Building the 4345/4344

[Robh3606]

Hello George

"Then, once that answer is in hand, how do you determine the number and value of resistors to satisfy a given requirement? Let's say the desired power handling capacity is 60 watts, and, as an example, let's take the 15 ohm resistor in Giskard's HF network in post 170. Using 10 watt resistors, I would need six 90 ohm resistors in parallel, correct?"

Yes

"Secondly, I have been researching numerous threads on the forum to learn more about what capacitors to get. What a blackhole topic that is !!! Along the way I have read quite a bit about charge coupled networks being very worth the extra expense. Presuming that would be true in this case, how would these circuits be modified?"

Take a look at the 4343 Crossover Mod Thread. It's simple just double the capacitance and put 2 in series with a 3 Meg Resistor at the junction of the 2 caps.

Hello Darryl

"Rob ,do you know of anyone who would have the drawings for the JBL 4550 BKA cabinet ? Any help would be much appreciated. Thanks... Darryl /Woodstock Sound"

No and I don't see them in the Library. Start a new thread in the Tech Help forum. I think there are people who do have them and this question is going to get lost. Someone will step up if they have them.

Rob

[Ian Mackenzie]

Orange is "old equivalent", green is "new equivalent", white is actual "tapped autotransformer" model.

Orange was done from 1981 Engineering Design Specification voltage drive using datapoints - all L-Pads open - all 8 ohm dummy loads.

There is enough slop in the L-Pads that both are viable - the new model should be cheaper to build with less parts.

Here is a request for another iteration.

Lets assume for a moment most will want to bi amp these and the 4344.

I note the 3145 midrange passband filter has 2.5 R fixed L pad and a variable pad.

The penny drops that the essence of bi amping to gain maximum power and damping on the cone drivers. So why do we need to Pad back the midrange filter at all and in doing so sacrific sensitivity and power amp head room?

(The midrange can still be balanced via the Hf amp level although not independently of the Hf & Uhf. Although I tend to think its much easier to set the levels of the horn and slot against the midrange anyway and then adjust the woofer level)

I would therefore like to propose running the 2122H without a fixed and variable pad and revising with fix pads for the Hf & UHf filters in the "new equivalent filter" by Giskard.

It would be much appreciated for Giskard to review the above and publish a revision when time permits.


Ian

[4313B]

That would be worthwhile.

[Robh3606]

Well just make sure that those resistors don't effect the Q of the bandpass filters. I have run the 2122 strarght up and you get a rather pronounced bump without the filter in place.

Rob

[Ian Mackenzie]

Well just make sure that those resistors don't effect the Q of the bandpass filters. I have run the 2122 strarght up and you get a rather pronounced bump without the filter in place.

Rob

Rob,

I assume you meant the resisters.

Ian

[Robh3606]

"Rob,

I assume you meant the resisters.

Ian"

Yes just see what happens without them may not do a thing but???

Rob

[4313B]

I would therefore like to propose running the 2122H without a fixed and variable pad and revising with fix pads for the Hf & UHf filters in the "new equivalent filter" by Giskard.

It would be much appreciated for Giskard to review the above and publish a revision when time permits.

You can go ahead and pull the 3 dB pad from the 2122H when bi-amping. That will be like doubling the power of your driving amp. The original network was based on the low pass inductor (1.8 mH) having a DCR around 0.6 ohms so try to maintain that. e.g. Using an inductor with 0.2 ohms DCR shifts the crossover down by roughly 50 Hz. A Zobel based on 10 ohms might also be beneficial for tweeking since you would be removing the L-Pad.

Well just make sure that those resistors don't effect the Q of the bandpass filters.

Absolutely. In this particular instance it is evident by inspection that were are looking at a 3 dB pad based on an 8 ohm model. The DCR of the inductor will affect the slope and should be kept around 0.6 ohms. The Solen 16 AWG or 18 AWG would be decent.

[Ian Mackenzie]

Okay,

But what I meant was the fixed pad drops the level 3 db, the L pad is normally also set at - 3db from max for zero position.

This suggests the 2122 is pad back more than -3 net.

If I dropped the fixed pad and turn up the L pad full I wonder if there would be enough level in the HF and uhf filters, even with them up full?

Ah, I think I get it There was a certain amount of real gain in the 2122H bandpass filter and some real loss of gain in the 2245H low pass filter due to choke 5.4 Mh dcr. That net difference could account for the fixed pad attenuation and the -3 db flat position for the mid range L pad. And of course it was voiced and then balanced to sound right..the numbers don't always tell the whole story!

They must have had a hell of a time sorting out those voltage drives.....
G.T did a commendable job calibrating the 4345 L pads.

