Jump to content

Re-engineering JBL crossovers to turn PA into HIFI !


taff
 Share

Recommended Posts

So i've bought myself a pair of JBL TR125's for messing about with and learning. The main thing that attracted me to them was the price, and the 99db sensitive 15" woofer. They are in a second system in my workshop, a nice 5.5 x 6 space. I have them basically in the corners and raised so the tweeters are ear height when I'm standing, which is how I spend most of my day!

Having had a look at the crossover:

attachment.php?attachmentid=56118&d=1340480142

A few things jumped out at me.. the first was that the whole tweeter protection circuit was pretty pointless for my use as I wasnt going to be firing hundreds of watts at it to get huge SPL's. So I bypassed the resistor to see what effect it had (I figured it was mainly attenuating the tweeter) This seemed to be the case, subjectively more high frequencies coming through and generally better to listen to.

Then I got inquisitive about the role of the 2 parallell 28uf caps parallelling the woofer. Having read this a few times :

http://lampizator.eu/SPEAKERS/Practical%20tips%20loudspeaker%20DIY.html

I assume they are a low pass filter to protect the woofer from low frequencies. I bypassed one of the caps, and was surprised that there was an (again subjective) increase in detail & volume - especially of voices. Im guessing its also allowed more higher frequencies through to the woofer so there is some reinforcement happening either side of the crosssover point.

Encouraged by these two small changes, I've started playing with some crossover filter calculators all of which give values for a second order crossover that are fairly different to those in the speakers. (both drivers are 8ohm, crossover point is listed as 2200 hz)

Reading around it seems good sounding speaker crossovers are arrived at by a mix of trial and error and calculation, I'm hoping for some input from those of you who have some experience here to add your thoughts or guidance! If not, then it'll be a log of my journey!

PS, yes, I intend to measure them in room soon..

thanks,

James

Link to comment
Share on other sites

The lamps act as a compressor to protect the tweeter from continuous overload rather than peak limiting. The two woofer capacitors are in series, not parallel, the effect is to halve the capacitance and double the voltage handling. By removing one, you have reduced the amount of treble in the bass unit.

If you really want to make these 'speakers work at their best, don't bother with a passive crossover, just go active.

S

Link to comment
Share on other sites

The lamps act as a compressor to protect the tweeter from continuous overload rather than peak limiting. The two woofer capacitors are in series, not parallel, the effect is to halve the capacitance and double the voltage handling. By removing one, you have reduced the amount of treble in the bass unit.

Thanks serge, of course I meant in series. Damn, forgot my basics.. so 2x 33uf = 1x16.5uf.. which is much closer to the 5-10 'ideal' value you get from calculators.

If you really want to make these 'speakers work at their best, don't bother with a passive crossover, just go active.

S

haha should have seen that coming..

Link to comment
Share on other sites

The 2 caps in series are to get the odd value of 14uF. By removing one of the caps you will have raised the knee at the the crossover point and will have more spl in the upper midrange/lower treble.

JBL are a very respectable company that really know what they are doing. You could try improving on the parts that JBL chose but I wouldn't modify the circuit without frequency response and impedance measurements of the individual drivers in the box. Once you've done that you need to import the measurements into some crossover modeling software (I use LspCad).

Ignore Lapizator - Designing a crossover by ear will most likely give you a speaker that will have a wonky frequency response with phase problems. It may sound great on some music but a lot of stuff will be sound bad. Scroll down and you will see what to expect from a Lampizator crossover - LINK

A good speaker designer will adjust by ear but only small amounts and measurements should be taken after to confirm what you hear.

- - - Updated - - -

THIS is worth reading just to give you a basic understanding of what's involved.

Link to comment
Share on other sites

When considering how difficult it is to design a passive crossover, and how difficult to make small adjustments or apply any EQ to correct for inevitable driver shortcomings, or to apply any time-alignment, an active DSP-based approach is just SO much easier and repeatable.

I wouldn't dream of modifying a passive crossover designed by a respectable company like JBL.

S

  • Upvote 1
Link to comment
Share on other sites

Thanks, exactly the stuff I was hoping for. I guessed the lampi stuff is very specific to the green cones, most of his mutterings need to be taken in a wider context I think.. if you take it as gospel it causes trouble!

