PDA

View Full Version : Another DIY 4425



John W
09-07-2005, 02:39 PM
Another DIY 4425



I am in the process of gathering the components for a pair of 4425 monitors. After reading parts of the quick and dirty 4430 thread I think I want to base the crossover off the NL200t3 network. Both networks use a 12 in woofer, the 2416h driver and cross over at 1200hz. Other than that, I am not sure on differences between the 120T3 and the 2214H woofers.

I combined the low pass from the 4425 with the high pass of the NL200t3 and came up with this:

John W
09-07-2005, 02:40 PM
This will be cheaper and easier to implement than the stock 4425 crossover, and if it falls flat, I can always rebuild the high pass. Has anyone tried anything similar?

Also, I can duplicate the 4425 box using the measurements from the Technical Manual, but I am unsure of the port dimensions. For a 1.9ft3 box tuned to 34hz I calculate 2 ports 2 in diameter x 6.25 inches. Can someone with a 4425 measure the ports to verify this?

Thanks for any input.

frank23
09-07-2005, 02:50 PM
Well, I haven't simulated the crossover, but I don't think this will work. The coil in the low pass is 3.8mH, a fairly large value is seems to me. The cap in the high pass is only 2.2uF, a fairly low value. Looks like there will be a large empty hole in the middle...

I'm sorry I don't know what the NL200 is, but I think the 4425 horn needs compensation as it is a full biradial like the 2344 horn on the 4430 and this crossover does not seem to provide it.

I have both 4425 crossovers, horns and 2416 drivers lying around, But I think it would be too much hassle to send them to the USA from Europe.

I think the 4425 crossover is the 3124, look for those schematics. Unless you take a whole different route and design your own, there is no shortcut in JBL crossovers, they do what needs to be done I think.

frank

John W
09-07-2005, 02:57 PM
This is good input:

Here is a link to the l200t3, which I guess is fairly rare:

http://manuals.harman.com/JBL/HOM/Technical%20Sheet/L200t3%20ts.pdf

Zilch
09-07-2005, 03:06 PM
It'll likely work fine. Use the parts list I provided in the Q&D thread for the HP.

I sent the ones I made to Robh, who is building similar.

If you want even simpler, just build the complete NL200t3 for starters.

I played it with 2418H-1 and 2426H, both with 2235H in B380, with good results.

Frank doesn't understand the circuit yet. The 2.2uF cleverly provides the requisite compensation.

The 4425 ports are 2.5" x 6" deep here....

John W
09-07-2005, 04:23 PM
Thanks for the port sizes. I guess for starters I will try the stock NL200t3.

Earl K
09-07-2005, 04:55 PM
John / Zilch

- One thing to watch for in this design exercise is a possible efficiency mismatch between the horn and the woofer . It might be necessary to insert a variable Lpad in the horn circuit for easier matching.

- John, If you tire of waiting around for decent 2214 woofs / consider using the 128H .
It has a long established pedigree . It may require a bit of rework to the lowpass circuit.

- If it was my project I would start with a newer woofer and get a pair of le120h(s) from Harman . I figure, that if they are good enough to matchup with the 050ti tweeter / then they must be quite good . Here's a bit of info .
( The pics are were provided courtesy of Paragon, who pulled this woofer from one of his L90s, for this "Show & Tell" photo-op .

:)

John W
09-07-2005, 05:03 PM
Actually, I already have the 2416H/2342 combo and the 2214H woofers should be arriving next week.

Earl K
09-07-2005, 05:12 PM
Actually, I already have the 2416H/2342 combo and the 2214H woofers should be arriving next week.

- Well then, you're good to go ! :dancin:

- I'll let my comments stand for the future generations reading these archives . ;)

:)

Guido
09-07-2005, 05:13 PM
John, I did 4425 DIY!

Why not build the stock networks? It's not that much work.

