Polarity and 4343 wiring

[Tom Loizeaux]

I've been following discussions about polarity in vintage JBL studio monitors and have been involved in discussions specifically about 4343s. Giskard sent me his opinion on correct wiring on these 4343s with the 3143 crossover. I found that I had not wired them in correct polarity and, after making the changes, found the imaging and sound stage to be vastly improved! I did reverse all wires at the drivers from JBL's original specs so these would be compatable with all my other systems, but I feel I've got these all working in phase now.
My immediate reaction is that these 4343s sound like silk. There is no harshness, yet there is plenty of low end, ample mids, and lots of top detail. And now, instruments seem to float from between these cabinets.

FYI JBL's spec:
2231 green to red terminal
2121 blk/wht to red terminal
2420 red/wht to red terminal
2405 gry/wht to red terminal

Thanks again to Giskard for his help with this.

Tom

[boputnam]

Hey, Tom...

A most interesting post - drawn to it like a moth to flame :flamed:

But, I'm not clear on the sequence.

Had you reversed factory original connections and "found that you had not wired them correctly" - i.e., a simple operator error, or did it simply not sound good?

This might get complicated, but the phrase "FYI JBL's spec" seems not right - that seems like your current configuration (rather than "spec"), in that from my reading of the schematic, you have reversed MF, HF and UHF (so they all as a group) are now phased 180-degrees from factory. They now are in-phase with the 2231H: Factory had them out-of-phase - that is LF was (-), while MF, HF, and UHF were all (+) - same configuration as the 4345 (3145).

Your change only impacts the area of crossover (or more correctly "roll-off") overlap between the 2231H and 2121H - that is, you are now getting a doubling of those overlap frequencies, where as-factory would have cancelled those. The inter-relationships between the MF, HF and UHF are unaffected by this, and so imaging for all frequencies above the first high-pass filter should be unchanged - unless you had conflict with "incorrect" interim re-wiring, or something else in your "Tom's Wall of Sound" ;)

I'm intrigued with the result. :hmm: Maybe Giskard's handy-dandy BassBoxPro/x-over Pro can show us what is occurring in the two different configurations? Pretty please?

Background observation: JBL convention on earlier schematics is for each lead from the RED Input cabinet binding post to be the upper lead to each transducer**. On some, JBL labelled the (+) and (-) speaker terminals. In every case where labelled, the solid (non-striped) lead connects to the (+) transducer terminal.

** - except starting with the 4410 and higher, where the BLACK Input cabinet lead took top position on the transducer leads. Maybe this was timed with the beginning of the Big Change at JBL? :confused:

[John Nebel]

Hi Bob,

After reading through all the earlier posts on this topic including yours from June and Giskard's from May, I called JBL and came up with:

LF blk or grn/blk > speaker blk

MF blk/wht > speaker blk

HF red/wht or yel/blk > speaker blk

UHF gry/wht or org/blk > speaker blk

On the 3143 I looked at, the color codes were different. The speaker was purchased in 1980, the schematic on JBL's web site was 1976 so a production change must have been made.

The above scheme would have a clear advantage to production - a red wire goes to a red terminal, a wire with black in
it goes to a black terminal.

The JBL engineer called a "speaker tech downstairs" for confirmation.

JBL seemed quite proud of the fact their old equipment worked well and was still repairable. "Buy it on e-bay, send it in, and we'll fix it."

YMMV

John

[boputnam]

Hey, John...


The above scheme would have a clear advantage to production - a red wire goes to a red terminal, a wire with black in it goes to a black terminal.
That was exactly my point (sometime...) yesterday. For production / assembly purposes, JBL would have:

solid leads -> red terminal
stripped leads -> black terminal

This part of your post:
LF blk or grn/blk > speaker blk
MF blk/wht > speaker blk
HF red/wht or yel/blk > speaker blk
UHF gry/wht or org/blk > speaker blk
is consistent with that.

The connections at/from the network would handle the inter-transducer polarity/phasing configuration as engineered.

[Guido]

Hi John!

Right!

I have the 4343 with "new" wiring.
Only connect cables with black in to black speaker terminal.
I checked what Bo wrote once and it is right, the phase turn is made internally.

BTW if you want to reverse polarity to "non JBL terminology" then of course you have to wire cables with black in to the red speaker terminals.


:)

[4313B]

Whatever...

Wire your loudspeakers the way you want to...

[boputnam]

Whatever... Wire your loudspeakers the way you want to... Oh come-on, please? You know this stuff backwards-and-forwards!! Did I mis-read the schematic? "Factory had them out-of-phase - that is LF was (-), while MF, HF, and UHF were all (+) - same configuration as the 4345 (3145)." No?

I'm intrigued by your suggestion. :bouncy:

[4313B]

I'm not saying anyone misread anything.
If Tom wired his 4343's the way I suggested (the way the system is phased from the factory) and he likes the sound that's great. :)
If anyone wants to wire them different that's great too :)

[boputnam]

...one last time. :banghead:

the way the system is phased from the factory And, was that:
LF (-)
MF (+)
HF (+)
UHF (+)

:confused:

Thanks,

[JBL Dog]

I don't have time to tear my 4343's apart today! :bomb: :banghead: :biting:

This message comes from JBL Dog :band:

[4313B]

The RED terminal on the loudspeaker enclosure goes through the passive network to the RED terminal on the LF and to the BLACK terminals on the MF, HF, and UHF.

It's the same with the 4345.

[boputnam]

The RED terminal on the loudspeaker enclosure goes through the passive network to the RED terminal on the LF and to the BLACK terminals on the MF, HF, and UHF.
Thanks, very kindly, Mr. Engineer (said the geologist, none too certain about these sorts of things).

I may have confused - I was mostly trying to understand whether them transducers were sucking or blowing. I believe, it'd be one sucking and three blowing, to use your recent metaphor :rotfl:

And as always, "Cheers" to you too - but I've got three-hours and a commute to wait through first! See you on the other side...

:cheers:

[John Nebel]

Giskard,

In your last post you said:

"The RED terminal on the loudspeaker enclosure goes through the passive network to the RED terminal on the LF and to the BLACK terminals on the MF, HF, and UHF."

This seems correct (!!) and agrees
with the way I read your 5/31 post.

