sensitive to resonances in the 500-600Hz sector and 1300-1400 Hz area. If that is fine I rarely see other problems if the gaps are clean. It is almost always audible when sweeping a sine wave. The sound changes to reminding of the sound from a Square wave. It is also visible on a impedance measure with something like a DATS3 or similar.

The drives are sensitive to if the back cap is on or off and also must see free air for at least 0,5-1m in front. If you put it on a table with the back cap facing down it must be on a soft surface as a hard surface may affect the back cap. I use some 1 inch foam on a short pedestal up from table when measuring the driver on its back.

And always change to foam to felt when you have it open. The foam may look fine but if you push with your finger it often falls apart. It is easy as it sits with a large double-sided self adhesive tape. It comes out easy. I typically glue the woolen felt to back back with contact glue. The size of the felt is easy too. The outer diameter of the back cap is perfect as a template it will cover the sides on the inside. You just have to cut a small piece for the terminals on the inside of the back cap. I apply contact adhesive on the inside (including the angled sides) and just put in the felt circle in. Then I put a quarter or a washer in the center to pust down the felt on the resessed square in the middle of the inside of the back cap and the just apply suitable weight for a few hours to. I often use a round flat bottom breakfast bowl that I fill with water.

I typically run about 80-75 db without horn with no capacitor but a series resistor of about 20-50 ohms.

One thing to consider with the newer Mexico diaphragms is that the lacquer on the coil often has blobs and spill. I typically "feel" with the fingers on the coil and on the inside to detect blobs and often find that I have to adjust them with a knife or a file.

It may prove more or less impossible to get a "Mexico diaphragm" with a lacquer blob silent and that can be darn annoying when the cores may be clean and in good shape.

The phase plugs are glued together from sections so they can differ quite a bit from core to core. It sounds like you were lucky finding two good cores with similar paterns. Good news.

I did spend quite some time a few weeks ago to find two well matched drivers from a sample of quite a numer of cores and diaphragms. I did find a few good pairs but it was very time consuming and then several of the cores were NOS and never used.

Have a good weekend
//Rob