Have a look here:
http://www.audioheritage.org/vbulletin/showthread.php?t=4877&highlight=4425+DIY
http://www.audioheritage.org/vbulletin/showthread.php?t=1005&highlight=4425+DIY

John W
09-07-2005, 07:47 PM
I saw these posts, its sort of what got me going on the project.:applaud:

The complete crossover is still an option. Mainly I want to hear what the simple crossover sounds like, then I can work my way up.

Zilch
09-07-2005, 08:20 PM
I guess for starters I will try the stock NL200t3.http://audioheritage.org/vbulletin/showthread.php?t=3838&page=23&pp=15

Note: If you use my layout, rotate the big inductor 90° horizontal. There was some crosstalk between LF & HF sections as built....

John W
09-08-2005, 09:17 AM
Note: If you use my layout, rotate the big inductor 90° horizontal. There was some crosstalk between LF & HF sections as built....



Thanks for the tip. I am debating between the big 15 gauge air core and the Erse 16 gauge iron core inductor. The 3mH air core has a DCR of .64 and the iron core a DCR of .18.

dB = 20 log ((Rl / Z) + 1) = 20 log (((.64 - .18) / 5.6) + 1) = .69 dB

Its not much, but I may need everything I can get out of the bottom end.

Zilch
09-08-2005, 10:40 AM
I am debating between the big 15 gauge air core and the Erse 16 gauge iron core inductor. The 3mH air core has a DCR of .64 and the iron core a DCR of .18.
That'd work too, most likely. JBL used laminated core in the 4430 LP.

I had an L-Pad on the HF output, but never used it. Turns out from the impedances that a 4-Ohm L-Pad would be required so's not to mess up the compensation frequency.

Next ones I'm building will have biased HF, as well. Just a couple of bucks more to do it, since there's only one capacitor. I'd try a pair of 4.3 uF for starters there....

Zilch
09-22-2005, 03:36 PM
http://audioheritage.org/vbulletin/showthread.php?p=70214#post70214

John W
09-22-2005, 03:59 PM
I biased the HF section also.
The initial results were a little too bright. I need to Lpad the HF section to bring it down a little.
Doing this I need to keep the 4ohm impedance, so I put together a little spreadsheet that calculates resistor values for an Lpad. It is set up with 4ohm, but if you use it, you can change the 4 to 8, or 16 or whatever.
I ordered a few combinations of resistors, and decided to finish the cabinets while I wait their arrival.
Next step is to get a more accurate measurement of the frequency response and incorporate the Lpad.

Zilch
09-22-2005, 05:11 PM
Another option: parallel the two gangs of 8-Ohm stereo L-Pads to make 4 Ohms for testing to determine what you ultimately need.

See Parts Express #260-251 or 260-260....

John W
10-19-2005, 07:37 PM
I have finally got these speakers up and running. The current version of the crossover is much like the one at the start of the post except that the 2.2uf was replaced with a pair of 4.3uf and biased. Also I found that a 3db L-Pad brings the top end down to just where I like it. I must say that I really like the sound of these. They are a little larger than I first envisioned, but they really fill the room with nice music.

Zilch
10-19-2005, 09:33 PM
YES! And just beautiful in their simplicity, too....

[In many ways.... :)]

duaneage
10-19-2005, 09:42 PM
Beautiful job. They look store bought original to me.


My counterfiit JBLs

http://audioheritage.org/vbulletin/showthread.php?t=6775&page=2&pp=15&highlight=4411+clones

John W
10-19-2005, 10:11 PM
One thing of note is that I was having a little trouble rounding up some 2 1/2 tubing for the ports. In the end I found some nice stiff cardboard shipping tubes at the local office supply store. They actually come in a range of diameters from 1 1/2 to 3 or maybe bigger and are about 2 feet long. Only cost me $1.50

Zilch
10-19-2005, 10:28 PM
Where did you end up on the LP of the crossover? Stock NL200t3 with 3.0 mH laminate core inductor, or 3.8 mH per 4425?

It's amazing how effectively that simple HF compensation works. Several others I know are building with it now. We'll have more results reported here soon.
The forum owes SOMEBODY a keg of their favorite beer, I say.... :cheers:

John W
10-19-2005, 10:36 PM
I used the 3.8mh inductor, like the 4425 schematic.
It sounds good...but I really need to work on my test setup.
When I measure this stuff using a mic hooked up to the PC its hard to tell whats going on.