This is different from Tom's post today:

"FYI JBL's spec:
2231 green to red terminal
2121 blk/wht to red terminal
2420 red/wht to red terminal
2405 gry/wht to red terminal"

Because the blk/wht, red/wht, and gry/wht are connected to the red cabinet terminal.

Or I'm really missing something.

John

Tom's post might make sense if someone is biamping and wants a positive cabinet, but that would depend on the phase shift of the LC network which was removed by the biamp switch and I'm seriously regretting not having paid more attention in freshman physics.

[boputnam]

Or I'm really missing something Thanks, John - me too. I felt I'm the only one in the wilderness out here.

Tom's starting post is inconsistent with the schematic, and contrary to what Giskard finally posted.

My head is spinning... :spin:

But to clarify, I'm not proselytizing any one phasing over another (never meant to). I'm just trying to sort-out what JBL's original phasing was, by transducer, and it looks like:

LF (-)
MF (+)
HF (+)
UHF (+)

based upon Giskard's last post:

The RED terminal on the loudspeaker enclosure goes through the passive network to the RED terminal on the LF and to the BLACK terminals on the MF, HF, and UHF.
All the involved transducers are negative, the LF Red terminal is seeing the Red Input (through the network) and so is running (-), while the MF, HF and UHF Red Terminals are all seeing Black (through the network) and so are running (+).

But, I fear we (me...) are unintentionally pissing off a good friend, and wish it wasn't so! :(

[4313B]

I'm not sure why it is different from Tom's post.

The following wires are "hooked" up to the RED input of the enclosure:
Green - which goes to the RED terminal on the 2231
Black/White - which goes to the BLACK terminal on the 2121
Red/White - which goes to the BLACK terminal on the 2420
Grey/White - which goes to the BLACK terminal on the 2405

The following wires are "hooked" up to the BLACK input of the enclosure:
Black - which goes to the BLACK terminal on the 2231
White - which goes to the RED terminal on the 2121
Red - which goes to the RED terminal on the 2420
Red - which goes to the RED terminal on the 2405

By convention, anything with a stripe in it usually gets hooked up to the BLACK terminal of the transducer. JBL has used a ton of different wire colors over the years. The "best" convention before the Ti series was green & green/black for LF, white & white/black for MF, yellow and yellow/black for HF and orange & orange/black for UHF.

*****

Ok, re-reading Tom's post I see the error. Tom, what was the message I sent to you? What did it say? I know what I meant but it may not have been what I said. Wasn't the 2121 out of phase or something? I purged my mailboxes so I can't go back and see.

"Because the blk/wht, red/wht, and gry/wht are connected to the red cabinet terminal."

Yes.

"Or I'm really missing something."

No. I probably am at this point....

"This seems correct (!!) and agrees with the way I read your 5/31 post."

Well that's a plus.

[Tom Loizeaux]

I read all this response to my original post - and got more confused.

I wired my 2231s so they move out with + voltage going to the red input terminal. (Though I bi-amp these 4343s, I don't see why this would make any difference.)
I wired my 2121s to be in phase with the 2231s, which put the white wire to the speaker's red terminal. Is this a mistake? (I ran the + voltage to the 2121s through the crossover to determine cone direction)
I wired the 2420s with the red wire going to the red speaker terminal.
I wired the 2405s with the red wire going to the red speaker terminal.
Remember that I'm trying to wire these to all "play forward" or be "positive".

I may still not have these correct for a complete "In Phase" system, so let's not drop this until we figure this out.

Thanks,

Tom

[boputnam]

Thanks, Tom...

Whew - this has been one struggle today. Too bad you were in surgery all day ;)

To start, for all not yet in therapy - there is no intrinsic need to be one polarity or another. It only matters when you are pairing cabinets that each same-frequency transducer be same-phased (all LF (+) or (-), etc.) to avoid frequency cancellation.

So, reading your most recent post, here's what you done:

2231's: (+) voltage to Red terminal = (-), and also factory. That is the GRN wire, correct?
2121's: WHT to Red terminal = (+), and also factory.
2420's: RED to Red terminal = (+), and also factory.
2405's: RED to Red terminal = (+), and also factory.

According to your latest post, you have connected the tranducers to factory spec, as Giskard recommended (and just re-posted ;) ), and your first post on this (07-17-2003 07:36 AM) was somewhat in-error.

However, your LF cone is moving IN on positive voltage, opposite what you posted.
+ voltage going to the red input terminal ...on the 2231H would make the cone move in.

Now, if you are pairing these 4343's with other cabinets, just make sure those other transducers are like-phased. That is, make sure the LF's are all (-), and all other transducers are (+) to compliment what you have in the 4343 configuration.

Thanks greatly for re-posting on this.

Now, to some fine old-vines Zin... Takers? :dancin:

[Tom Loizeaux]

To Bo's statements:
(quote) Your change only impacts the area of crossover (or more correctly "roll-off") overlap between the 2231H and 2121H - that is, you are now getting a doubling of those overlap frequencies, where as-factory would have cancelled those.
Are you saying that the woofer and mid bass are supposed to be out of phase with each other? That doesn't sound right to me.

In answer to the questions from your most recent post:
First, remember I've reversed all these drivers to "playing forward" (positive).

No, I put the green wire to the black speaker terminal on the 2231s.
I put the white wire to the red speaker terminal on the 2121s.
I put the red wire to the red speaker terminal on the 2420s.
I put the red wire to the red speaker terminals on the 2405s.

I realize this is not consistant with the JBL 3134 schematic, but when I wired it to the schematic the music was not coming from between the cabinets the way it should.

My quote of "+ voltage to RED terminal" was refering to the INPUT terminals - on the cabinet - not the speaker terminals. Is any of this helping? I reversed the 2121s, 2420s and the 2405s - and things got better. I don't know what I should try at this point. All I know is that my cabinets sound better then the way they were wired. But I still want to know "the truth!"

Tom

[boputnam]

Are you saying that the woofer and mid bass are supposed to be out of phase with each other? That doesn't sound right to me. Yes, that was factory spec. It may improve the imaging of the 2121H. A "doubling" as you have done is not bad, only not-as-designed.

