Jump to content

A JBL white paper and a bunch of questions about what’s possible


Recommended Posts

I found a JBL 4367 white paper today that has piqued my interest in a couple things. 

First I should mention that it’s really more of a white paper lite, something the marketing guys worked out with a couple engineers. But it’s still a good read. 
 

The first thing that caught my eye was the biased crossover using signal power rather than a battery or external source of power. Not sure I understand how they pull this off. 
 

The second thing that looked cool was the passive eq they allude to that corrects the raw frequency response of the driver in the horn. 
 

The 4367s use differential drive woofers and some advanced compression drivers. 
 

If you have a minute please go check it out 

https://www.jblsynthesis.com/on/demandware.static/-/Sites-masterCatalog_Harman/default/dwc07f5318/pdfs/JBL Synthesis 4367 White Paper.pdf

 

 

Link to comment
Share on other sites

12 hours ago, jwgorman said:

The first thing that caught my eye was the biased crossover using signal power rather than a battery or external source of power. Not sure I understand how they pull this off. 

Would be interesting to see a schematic of the diode and resistor circuit for the crossover network's self-bias function.

Link to comment
Share on other sites

20 minutes ago, Khornukopia said:

Would be interesting to see a schematic of the diode and resistor circuit for the crossover network's self-bias function.

Agreed! Any idea how they passively eq’d the compression driver/horn? 

Link to comment
Share on other sites

12 hours ago, jwgorman said:

The second thing that looked cool was the passive eq they allude to that corrects the raw frequency response of the driver in the horn. 

 

Klipsch has been doing this one for ages.  It's why Klipsch internally uses the term "balancing network" instead of crossover network.  It both divides the signal per driver and eq's the driver's response if needed.

Link to comment
Share on other sites

Klipsch’s networks (the classics that have schematics on this forum) do pass specific frequencies to drivers, and by increasing the slope of the crossover they can effect, to a limited extent, how much high treble comes out of a K77 (for example) but beyond that there is no eq. If you look at the raw frequency response of the JBL driver/horn in the white paper I linked to they have provided considerable eq to get their driver/horn flat. 
 

Link to comment
Share on other sites

19 minutes ago, jwgorman said:

Klipsch’s networks (the classics that have schematics on this forum) do pass specific frequencies to drivers, and by increasing the slope of the crossover they can effect, to a limited extent, how much high treble comes out of a K77 (for example) but beyond that there is no eq. If you look at the raw frequency response of the JBL driver/horn in the white paper I linked to they have provided considerable eq to get their driver/horn flat. 
 

 

At Roy Delgado's 2019 speaker class he described the networks doing both functions and showed some examples of the how it is used.  While it's true that it was mostly done near the crossover point, it doesn't seem a great stretch to imagine this applied elsewhere in a driver's operating band if needed.  All I'm saying is that this is not an innovation exclusive to JBL.

  • Like 1
Link to comment
Share on other sites

14 hours ago, jwgorman said:

I found a JBL 4367 white paper today that has piqued my interest in a couple things. 

First I should mention that it’s really more of a white paper lite, something the marketing guys worked out with a couple engineers. But it’s still a good read. 
 

The first thing that caught my eye was the biased crossover using signal power rather than a battery or external source of power. Not sure I understand how they pull this off. 
 

The second thing that looked cool was the passive eq they allude to that corrects the raw frequency response of the driver in the horn. 
 

The 4367s use differential drive woofers and some advanced compression drivers. 
 

If you have a minute please go check it out 

https://www.jblsynthesis.com/on/demandware.static/-/Sites-masterCatalog_Harman/default/dwc07f5318/pdfs/JBL Synthesis 4367 White Paper.pdf

 

 

 

I took a quick look.  I'll have to read this in more detail. 

 

It looks like they are charge coupling the caps using diodes and possibly cutting the cap count in half compared to using a battery.   If I recall, you quadruple the capacitor count (I might be wrong). 

 

When researching my JBL 250s, in some old posts Greg Timbers (JBL's 'Roy' in the 80s) didn't believe in charge coupling that some users were doing.  A while later after some research and testing, he recommended this in threads for hobbyist for their old speakers and the Everest used charge coupling.  Seems interesting, but huge parts count int he JBLs as they have very complicated crossovers compared to a Type A. 

 

The diode method seems interesting.  I still need to do the crossovers in my JBLs but that is a ways off. 

 

  • Like 1
Link to comment
Share on other sites

 

56 minutes ago, jwgorman said:

Klipsch’s networks (the classics that have schematics on this forum) do pass specific frequencies to drivers, and by increasing the slope of the crossover they can effect, to a limited extent, how much high treble comes out of a K77 (for example) but beyond that there is no eq.

 

"Modern" Klipsch networks, like the AL-4 pictured here, do smooth the frequency response.

 

IMG_1555.JPG.3151df3e559ae9f8494aec1afb42afa9.JPG

Link to comment
Share on other sites

They explained in the paper they rectify the AC signal to DC with diodes, same way the electrovoice tweeter protection device creates DC to power the relay to trip.

 

To me the whole DC bias is silly compared to the distortion caused from nonlinear rectifier load. The charge currents might diminish to a very low value once a charge is built up on the capacitors. I just don't see how a polypropylene capacitor with a very low dissipation factor already show any non-linearities around the zero crossing point. I can see the argument if they were using aluminum electrolytic capacitors but polypropylene no.