Zilch
10-19-2005, 10:41 PM
Yup. We need a thread on DIY RTA.

I can NOT do it without RTA.

No way....

duaneage
10-19-2005, 11:14 PM
Electrical pvc conduit comes in 2 1/2 in ID It is grey and sold at electrical supply houses. After I sank my power feed for the house I ended upwith plenty. They also have 3 inch in electrical pvc conduit.

I scored 5 Inch ID pvc couplings to make midrange enclosures. 6 bucks each.

Earl K
10-20-2005, 06:01 AM
Beautiful John just Beautiful ! :applaud:

:)

John W
10-22-2005, 07:28 PM
Someone was selling a Behringer DSP8024 locally for a good price, so I snatched it up and added a measurement mic. Finally I can get some fairly accurate measurements. It was a nice fall day here in the NW so I pulled a setup outside to preclude any room reflections and hooked it up.

Here is a shot of my test setup:

John W
10-22-2005, 07:30 PM
Here is the pink-noise response measure with a 1db scale grid, followed by the response with a 0.5 db grid.

John W
10-22-2005, 07:32 PM
Overall, not to bad. :)

Zilch
10-22-2005, 07:51 PM
WAY cool! :applaud:

No WONDER they sound so good, huh?

Looks like you're within +/- 3 dB the whole way, if I'm reading it correctly.

JBL'd say "SHIP it!"

Now try the NL200t3's 3.0 mH/20 uF combination in the LF and see it it'd be an improvement.

You probably do want to use the laminate core for a little more low end.

[John's a member of the Lansing Heritage research staff now.... :p]

johnaec
10-22-2005, 08:03 PM
Is that really +/- .5dB 40-20K?? No EQ? Wow! :)

John

John W
10-22-2005, 09:23 PM
I did a little recheck of the scale on the new toy. The dsp8024 has a 1db and a .5db setting. It looks like when in the .5 db setting there is a 5db difference in the grid squares, and for the 1db there is a 10 db spread. So it is more like +/- 5db.

Zilch
10-22-2005, 10:10 PM
I'm ignoring whatever's going on down there at 180 Hz.

It has nothing to do with the crossover (assuming you've got the two 100-Ohm resistors in there,) and a different measurement technique is required to read what a woofer's doing below 250 Hz.

8024 manual nowhere says what the scale is on the RTA display that I can find. Resolution can be set to either 0.5 or 1.0 dB per pixel. I count 8 pixels per dotted line using a magnifier. Thus, the scale is either 4 or 8 dB, depending on the resolution setting.

So, +/- 3 dB, then, and the HF compensation is certainly "happening." :)

Earl K
10-23-2005, 09:44 AM
I count 8 pixels per dotted line using a magnifier. Thus, the scale is either 4 or 8 dB, depending on the resolution setting.

- Yes, you've got it correct .


:)

Earl K
10-23-2005, 09:58 AM
John,

- Those are really cute speakers . Maybe I'm the one who should be buying some le120h woofs for a similar project . ;)

- It's nice you got some measurement stuff . I also use the 8024 as a portable RTA . It's quite utilitarian in that regard .

- Personally, if it was my project / I'd like to see a couple more db of response between 1.6K & 3.2K . But, that's a preference thing. This would unfortunately require a rework of at least one portion of the circuit .

- Since this is an area of great personal preference , that curve could easily be working for you, so I'll shutup now. :p ( Fletcher Munson curves , etc., etc. )

- That perfect Vee "hole" at 160 hz is likely just a "hard surface bounce" reflecting ( at 180° ) into the mic. This will contaminate the FR measurement in that area .


:)

Dutchman
10-25-2005, 08:46 AM
:applaud: These DIY 4425 are very nice!!

This might be a bit off topic but I hope someone has the answer on this one. I'm building cabinets for my DIY 4430's and I'm wondering if the horns must be directly placed underneath the bass-speaker (I hope not from a cosmetical point of view).