And to the rest:


No, I put the green wire to the black speaker terminal on the 2231s.
I put the white wire to the red speaker terminal on the 2121s.
I put the red wire to the red speaker terminal on the 2420s.
I put the red wire to the red speaker terminals on the 2405s

...is fine. No problem. The only change you have affected is you are running the 2231H's (+), giving a doubling in the "area" of the crossover between the 2231H and 2121H. If YOU like it, GREAT!! The entire upper end (MF, HF, and UHF) is (+), as is factory, so the imaging in that range is spec.

Is any of this helping? Yes - all this improved description is helping a great deal - many thanks. Many mysteries resolved.

Now this, could be confusing...
I reversed the 2121s, 2420s and the 2405s ...wrt the above quote as to "white and red wire(s) going to red terminals". If the wiring quote is correct - white and red wire(s) going to red terminals - whatever "reversing" you did, corrected the MF, HF and UHF to factory spec.

Now - out of interest to your initial "imaging" frustration - are you pairing these 4343's with anything? If so, what? They may have been out-of-phase with the "factory" wiring for the 2231H.

All I know is that my cabinets sound better then the way they were wired. That is what matters.

But I still want to know "the truth!" The truth is "out there somewhere" and if "factory" truth is what you seek, it is as Giskard and I posted:
LF would be (-)
MF (+)
HF (+)
UHF (+)

To get there, all you need do is put GRN wire to Red terminal of the LF.

[Tom Loizeaux]

This has been a real goose chase and I must apologize for some of this.
Something Bo said got me thinking. He said that the mid bass 2121 should be out of phase with the woofer. I assumed that these two should be in phase, so testing them with a battery seemed logical. Thinking I had this right is what started my confusion. See, a little knowledge can be a dangerous thing! I also colored in the speaker terminals on my copy of the JBL schematic, but then I started to question my accuracy on this.
Anyway, after reading all the above posts, I went down to my studio room this evening and pulled the top panels off my 4343s. I switched the wires to the 2121s, 2420s and 2405s. After re-assembly I found that I still had the sound stage I found recently and, when switched to MONO, had the music coming from between the cabinets, proving the drivers were working in phase! They sound wonderful!
So now I have these wired to match JBL's schematic...only reversed at each driver!

Thanks to all of you for your help and patience. The next time someone asks about wiring a vintage JBL 4 way cabinet, we'll know!

Tom

[4313B]

"The next time someone asks about wiring a vintage JBL 4 way cabinet, we'll know!"

The previous discussion applies to the 4343, 4344, and 4345.

The 4315B is wired as follows:

The following wires are "hooked" up to the RED input of the enclosure:
Green - which goes to the RED terminal on the 2203
White/Black - which goes to the BLACK terminal on the 2108
Yellow - which goes to the RED terminal on the 2105
Orange/Black - which goes to the BLACK terminal on the 2405

The following wires are "hooked" up to the BLACK input of the enclosure:
Green/Black - which goes to the BLACK terminal on the 2203
White - which goes to the RED terminal on the 2108
Yellow/Black - which goes to the BLACK terminal on the 2105
Orange - which goes to the RED terminal on the 2405

Note that JBL did use several previous color code schemes for earlier production runs of the 4315's. The point to note is the phasing as opposed to wire colors.

[boputnam]

switched the wires to the 2121s, 2420s and 2405s. ... by doing this, you put these out-of-phase with the LF (2231H's), as was the design.

Now, your 4343's are reversed from factory, but honor the engineered phasing for the cabinet (only 180-degrees out-of-phase).

Your report confirms that switching the wiring to all the HF transducers as a group keeps them in-phase (as a group) as designed, and that would have no effect on the "recent sound stage" you enjoyed. But by doing that, you re-established the cancellation in the crossover point between the 2231H and the 2121H which might improve the imaging in the ~290 Hz range.

Thanks to you for pushing on through this confusion - it provided a great deal of confimation, and much to ponder! :hmm:

[4313B]

"Now, your 4343's are reversed from factory, but honor the engineered phasing for the cabinet (only 180-degrees out-of-phase). "

Yeah but if he's using 4355's as mains and these 4343's as rears everything on the top end will be in phase ;)

[boputnam]

Yeah but if he's using 4355's as mains and these 4343's as rears everything on the top end will be in phase. THAT is an interesting point, and correct! But the LF's will be regrettably out-of-phase. See? You too can teach a geologist to read network schematics... :banghead:

It seems that this is the point in the discussion that we hear from martinleewin... ;)

My head is :spin: .

[boputnam]

The point to note is the phasing as opposed to wire colors. From the Sage of the Forum.

Perfect.

As always, thanks, Giskard.

[John Nebel]

Thanks to Giskard and Bo for clarifying this.

So recently I learned that surrounds rot, and indeed mine were. Spiders fatigue, and mine likely were. Things inscrutably change with frequency, phases shift, sometimes cancel as the factory intended, but maybe that is not what you wanted. Cabinets are negative, the room is full of standing waves, and 1 watt
goes a very long way.

Maybe computer nerds like myself should stay away from audio, it's too complex.

John

[boputnam]

Maybe computer nerds like myself should stay away from audio, it's too complex. :rotfl:
No WAY!! We NEED you guys to provide us with these Forums and to help us build avatars!

[Tom Loizeaux]

One last question concerning crossovers and my 4343s:
I am bi-amping these 4343s and using a 24dB/octave Rane active crossover, and setting it at the suggested 300Hz. When I adjust the delay on the crossover, I find that max. level at the crossover point requires some delay. This delays the woofer to be in phase with the rest of the drivers, though I'm only going to hear the relationship of the 2231 with the 2121s. In this "delayed" setting the low end is slightly stronger. Is this correct, or am I incorrectly excentuating the crossover point? Is "no dely" what JBL intended (having these 15s and 10s out of phasewith each other) even when using these 4343s in bi-amp mode?
What do you think? Bo, Giskard, ...anyone?

Thanks,

Tom

[Robh3606]

Hmm

My question would be if the 5234 or other JBL electronic crossovers designed to be used with the 4343 inverted phase on high pass output??? Do those crossovers maintain phase coherence at both high and lowpass outputs???

Tom can you change phase on the Rane?? On the M552/553 you can invert phase on High Pass outputs.

Rob:smthsail:

[Tom Loizeaux]

[QUOTE]Originally posted by Robh3606
[B] outputs???
Tom can you change phase on the Rane??