 

Just marketing BS to sound cutting edge and sell speakers.

Link to comment
Share on other sites

10 minutes ago, jwgorman said:

That good to know. I’m very curious as to the passive correction done in the 4367. 

From about 8db down at 10k to essentially flat. I have no doubt Roy has this figured out but I would like to know how. 

 

The paper states "The UHF control provides similar adjustment from 4 kHz to beyond 20 kHz"

 

The frequency chart of just the driver is with a flat input frequency, the network adjusts levels to compensate for driver inadequacies.

Link to comment
Share on other sites

27 minutes ago, jwgorman said:

That good to know. I’m very curious as to the passive correction done in the 4367. 

From about 8db down at 10k to essentially flat. I have no doubt Roy has this figured out but I would like to know how. 
 

232C8B3F-3C48-4780-8B98-327A47907286.jpeg

20A3185E-789C-47A1-B31D-9FF0C913AD1D.jpeg

2033156175_ScreenShot2022-02-04at1_52_55PM.png.e8e0bd4c7e1c93cc9d3cf4e4a65b81fa.pngHere is one form of passive tweeter EQ. The L pad is bypassed by a cap, and so can flatten out (to an extent) a tweeter's response. Don Keele clued me in to this many years ago.

Link to comment
Share on other sites

24 minutes ago, boom3 said:

2033156175_ScreenShot2022-02-04at1_52_55PM.png.e8e0bd4c7e1c93cc9d3cf4e4a65b81fa.pngHere is one form of passive tweeter EQ. The L pad is bypassed by a cap, and so can flatten out (to an extent) a tweeter's response. Don Keele clued me in to this many years ago.

Thanks. I was actually aware of this. It makes sense that, given the correct capacitance, that would help bypass hf attenuation. Do you think it could be as simple as a bypass cap in the attenuation circuit? Or maybe several caps connected to the hf and uhf boost knobs on the front panel?  

Link to comment
Share on other sites

A picture is worth a thousand words.

 

I was curious why in gods name would they even consider DC bias for polypropylene capacitors. The white paper failed to mention they also use bipolar electrolytics which of course has higher dissipation factor and nonlinearities, and where you usually see people using a DC bias.

 

The white paper also states using air core inductors because they don't suffer hysteresis from core saturation. The extra DCR from all the extra copper needed to get the same inductance as regular ferrite core inductor is in my opinion more an issue instead of just sizing the ferrite core inductor high enough to not saturate. I have never been wowed ever when someone has changed a regular core inductor to an air core inductor. IF you can get the same DCR which I highly doubt without it being massive in size yes I can see the advantage but normally I see them about double the DCR which I would think it would matter on how mechanically damped the woofer is and the increased DCR is moot. Either way there are always tradeoffs, I just don't see a huge difference in sound going air core if the regular core was sized adequately. Measurements confirm this.

 

The DC bias now makes sense getting a look at the crossovers, they use cheap electrolytics in their networks.

 

spacer.png

Link to comment
Share on other sites

20 minutes ago, jwgorman said:

Do you think it could be as simple as a bypass cap in the attenuation circuit? Or maybe several caps connected to the hf and uhf boost knobs on the front panel?  

 

It's just how capacitors naturally function in relation to frequency, at lower frequencies where the driver has higher output the impedance of the capacitor is higher giving a signal drop across it effectively reducing the signal to the driver, as frequency increases into the lower output region of the frequency plot the capacitor has decreased it's impedance not giving a drop across it, this naturally flattens the driver plot out.

 

For example the Klipsch AA network has the tweeter down -6db at 5kHz compared to input of the network. This -6db loss across the filter network is reduced as frequency increases.

Link to comment
Share on other sites

Basically, don't look at it like 3kHz-8kHz is 0db and the engineer needs to boost 8kHz-20kHz +6db or whatever, look at it as they are reducing the 3kHz-8kHz by -6db and due to filter slope the signal returns to 0db where the driver would have it's low output region, 8kHz-20kHz.

 

Numbers not exact, just an example.

  • Like 1
Link to comment
Share on other sites

17 hours ago, jwgorman said:

Thanks. I was actually aware of this. It makes sense that, given the correct capacitance, that would help bypass hf attenuation. Do you think it could be as simple as a bypass cap in the attenuation circuit? Or maybe several caps connected to the hf and uhf boost knobs on the front panel?  

Can't speak to the JBL circuit, I know that what I have works for me, the cap is a 5 uF which in this application bypasses the L pad above 8 KHz. I did this to compensate for my old ears, so the Great Heil retains its characteristic rising response, meaning it is "flat to my ears" 

 

Infinity used a DC bias in their Kappa speakers crossover about 25 years ago. I'd like to see an AES paper that really discuses this from a scientific POV, and not just marketing. For this to make any difference, the caps would have to be very closely matched and good film types to begin with. 

Link to comment
Share on other sites

So, if you can electrically bias your electrolytic caps to avoid their hysteresis, you should also be able to magnetically bias your iron-core inductors to avoid their hysteresis, right? This might only work with I-core inductors, where you can stick a magnet on the end...

Link to comment
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

 Share

×
×
  • Create New...