All 4425's seem to have a certain angle between the bas-speakers and horns (2214 placed in the middle of the cabinet an the horns on the sides). Is there a reason?? :blink:

John W
10-25-2005, 08:55 AM
I have a follow-on question. Are the 4425 horns ever rotated 90 degrees and the speakers placed on their side?

Zilch
10-25-2005, 01:41 PM
I have a follow-on question. Are the 4425 horns ever rotated 90 degrees and the speakers placed on their side?We've seen an occasional pair of 4430's used sideways with the horns rotated.

Both 4430 and 4425 are symmetrical dispersion horns. Lay them on their side and the distribution should be the same. There are differences in detail, but I would just do it and not rotate the horns....

Zilch
10-25-2005, 01:46 PM
This might be a bit off topic but I hope someone has the answer on this one. I'm building cabinets for my DIY 4430's and I'm wondering if the horns must be directly placed underneath the bass-speaker (I hope not from a cosmetical point of view).If the drivers are not in vertical alignment, the plane of time alignment is not horizontal. Some think that matters, though I could conceive an argument that it might be better in certain situations.

Time alignment's a third- or fourth-order consideration to me. For some, it's prime....

clmrt
10-25-2005, 02:04 PM
What's 1st and 2nd? Group delay? Flat response? Dinner and a show?

Zilch
10-25-2005, 02:44 PM
First order is frequency response. Do these drivers play well together? Is there a compatible crossover region? Is the box sized and tuned for optimum woofer performance?

Second order: How do it sound? Have I selected the best crossover frequency and slope for this combination? Is my HF compensation right? Do these drivers just plain suck in this configuration no matter what I do?

Third order: Sound field. Is this the dispersion and imaging I was aiming for? Are there reflections and/or resonances I gotta attend to? How do I fix this stuff?

Fourth order: Time alignment and aesthetics. I don't give much of a whit about either, but I TRY to pay some gratuitous attention to them.... :p

Ian Mackenzie
10-25-2005, 05:21 PM
There is a paper on this horn and monitor (around somewhere)

Its explains why and how they did what they did at the time.

The LSR paper is also a good read and explains how JBL's evolution of monitor design has moved on in so far as studio applications go. Far more importance is given to controlling early reflections and the off axis power response as opposed to flat on axis response. In these soa designs flat on axis reponse is a given, but only by virtue of complex system design innovations like proprietry drivers & waveguides, elaborate crossover networks.

The 4425 and 4430-35 are what they are so there's not much point messing with them too much.

Ian

JBL 4645
06-21-2006, 11:04 PM
John W

Behringer make a lot of affordable products in the catalogue range, hare in the United Kingdom they are selling them at crazy prices! The site is called “Trunkey”

http://www.turnkey.co.uk/web/homeAction.do?dispatch=homePage (http://www.turnkey.co.uk/web/homeAction.do?dispatch=homePage)

I’m keeping my eye on there products, and will soon upgrade my EQ’s to 1/3 octave for my home cinema, but the model you have there, how much did that sweet baby set you back?

John W
06-22-2006, 08:18 AM
I have a Behringer 8024 which is an older model ultracurve pro unit. I bought mine used for about $120 then also needed to get a Behringer mic and some cables, so altogether about $200.
Normally I only pull it out to use the RTA functions, which it handles very well. The equalizer seems to add a little noise and fuzz so I don't leave it connected to the signal path.
The newer Behringer DEQ2496 is supposed to be a nice step up.

JBL 4645
06-25-2006, 12:22 PM
John W

The more I read the reviews from the that are around, the more and more it convinces me to purchase around x5 EQ for the home cinema, matching of course, as my current EQ set-up is very tricky too get it all aligned and calibrated.

Once that is done, I may look forwards to a brand new (RTA) from the same manufacture, but this is a whole long way off, but it will happen…

Thanks for the comment; I’m sure it will come in very handy.

PS. As for noise, could be the input cables, does the model have (XLR) inputs?:)