Robh,
The Rane lets you dial in delay on the low pass section to bring the low and high signals in to phase.
If I was using a 12dB/ octave crossover, which would have a 180 degree shift, then would the 2231 being wired "out of phase" with the 2121 make sense? But since I'm using a 24dB/ octave crossover, which has a 360 degree shift, shouldn't these drivers be wired "in phase"?
Just looking for an informed answer.

Thanks,

Tom

[Robh3606]

Sorry but the best I can do is uninformed!

Simply the crossover cards are either 12db or 18db on the 5234 so I am not sure what the crossover does?? Does it maintain the proper phase relation ships between high and low pass or invert like you would a passive crossover so you would have the same phase relationship between drivers as the passive crossover??

I use a 24dB L/R set-up to. I have my drivers in phase. I know the crossover does not invert phase from high to low pass. It has the option to do so if required. My crossover does not have a delay option so I can't fine tune or compensate for driver distances IE horn loading of the bass or compression driver. Theoretically without path differences you would be one cycle out of phase at the crossover point.

I am a babe in the woods with active crossovers very much on the learning curve, hope I am correct hope this helps

Rob :smthsail:

[Robh3606]

Hey Rob

Is there a 1 cycle delay with no path diferences??

Rob:smthsail:

[Tom Loizeaux]

I hope these will be my final words on wiring my 4343s.
I realize, now, what the schematic calls for and believe JBL was correct with their color coding.
I also feel the JBL intended that these studio monitors be used either in the INTERNAL mode (4-way passive crossover) or EXTERNAL mode (bi-amped) with the cabinet being driven by outputs from two amps being fed by an active crossover at 300Hz, in the case of the 4343, and that a 12dB/octave be used. JBL probably hoped that their 5234 or 5235, with the appropriate cards be used, or that another crossover brand be used - having a 12dB/octave slope.
It seems to me that correct polarity, as set up in the 3143 internal crossover, expects the low pass crossover to be a 2nd order, 12dB/ octave slope with a 180 degree shift. I also suspect that this 3143 crossover uses the physical set-back of the 2420 to acheive phase in this design. By using a different slope on the low pass crossover point you are changing the phase relantionship of the woofer to the rest of the componants, particularly noticeable between the woofer and the mid bass driver!
So, if I use a 24dB/octave slope (360 degree shift), I will not provide the required 180 degree phase shift and, in reality, would be running the woofers 180 degrees out of phase with the rest of the drivers! I can either switch the wires to the 2231, or use my active crossover's delay (phase adjustment) to correct this effective 180 degree disparity!
I think I have this right (finally) and I wonder if this makes sense to those who might know more about this then I.

Looking forward to your comments.

P.S. I truly think I'm done with this issue and promise not to continue to chase this...unless, of course, someone shines a new and compelling light on this issue!

Tom

[Tom Loizeaux]

I hope I didn't scare off any responses to my last post. I AM curious about responses to my idea that the 24dB crossover shifts the woofer polarity from what JBL expected...etc.
My comment about being done whit this was really ment to say that I'm going to try not to generate more questions about this because I feel a little guilty about stiring up some of this confusion.
I am looking for truth and would like to hear from others about whether I've found it on this issue.

Thanks,

Tom

[Robh3606]

I have always been fascinated with the whole phase issue. Between the the different crossovers orders 1st-4rth and types Butterworth L/R and there effects, then there is baffle position and path lengths for horns. Complicated in the sense that there are many things to consider. Especially when you are trying to coble a set of speakers together from scratch. That phase switch is the way to go!!

I am curious about responses to my idea that the 24dB crossover shifts the woofer polarity from what JBL expected...etc.

Makes sense an extra 180 degrees of shift so you have to reverse and there is no reason to have them out of phase unless path length differences cause a null at crossover. I would still like to know if the 5234/35 inverts phase on its outputs. I think it does as it only makes sense to keep the same passive crossover phase relationships.

Rob
:smthsail:

[boputnam]

Tom and Rob...

Just returning from a trip, and trying to catch-up on your discussion. There's a nearly parallel Thread on the Technical Help forum that is about to converge with this one... :eek:

From there, as per usual, Giskard posts the most remindful:
Utilize the phase condition which yields the most natural reproduction. ...but that doesn't answer your question on the active crossover sending out-of-phase signals to the LF and HF (2-way).

From Rob:
On the M552/553 you can invert phase on High Pass outputs. I wonder whether JBL incorporated this option into their later-model active crossovers because of their legacy polarity convention...? :hmm:

Let me see if I can wire-up that Galaxy "Cricket" to answer your question(s). I believe I can insert it pre- active crossover and check for phasing effects (w/ and w/o the actives), but this is will take making some cables and such. I have both the 5234A and the 553, so if I get it done, it might be useful. I regrettably won't get to this until closer to the weekend, and by then you may have formed your own crossover company!

[boputnam]

Rob...


...I am not sure what the (active) crossover does?? Does it maintain the proper phase relation ships between high and low pass or invert like you would a passive crossover so you would have the same phase relationship between drivers as the passive crossover?? I'm not sure this will relieve some of the mystery, but both these schematics are portrayed in 4-way, passive mode. That is, the internal switch converting the cabinet binding posts (inputs) from bi-amp configuration to internal crossover is shown in the the "internal crossover" setting. The 3145 is a bit more clear to see this, and if I'm reading it correctly, the phasing of the cabinet is unchanged whether bi-amped mode (and using the 5234A), or not.

So, in the case of the 5234A active crossover, the 3145 schematic suggests LF and HF outputs are in-phase. This confirms my hunch, and I think I will forego the Cricket test.

[Robh3606]

Hey Bo I was typing this when you were doing yours. You beat me to the draw. I see your point.


"On the M552/553 you can invert phase on High Pass outputs."

I took another look you change phase on all outputs.

"I wonder whether JBL incorporated this option into their later-model active crossovers because of their legacy polarity convention...?"

Could be for the most flexability.

"Let me see if I can wire-up that Galaxy "Cricket" to answer your question(s). I believe I can insert it pre- active crossover and check for phasing effects (w/ and w/o the actives), but this is will take making some cables and such.

