Passive CD compensation

I'm being a bit lazy here and wonder if there is someone here that has a schematic for a CD horn compensation circuit to put between an active crossover (analogue without CD comp.) and a 2371 horn with a 242X driver.

A friend is just starting in the JBL world and will build two way system on a budget. Woofers are 2225 and over that a 2418 driver on a 2371 horn. He will be going active XO but the XO has no CD compensation built in. A compromise to keep costs at bay.

Should be just a RCL link to ground in parallell with the driver, but as I normally use DSP I don't have this at hand. And unfortunately I don't have any 1" drivers to experiment on either.

All input appreciated

Kind regards

//Rob

Hello Rob


You could use the basic layout in the 4430 as an example without the mid pad of if you need parametric you could use what I posted here as a starter.

You might get away with just a series capacitor depending on the horn/CD raw response.

Look for a PA box using the 2371 with a passive crossover and start with that for the compensation. Look in the G SERIES

I have a 2418 but the wrong horn 2370.

Rob :)


https://www.audioheritage.org/vbulle...rmat+listening


https://www.audioheritage.org/vbulle...peaker-Systems

http://www.rcfaudio.ro/suggested-des...club-15-LP.pdf

Did you mean the 2370A horn? I looked it up in a jbl parts list and my impression is the 2371 is a screw mount version of the 2370A.

https://jblpro.com/en-US/product_documents/2370a-pdf

I modelled the compensation for that horn recently for someone. It only needs about 2-3 db compensation above 4500 hertz. To answer your question yes it’s a series resister + capacitor in parallel with the fixed L pad.

This horn does not need the typical CD compensation if you check out the jbl 2370A spec sheet. I used a 1200 LR 12 db network and with the horns nature roll off below 2000 hertz it should work out fine.

The bump in the response from 1600-4,000 could be notched out with a series tank passive EQ but the complexity is hardly worthwhile with this horn. It’s not a particularly good horn for sound reproduction. It’s a musical instrument SR horn.

I will upload the schematic later.

If you look at the schematics for that horn/CD combo they are 2.5-5 uF series capacitor and a .4mH shunt to ground as the crossover. I would just start with the series cap and go from there. Here are a couple of examples.

You may want to add a fixed pad to keep the hiss down. I do that and it works wonders towards a silent system.


Rob :)

Quote:

---

Originally Posted by Robh3606

You may want to add a fixed pad to keep the hiss down. I do that and it works wonders towards a silent system.

---

Or you can buy uber expensive electronics. Another way to achieve the same goal.:D


Widget

Quote:

---

Originally Posted by Mr. Widget

Or you can buy uber expensive electronics. Another way to achieve the same goal.:D


Widget

---

:Dyeah that's one way to go!

That’s funny…Lol


I checked out the G series here
https://jblpro.com/en/site_elements/...es-information

That crossover would work but with a bit more investigation it can be improved for matching the 2371 horn to the 2225 woofer.

Looking at those G systems in a bit more detail the G series woofers are higher sensitivity (102 db) than the 2225 (97db) and have a more extended mid range with the light weight aluminium VC. The crossover point in these systems varies from 1.5 to 3 khertz.

The 2225 woofer on the other hand has a max recommended crossover according to the spec sheet link below of 1200hertz. It’s a old jbl design. So a custom passive crossover network designed is warranted.

https://jblpro.com/en/site_elements/2225h-j-information

This is what l came up with. I will attach a better image later today when l have more time.

The low pass filter has impedance control for the 2225 Re and Le. I have not simulated the 2225 low pass filter design. Looking at the 2225 response in the spec sheet the Linear X Leap LR 12 db passive filter set for 8 ohms and 1200 hertz should work fine. The high pass network is also a LR 12 db network with a crossover point of 1200 hertz. The attenuation for the 2371 is set at -13 db. A passive RC network is optimised to provide a positive shelf EQ above 4,500 hertz to offset the negative HF shelf response of the horn. I have shown alternative R values for adjustment of the EQ for 1, 2 and 3 db of HF shelf. Parts have been specified in the notes.

Note the 2225 voltage drive looks odd because the RC network compensation for the rising impedance of the driver.