I will try bypassing the 2122H fixed pad this weekend.

Ian

[Ian Mackenzie]

You can go ahead and pull the 3 dB pad from the 2122H when bi-amping. That will be like doubling the power of your driving amp. The original network was based on the low pass inductor (1.8 mH) having a DCR around 0.6 ohms so try to maintain that. e.g. Using an inductor with 0.2 ohms DCR shifts the crossover down by roughly 50 Hz. A Zobel based on 10 ohms might also be beneficial for tweeking since you would be removing the L-Pad.
Absolutely. In this particular instance it is evident by inspection that were are looking at a 3 dB pad based on an 8 ohm model. The DCR of the inductor will affect the slope and should be kept around 0.6 ohms. The Solen 16 AWG or 18 AWG would be decent.

I am also busy winding coils for the new equivalent design for trial over the Christmas break.

What assumptions/values did you make/use in calculating the fixed pad. trying to work out your values for the horn fixed pad and 2405 ?


Am I correct in assuming you were shooting for a 10-11 ohm impediance on the horn for the fix pad values of 5.6R and 8.6 (15/20)?

Given the 2307/2308 sensitivity of 108 my guess a 10-11 ohm impediance load seen by the crossover, is about -7 db for the fixed pad, about -2 db from the 20R shunted on the horn and about -4 slack on the variable L Pad = 95 db for 0 position.

The 2405 would also appear to be perhaps about 10 ohms impediance hence the 6.8R / 6.8R values which gives about -8 db (105db -8 for the fixed pad -3 db in the variable L pad =95 db for 0 db position.

Ian

[Ian Mackenzie]

Thanks Giskard,

If I recover sufficiently today I will have a crack to firing it up tonight!

(had to polished off most of a bottle of sparkling Shiraz yesterday ..goes well with Turkey...nice)

Ian

[Tom Loizeaux]

I like the way this discussion is going. Designing a crossover for a 4344 (45) would be a very useful exercise!
I agree that if one were building a new pair of these speakers, they'd probably want to run them in bi-amp mode. Given that, the passive crossover would only need to deal with the 10", HF compression driver and 2405 slot. Since the 10" (2122 or 2123) would be the least efficient of these three componants, it should run without any pad (except to deliver the correct slope with the compression driver). Then it may be best to pad the compression driver and slot with stepped attenuators. You wouldn't need more then 6 dB of level adjustment. This would keep variable L Pads out of the crossover and should lead to superior performance.

Tom

[Ian Mackenzie]

Glad someone caught on!

[Tom Loizeaux]

I think there would be great interest, at least here in the Forum, for buying a complete set of detailed plans for building the 4344.
Since a few of us have built, or are building these, it would be smart to have you guys pool your plans, have someone refine, adjust and proof these, and ultimately make them available to prospective 4344 builders. The proceeds of the sale of these could go to support the Forum.
Along with the plans could also be the schematic for the crossover for these "upgraded" bi-amped 4344s. The details of this crossover are currently being worked on (thanks Ian and Giskard).
I, for one, would love to have accurate plans, maybe even including some life-size templates, etc, so when I'm ready to undertake that project, I'd have all the detailed info I'd need.
Though I love my 4343s, I would consider building the 4344s, using the "upgraded" drivers and crossover. I'm sure they would be an improvement, even if only slight, over my vintage 4343s. Having these plans and knowing I could get the correct drivers and crossover parts would make this a very attractive project down the road.

So who's going to step up and agree to coordinate getting these plans and crossover info together?

Tom

[4313B]

I've already done too much for this forum in the past five years. I don't need to step up to anything...

I've gone over this 4344 stuff a few times. Ian has real life hands-on experience and knows what to do.

[Ian Mackenzie]

Look, my simple response is use the search button and keep searching and asking the right questions for about 6 months. Then you will have an inkling of what the product is about......

Building one of these is not something that comes straight off the blue print and I think any notion of "selling" plans is totally off the wall. What would be worth while and is sadly lacking is a product profile like the 4430 in the Library.There's more factual knowledge there than half the crap slapped over the forums by people who are simply clueless and guessing.

The same perspective applies to the 4343-4344-4345. Its very complicated, a large investment, an absolutely huge job and you need a shit load of resources and its a steep learning curve all the way to the end. ie Its not just case of having a router, but how to use one skilfully in new and challenging situations, devising and making up special jigs etc.

A mentor is a must!

I generally find a good deal self satisfaction from first collecting the facts about something I know little or nothing about rather than filing a request for everthing on a silver platter.

That way you take less for granted and have a lot more respect for those who put in the first hard yards.