I have been thinking about that. Could be I am making this harder than it is. This stuff always mess's me up. :banghead:

When you switch the Biamp switch you keep the same polarity on the woofer so it already is correct. So all you are doing is splitting the signal at the appropriate frequency and slope. So why would the crossover have to invert the phase?? As long as the slopes are the same all should be well.

Rob
:smthsail:

[boputnam]

Hey, Rob...


I took another look you change phase on all outputs (552/553). Interesting, and potentially very useful measure of flexibility!


When you switch the Biamp switch you keep the same polarity on the woofer so it already is correct. ... and as well, the phasing of the MF, HF and UHF are unchanged whether bi-amped/external or 4-way/passive.


So all you are doing is splitting the signal at the appropriate frequency and slope. So why would the crossover have to invert the phase?? As long as the slopes are the same all should be well. Agree. I've not tried different slopes for different crossover points in the frequency range - never had want to do so. I have paid more attention to the potential benefits of "null" or boost points at crossover (vis-a-vis David F's post, 07-18-2003 10:18 PM, Technical Help), but not the use of different slopes.

[John Nebel]

This is somewhat off-topic but touches on the issue of biamping 4343s. Doesn't seem to deserve a new thread.

I put an old Pro Note prepared by Garry Margolis up on

ftp://atlas.csd.net/jbl

I'd forgotten that JBL said the 4343 had the same bandwidth at a 4350.

Their comments on bi-amping are on page 4.

I think the hand-written numbers are the prices for the components of the 4343.

Page 5 had an interesting caveat - clearly JBL thought 4-way design was hard to get right.

John

[Tom Loizeaux]

John, I just tried your "atlas" link and none of the files (pics) would open. ?

Tom

[Earl K]

Hi John ( Tom )

Those JPEGs opened within my browser - from that ftp file server. ( I use MS Explorer - 5.0 , on a Mac that is running system 9.2 )

<. Earl K

[John Nebel]

Tom,

They are standard JPEGS on a bog standard Unix server
and have been tested with

(1) Mac OSX 10.2, Netscape 7
(2) Win2K, Netscape 7
(3) Mac OS9, Netscape 7
(4) Mac OS9, IE 5
(5) WinNT, IE 5

Computer nerds do this out of habit.

In each case one could click on the image file name or icon,
depending on the variant and the page would display. The images are scaled to display normal size on a 1280x1024
monitor.

Which error message do you get and what is the OS/browser combination?

Computer nerds also ask questions like that.

John

[boputnam]

Which error message do you get and what is the OS/browser combination? Here's what my repeat experience has been, but me thinks it is related to a popularity contest!

Oh, and running XP Pro and IE 6.0.2800 sp2.03, but as I say - I think it's traffic...

[boputnam]

Rob and Tom...

Just had a thought, while out walking the dogs this evening...

I don't know where/why you got into that lengthy discussion about active crossovers and polarity - but it occured to me that the color of wire coming out of the passive crossovers and their relations to phasing might have in-part been involved.

My grab is, the networks were not designed to change phasing, per se. Decisions about phasing were part of the cabinet design, and to add to what has been posited here, the networks were merely wired with the appropriate colors and stripes/non-stripes to facilitate factory assembly.

So, with all transducers negative, (+) to all the Red terminals would have all transducers (-). But, as example, if the MF was to run out-of-phase from the LF (for whatever imaging reason), and the LF was running (-), the leads coming to the MF from the network would be installed such that the Red Input on the cabinet would path through to a striped wire for the MF, so that during factory assembly the MF striped lead could be blindly connected to the Black terminal, yielding a (+) phase. This is what the like posts from Giskard were illustrating.


:banghead:

[John Nebel]

Bo,

Interesting. It can't open the directory. MS
doesn't bother with standards.

Does the complete file name work?

ftp://atlas.csd.net/jbl/JBL_pn_1.jpg

They are named 1-5.

John

[John Nebel]

Bo,

Just tested with the latest IE 6 and it works.

I'll e-mail Don and ask him if he wants them for the library. If he doesn't and you or anyone else can't get to them, I can e-mail, total size 1mb or so.

John

[boputnam]

Does the complete file name work?

ftp://atlas.csd.net/jbl/JBL_pn_1.jpg Yea, John - perfect. Thanks very much, and I appreciate the posting of this info. Nice.

:cheers:

Also, I just futzed around with the ftp, and found that if entered with a closing foreslash, e.g.:
ftp://atlas.csd.net/jbl/ MS recognizes it. Bingo. I should have noticed that, but it was late (for me...) last night. :snore:

Got any more goodies? :yes: ??

[John Nebel]

Bo,

A few things and I'll put them up.

Also, Don can take them for the library if he
wants.

John

[Tom Loizeaux]

John,
Today those files opened!
Anything else on thec 4343 or studio monitor crossovers, phasing, etc. would also be appreciated!

Thanks,

Tom

[John Nebel]

Tom,

The 5235 crossover manual has a section on polarity and phasing.

http://www.jblproservice.com/pdf/Vintage%20JBL-UREI%20Electronics/JBL-5235%20manual.pdf

The manual also has the RC values for the various plug-in cards so the 4343 variant may be put together easily enough.

A 5235 should be available for $100-$200 from a 'net search or on Ebay if you ever want to go that route.

John

[boputnam]

Hey, John...

This quote has been sticking with me:
Page 5 had an interesting caveat - clearly JBL thought 4-way design was hard to get right. that is, unless, JBL do it. ;)

JBL (or at least Margolis...) does think 4-way provides better accuracy (from page 1):
"... but because (the 4333A) only has three transducers, it is not quite as accurate as the four-ways."

That has certainly been my experience, too. But I don't think I'd try and design one. :eek:

This Pro-Note has been a real interesting read. Thanks for posting it.

[boputnam]

Hey, Tom and Rob...

Grist for the mill, here, boys! :D I hesitate to even post this, and restart the grind :eek:


I've not tried different slopes for different crossover points in the frequency range - never had want to do so. - remember who wrote those words...? ;)

Well, looking (yet again), at a very clean full-color copy of the 4345 brochure lately provided me, I note the 4345 employs two different slopes, as follows:

The active 5234A with 51-5145 card crosses over at:
290Hz using 18dB slope (versus 12dB for the internal network at same crossover point, see below).