Quote:

---

Originally Posted by sebackman

I'm being a bit lazy here and wonder if there is someone here that has a schematic for a CD horn compensation circuit to put between an active crossover (analogue without CD comp.) and a 2371 horn with a 242X driver

//Rob

---

Hello Ian

You do realize he is going to be using an analog active crossover set-up?? All he needs is the horn compensation, attenuation and slope he already has covered.

Rob :)

Getting old …Lol

New glasses?

Well the attenuation and EQ will work fine.

Dear all,

Thank you for the input.

Yes, just looking for the correction link only, not XO. I will probably add a constant impedance attenuator with two resistors also to reduce noise and hum.

My memory is getting worse, but I think I have seen this done with an RCL link placed between the XO and the driver, ie just a capacitor, a coil and a resistor in series i parallell with the driver (and any attenuator if that is used) pushing down the lower mid to compensate for the falling response in high frequency.

Kind regards
//Rob

Hi Rob,


Your friend will need something like the following ( the 8-ohm variable LPad has about 10db of attenuation dialed in );

Attachment 91026

R3+R4 together, represent the variable 8 ohm LPad.

R1 exists here to keep the impedance from dipping below 4 ohms ( the 2418 might not need R1 since it looks to be slightly higher in impedance ).

Also, the 2418 might not need L1 ( a PEQ resonating coil ) since I think the 2418H has a bit more HF than the older 2425/6/7 series of drivers.

Here are Guido's old ( + rough ) measurements of a 2418 vs a 2416 on a 2342 horn ( the 2416 is pretty much a stripped down 2426 ).

https://www.audioheritage.org/vbulle...1&d=1109198840

https://www.audioheritage.org/vbulle...1&d=1109199101

One can see the 2418 has a bit more HF than the 2416 .

:)

I think we are all talking’s about the same thing here.

Looking at the jbl 2370A spec sheet below the response can be interpreted as a bump between 2-4 khertz and a couple of db loss of level up to 16 khertz. The 2370A and the 2371 only differ in the screw mounted driver as far as l am aware. Jbl did variations of horn mounting with their odd and poorly documented model numbering systems over the years.

It’s not a CD EQ problem in the traditional sense of a 6 db per octave lift above 3-4 k hertz.

To keep it simple and to correct my misunderstanding of the passive EQ only requirements please find attached a re worked schematic and simulations.

2371 horn and compression driver model (Green Curve)
2371 horn and compression driver model with active crossover high pass filter only Purple Curve)
2371 horn and compression driver model with active crossover high pass filter and the passive attenuation and passive EQ filters (Blue Curve)

Horn and driver model
I spent some time on the horn model using JBL tech sheet data and modelled the horn response as accurately as possible in Spice. The curve is directly comparable to the response curve scale in the JBL sheet The model is assumed to be essentially a resistive load. The horn loading will modulate the impedance of the actual driver. A 12 ohm impedance is assumed for the 2428J real drive load. If the 2418J version remove the 20R resister. Note the divisions in the attached simulations are 5 db for close inspection of the data.

Notes
The compression driver impedance ripple is well damped with a 20R parallel resister and the 2.3R of shunted resistance used for attenuation of about -11 db. The impact of either an H or J impedance will be minimal with the low impedance presented at the attenuation network. This should provide adequate protection for the compression driver with need for a protection capacitor.

The active crossover is assumed to be LR 4th order 1200hertz Hp and LP. The horn tracks the Opt Hp 1.2 khertz curve closely. The mounting of horn and the woofer on the baffle will impact on the final response. Given the 24 db slopes the interactions between the drivers should be confined to a narrow band in the crossover region.

The passive EQ control consists of two functions.
1. Level reduction of 11db
2. A two stage passive EQ network which provides removal of the 2-4 khertz bump and a mild 2-3 db high pass Shelf EQ above 4500 hertz.

I would expect the horn to sound as good as it’s ever going given it’s a musical instrument SR horn.
Note the off axis response in the JBL tech sheet.

Build the network just like the schematic and use 10 watt power resisters.

Edit l have added the JBL impedance data on a 2426H and noted on use of the 2418J or 2418H version.

Big Thank You Earl & Ian,

I will do a mockup and see if they get the "ok sign" by my friend. I should have components to make a set of each as needed.

This group always delivers. Brilliant.

Thank you!

//Rob