The passive, internal four-way crossover crosses over at:
290Hz using 12dB slope
1.3kHz using 12dB slope
10kHz using 18dB slope

So, what's it all mean? Are you two right in a 18dB slope being out-of-phase from the same signal with a 12dB slope of same polarity? If so...

¿Why would the 4345 change phase of the LF when in passive mode?
¿What is the REAL polarity of the UHF?

:spin:

NB: I reserve the right to delete this post until my time to do so reliquishes by natural causes... :banghead:

[John Nebel]

Bo,

There was probably a lengthy trial and error process getting the placement of drivers on the baffle, the drivers, and crossover design matched up. If one had a lot of time and the right room it could be done, but it would only make sense for the cost of that process to be spread over many speaker systems.

The more I read these threads, the more I appreciate the difficulty of getting it right.

JBL did seem pretty definite that biamping 4343s and presumably the 4345s which came later was a good idea.

I've read articles to the effect that one should not try biamping at home, but the JBL situation appears different in that it
was designed that way.

There was an interview with Peter Walker of Quad an aeon ago in Audio Amateur in which I believe he counseled against biamping, but that was due to amplifier power issues which have likely been transcended by the immense power for handling transients in modern amplifiers.

John

[boputnam]

"Don't let the math get in the way"

Got it! ;)

But "The UHF slope is actually 24 dB/octave" has me puzzled. Simply for clarity sake, howzit? Is my nifty brochure in-error, or is there something here I'm ignorant of...? :hmm:

[boputnam]

Giskard...

Great post - thanks.

Gimme more... :yes: ?? I'm guessing the impedence of the transducer (2405) contributes to the increased rolloff? Close?

I am repeatedly stunned by the imaging and "accuracy" of the 4345 - I had expected much greater evidence of crossover points in the listening. Not so. The 4345 design overwhelms, and has ruint my sleep habits. As well, the TV simply ain't getting used...

But, you've been notably quiet on the "timing vs phasing" questions raised here. Let me phrase this clearly as I can:

¿Does the effective 24dB rolloff on the 2405 change the "effective" phasing so that as-heard the 2405 is actually out-of-phase with the 2421B (versus in-phase as per the schematic)?

I know, "don't let the maths...", but still curious ;)

[GordonW]

Originally posted by boputnam
Giskard...

Great post - thanks.

Gimme more... :yes: ?? I'm guessing the impedence of the transducer (2405) contributes to the increased rolloff? Close?



Sort of.

What's actually going on, is that if you look at a graph of the frequency response of a 2405, you'll find out that it starts rolling off at about 6000-7000 Hz or so. The initial part of the rolloff (say, between 7000-3500 Hz) is about 6 dB/oct... and that, combined with the ELECTRICAL rolloff of the crossover, gives an EFFECTIVE 24 dB/oct. rolloff rate, measuring the output of the speaker PLUS the crossover.

In essence, the speaker is PART of the crossover network. Many great speakers are designed just this way... letting the driver act as part of the crossover, and therefore getting to use a simpler (and usually more transparent, due to less stuff "in the signal path") and less expensive crossover... :coolness:

Regards,,
Gordon.

[boputnam]

Giskard, that is exactly what I guessed, as I considered all the posts and reposts over the past few weeks, and tried to incorporate it all into my understanding. Lucky for me, I worded the query in a sensible manner (and minded my syntax ;) ).

And, GordonW thanks very much - I only lack good quality frequency response plots to visualize what we are discussing - but it makes sense.

[B&KMan]

Hello,
(sorry for my poor english: I'm french canadian)

Well your question is a big 100$ question. And I paid big time and money for the response.

I aquired my 4343 in used and many problem appeard on the start.
third part recone installed, cabinet dammage and wirring inconsistence...
and of course classical position of the dimmer for high drivers.

Well , i sit down depp in my chair and I start a bigest scientifict procedure.
I buy and studdy a reference material mesurement Bruel and Kjaer. Very more expensif of the any system...
Well, I pick-up the diagram of cross-over, read many forum information...
and finnaly I attemp to ask a response.

1--- Confusion of phase and polarity...

Many peoples write the subWoofer is inverted phase and all driver is foward phase.

--- the phase is rotary response in real and imaginary part of wave sound and flutuate depens on hz production. If you put phase meter and generate sinewave with x frequency the y frequency is different reaction phase.
The polarity is the way of flux electrical. If you put a mic with analyser, not significant difference in polarity good and false. but big difference in ear result.

just the Jbl factory have a data for interpreting the phase relation in polarity...
and determine who driver is out of polarity depend phase reaction in X frecquency.

I pick-up a Dual channel analyser Bruel and Kjaer (2032) for analysing signal...
B&K mic is diffuse field type and preamp. the built-in generator on Analyser produce many type of signal: for this test I choose the impulse pulse (very short in ms).

If you applied impulse signal into the driver : Normally ( if the polarity is good) the diagram is go foward (pressure signal) and back little more slowly (after cut signal). (the invert reaction is called rarifaction response)

this test is easy to applied with the big driver : just put small 1.5 v batteryin terminal for very short time and check if the cone is go front or go back ...
(attention the driver is not built for dc current so be-carreful do not test is very hight sensitive driver or you burn driver)
Unfortunately, For any small driver of compression driver the movement is non visible and impossible to determine the side.


In the laboratory test...

I unbuilt my 4343.
I check the polarity terminal speakers for good corresponding of the wiring diagram of the internal cross-over. And yes the black terminal is the negatif terminal and red terminal is positive terminal. in my version of 4343 (serial # 12654)
This test is easy and any ohm-meter to help to check any wiring confusion :
the negative wiring is direct with no resistance at any driver... Just positive wiring is treated by the cross over... just look the diagram. This rule is universal !!!

But the bigget problem and where the confusion is start, All driver is connected according to positive polarity. No exception !!!

So if you connect the Woofer in battery , the driver is go forward or backward ???.

Ok Ok OK , I continue my experimentation:
I check the polarity in all driver : all driver is pull out of speaker and tested individual : the BIG SURPRISE: all Red terminal Driver is NEGATIF terminal.

YES If you put impulse signal : you discovery the driver positif terminal is BLACK.

So the BLACK terminal of all driver is connected to positif wiring internal cross-over.

So I connected all driver in respect of polarity respons and I rebuild speakers.

Plug and play Speaker and AMAZING More 25% more details fast respond, elimination of box sound, the presence and transient is realy better and more realistic.

So you run in your way. But for me, the mistake of commun sense to match red wire with red terminal driver is start a big deception...

But, because this speaker is pass in many hands and probably reconing, you have nothing idea if the tecnician is inverse the terminal connector for build modern correlation in red and positive convention or just rebuilt in accordance of old JBL chart...

Just scientific test is reveal the real Situation. So in doubth Check around you if guy have instrumation for make test...

Best regards.

[boputnam]

Originally posted by boputnam
So, reading your most recent post, here's what you done:

2231's: (+) voltage to Red terminal = (-), and also factory. That is the GRN wire, correct?
2121's: WHT to Red terminal = (+), and also factory.
2420's: RED to Red terminal = (+), and also factory.
2405's: RED to Red terminal = (+), and also factory.

According to your post, you have connected the tranducers to factory spec...

[B&KMan]

Unfortunately the positof polarity in regard of my analysing internal cross-over and driver respons is

green LW is positif wiring and is connected to black terminal Low driver.
the black wiring cable is connected to red terminal driver...

and

black /wht wiring is positiv wiring and connected to black terminal mid driver.

the red/wht wiring is positiv wiring and connected to black HF terminal

and gry/wht wiring is positiv wiring and connected to black terminal UHF driver.
And this all driver is connected according in correlation in internal wiring polarity cross-over and polarity driver ...

My test is absolutely scientific... but many hand touch this speaker before me and I have not shure is this result is a standart result...

But in this test all positv terminal driver is connected in positive wiring cross-over and th cross over is connected to terminal speakers.

Cheers.

[boputnam]

The only difference seems you're running the LF reversed from JBL design (this assumes the internal cabinet wiring is unmodified...), but that's fine especially since you like the result!! The difference affects only the range of "acoustic overlap" between the 2231 and the 2121.

Thanks for posting all your testing, and welcome to the Forum! :wave:

[Alex Lancaster]

If You REALLY have to check a CD´s polarity, and do not trust Your fingertips, after removing the cover, You may want to do what We did after a long ethanol-based argument:

Get a .001" div dial indicator, use the wide base for it, not the kind of ball point, connect the 1.5V battery, and watch the dial, You have to be intoxicated, otherwise You wouldn´t do it.

[B&KMan]

Thanks for your feeback and your welcome,

well, for the lower driver the JBL spec it is invert polarity in this monitor??

The green, in the electric diagram available on this site web, is positive wire on cross over...

but my Driver 2231 is recone by a cone kit C8R2235 JBL FACTORY KIT

I suppose the original driver : the positive terminal driver is connected to positive wiring cross over...

So in this case the positive terminal is black. (exactly like all other driver)

Do you have any information who JBL explain the invertion of polarity of this driver is the correct set-up ?? The negative terminal driver is connected to the positive wire cross over ???

P.S. if I test the polarity the black is the positive terminal driver like all other driver...

So If you right , JBL plug voluntary the bass driver with invertion of polarity and load the cabinet before ???

[boputnam]

Originally posted by B&KMan
...JBL plug voluntary the bass driver with invertion of polarity and load the cabinet before ??? :yes:

And, here's the schematic - in case you did not already have it.

[John Nebel]

On the 4343 any black wire or wire with a black trace should be connected to the corresponding black speaker terminal.

It is that simple and JBL says that they made it that way to make the system simple to wire on the production line. They were not thinking of people like us, but it helps here too.

John

[boputnam]

Originally posted by John Nebel
On the 4343 any black wire or wire with a black trace should be connected to the corresponding black speaker terminal. Perfect! Thanks, John.

Now, back to your Servers... :shock:

Say - what you going as for Halloween? :hmm: Tell me, and I'll tell you...

[Zilch]

Originally posted by B&KMan
well, for the lower driver the JBL spec it is invert polarity in this monitor?? You are trying to read JBL wiring according to positive phase convention, whereas JBL standard is NEGATIVE phase.

According to the shematic, it is the OTHER three drivers that are inverse wired. The black terminal of the crossover input connects directly to the red terminals of the MF, HF, and UHF drivers, but to the black terminal of the LF driver, by design.

Connect as John just said. Any crossover output wire with black in it connects to the black terminal of the corresponding dirver....

From what you've said, you've got your LF wired wrong. The black wire should go to the black driver terminal, and the green wire to the red driver terminal....

[B&KMan]

Thanks again for reply Mr boputnam.

Well according to your link (same as my reference diagram)

the
green is + wiring for LF
black/white is + wiring for MF
Red/White is + wiring for HF
grey/white is + wiring for UHF

All pure color: black, white, red, red is the negatif wiring
(so if the black is black the red is not red??) :eek:

well according to test of each 8 driver in speakers, all appear to black terminal driver is positive polarity.
In case of Lf the test of dc current is simple and confirm the experience of the rest of polarity test.

my result is inverse of the general common ... :confused:

well your reply supposed you have a offical paper for connection, it is possible to send a pict of this.

Thanks...

P.S. I tried to send pict but I'm not good result (if you have tip for this, I appreciate)

[Zilch]

Originally posted by B&KMan
This test is easy and any ohm-meter to help to check any wiring confusion :
the negative wiring is direct with no resistance at any driver... Just positive wiring is treated by the cross over... just look the diagram. This rule is universal !!!

While it's true it's the red input connections (what you call "positive") that get the "treatment" in the crossover, those are connected to the BLACK terminals of the MF, HF, and UHF drivers, but the RED terminal of the LF driver:

blk/wht connects to the black terminal of the MF (wht to the red terminal)

red/wht connects to the black terminal of the HF (red to the red terminal)

gry/wht connects to the black terminal of the UHF (red to the red terminal)

black connects to the black terminal of the LF (green to the red terminal)

If I'm wrong, tell me soon enough for me to delete this. (And deny ever having posted it, of course.... ) :D

[B&KMan]

Originally posted by Zilch

According to the shematic, it is the OTHER three drivers that are inverse wired. The black terminal of the crossover input connects directly to the red terminals of the MF, HF, and UHF drivers, but to the black terminal of the LF driver, by design.


Thanks, well your response is correspond on my bench test
all 3 drivers (mf, hF,UHF) is connected as your description.

but the LF is little enigme.

actually my LF is go forward if I put + signal on black terminal.
the green is the + signal in 3143, so the driver is go backward volunteer by design JBL. Do you have any doc in this.

2--- if I abstract the color, normally the LF positive terminal it is connected in positive wire ??
In this design speaker, the driver is inverted polarity or my terminal color is mounted in inverse by recone guys??

thanks for your lights

[Zilch]

Originally posted by B&KMan
my result is inverse of the general common ... :confused:
You have "discovered" that JBL speakers (with few exceptions) follow the NEGATIVE polarity convention....

[B&KMan]

Originally posted by Zilch
You have "discovered" that JBL speakers (with few exceptions) follow the NEGATIVE polarity convention....

Wow !!! :jawdrop:


But if any black terminal is positif signal why plug black terminal to ground signal ???
What is interest to load driver in back motion???

This is not sense???

:banghead:

(...)

[Zilch]

Your LF drivers are correct if a positive DC voltage to the RED terminal pulls the cone in. In this system, the MF, HF, and UHF are intentionally wired opposite polarity from the LF, to compensate for phase reversals in the crossover itself, and possibly, the physical configurations of the drivers.

Yes, normally the LF red terminal is connected to the crossover red input terminal. Thus, a positive DC voltage to the red crossover input terminal causes the cone to move inward.

Read up here. (http://www.jblpro.com/pub/technote/tn_v1n12b.pdf)

Recognize also there is no "ground" or "positive" involved here. The red and black are merely arbitrarily assigned phase designations. Speakers are AC devices; there IS no electrical "polarity" here. You can reverse the wiring to a system with impunity, as long as you wire all systems the same so as to keep everything in phase, or out of phase, as designed. That's how this thread started. Tom wanted to reverse the phase of all of his drivers. He could merely have reversed the leads on his input terminals to accomplish that....

Read up here, also. (http://audioheritage.org/vbulletin/showthread.php?s=&threadid=3015&highlight=negative+convention)

[Alex Lancaster]

I guess Mr Lansing is having a lot of fun after 60 years of this.

[B&KMan]

Originally posted by Zilch

Recognize also there is no "ground" or "positive" involved here. The red and black are merely arbitrarily assigned phase designations. Speakers are AC devices; there IS no electrical "polarity" here. You can reverse the wiring to a system with impunity, as long as you wire all systems the same so as to keep everything in phase, or out of phase, as designed. That's how this thread started. Tom wanted to reverse the phase of all of his drivers. He could merely have reversed the leads on his input terminals to accomplish that....


Thanks and I read many link you give.

Yes you are right in all the way...

And my LF ir react in good way polarity + applied in red put motion in...

For the phase (where info JBL who claim this driver is connected is polarity inverse for best match phase ???

Or, if you prefer, what is protocol for phase test who determine what is exact polarity connection is better ???

thanks again.

[B&KMan]

Originally posted by Alex Lancaster
I guess Mr Lansing is having a lot of fun after 60 years of this.

Yes and lot of money in pocket because the responses is only acess by tech dealer and charge for ... :smthsail:

[Zilch]

Originally posted by B&KMan
For the phase (where info JBL who claim this driver is connected is polarity inverse for best match phase ???

Or, if you prefer, what is protocol for phase test who determine what is exact polarity connection is better ??? Now you are inquiring about the subtleties of speaker system design. Crossover components introduce phase shifts of varying degree depending upon the "order" of filter. However, the physical configuration of the drivers also influences the final outcome.

Generally speaking, the effects occur at and about the crossover frequency, attenuating the output at those frequencies at certain listening angles, as was mentioned earlier in this thread. Sometimes the designer chooses the final phase connection to optimize the overall performance in various listening environments. Users can, and sometimes do, the same. It's as much art and preference as it is technology.

Choosing phase connection by subjective evaluation is mentioned in the several references I cited. It's not absolute. Start with what JBL establishes as standard for any given system and tweak to your heart's content. Joy or curse, it comes with the territory here.... :)

[B&KMan]

Originally posted by Zilch

Choosing phase connection by subjective evaluation is mentioned in the several references I cited. It's not absolute. Start with what JBL establishes as standard for any given system and tweak to your heart's content. Joy or curse, it comes with the territory here.... :)

Thanks very well for your light.
Is my turn to give info...

I just finish run test impulse response of the driver with polarity with and without inversion.

My Bruel & kjaer bi-canal analyser is ability to mesure phase response drivers in same time of the spectrum response. I put mic under one meter for limited introduction of noise.

the response result is effectively really better with black wiring is connected to black pole of driver. Yes the polarity is inversed in this (the cone in go inside).

The phase is non discontinued with the MF. plus : the higher spectrum response of LF is really better and more linear. Of course the 1/3 octave analyser have difficult to explain this effect, the FFT analyser where each frequency appear expose more the big difference of two connection and spectrum response.

The response is really change and I have obligation to re-calibration of the pot for other driver because the LF is more power... what big surprise. :dancin:

I appreciate your information: Is confirm exactly the experience but it is not shure before double expertise in two way ( scientifict method).

:cheers:

[boputnam]

Originally posted by Zilch
If I'm wrong, tell me soon enough for me to delete this.
:D Nope, you've got it 100% correct!

To augmnent what "Dr." Nebel posted, (and I hope this is uniform...):

All solid colored wires (non-black) connect to the Red terminals
All striped wires with white or black or solid black connect to Black terminals.

:thmbsup:

[Zilch]

Originally posted by B&KMan
I appreciate your information: Is confirm exactly the experience but it is not shure before double expertise in two way (scientific method).You confirm what is generally accepted as fundamental here: JBL designers and engineers know what they're doing.

It' sometimes difficult to understand just exactly what that is, is all.... ;)

[B&KMan]

I just send a pict to mesure phenomenon with and without good connection
and the result of the phase and spectrum response...

pict is 1000 words value...

:hyp:

[B&KMan]

:smthsail:

same test in all parameter except conection in in good polarity but is in out pf phase design.

:spchless:

[boputnam]

Originally posted by B&KMan
same test in all parameter except conection in in good polarity but is in out pf phase design. Nice work, B&K Man! Thanks for sharing your findings, and for taking the time to collect the data, both right and wrong connections. Interesting...