My initial thoughts were to rebuild with better components, but as the values are away from what the calculators suggest and I'm changing the design criteria I'm wondering if they are optimal for my use and if they can be tweaked to suit hi-fi use more readily. Hopefully I can learn more too.. which I already am!

Sent from my Nexus 6P using Tapatalk

Link to comment
Share on other sites

Looking at the spec, HF is limited to 15kHz @ -10dB, so you might want to restrict the horn to, say, 8kHz and provide an additional tweeter to go above that. The Constant Directivity of the existing horn should be matched with a similar directivity of any such additional tweeter.

As the power handling and sensitivity of the LF unit is good, you might want to use some EQ to extend the LF at the expense of headroom, as you're unlikely to need so much in a 'hifi' context.

S

Link to comment
Share on other sites

First thing to do is ignore online calculators. I've never even looked at one but I assume that they assume that the frequency response and impedance of the driver is perfectly flat and there's no baffle diffraction to worry about. You need measurements of the drivers in the box or you need to trace the data from the manufacturers graphs and model the baffle diffraction in software.

Link to comment
Share on other sites

In early and swapped the now 33uf for a 7.5 I had lying around.. sure enough we have more bass!!

Have just found this very useful in helping my understanding:

image_preview

' In the network pictured in figure 2, the HP blocking capacitor, C1, increases in impedance with decreasing frequency, while at the same time the shorting inductor L1, decreases in impedance with decreasing frequencies, hence shorting out the low frequencies while the capacitor is increasingly blocking them. The inverse is true of the LP filter. L2 is blocking highs and passes lows, while C2 is shorting out the highs and passing the lows. This is how the LP filter blocks highs and passes lows while the HP filter is performing the inverse function. It is this dual action that allows this kind of network to create twice the blocking action (12 db) per octave of frequency than the simpler 6 db/octave network. If you did not understand this on the first read, that is OK, you have a lot of company. Think about how the capacitor and inductor change their impedance with frequency, and take another look at this circuit diagram. If you try, you can figure out what is going on as the frequencies change. '

Link to comment
Share on other sites

Yeah, everything that Fatmarley said is right. Those calculators give you a a good starting point when you start out but the crossovers that JBL have designed will have taken actual electrical and acoustic measurements into account. They know their onions basically. If you can do some real measurements and a bit of modeling go for your life, otherwise assume they have done a pretty good job.

Stefan

Link to comment
Share on other sites

Thanks for the continued input guys. Iteration number 3 now, have paralelled a 2.2uf with the 3uf for the tweeter to reduce the high pass frequency and get more crossover in the 1Kish region. This was successful, but too harsh. I then unsoldered the bypass on the resistor so the tweeter was attenuated as original. Very subjective I know, and I'll measure them soon.. but this is the best iteration so far. My guess is the bass driver doesnt go that low really, so the tweeter is attenuated to bring out more bass for a given volume.

learning.. and a better sounding stereo.. ace!

ps. shocked at the big differences little changes can make. Also the perceived increase in detail can be just a volume increase in those frequency bands. My guess is a lot of work goes into voicing speakers!

Link to comment
Share on other sites

Thanks for the continued input guys. Iteration number 3 now, have paralelled a 2.2uf with the 3uf for the tweeter to reduce the high pass frequency and get more crossover in the 1Kish region. This was successful, but too harsh. I then unsoldered the bypass on the resistor so the tweeter was attenuated as original. Very subjective I know, and I'll measure them soon.. but this is the best iteration so far. My guess is the bass driver doesnt go that low really, so the tweeter is attenuated to bring out more bass for a given volume.

learning.. and a better sounding stereo.. ace!

ps. shocked at the big differences little changes can make. Also the perceived increase in detail can be just a volume increase in those frequency bands. My guess is a lot of work goes into voicing speakers!

Looking at the original circuit diagram, the tweeter is attenuated by the two lamps in series and the unmarked resistor in parallel. Normally, tweeters are more efficient than woofers so need attenuating. The 'normal' amount of attenuation will be whatever the cold resistance of the two lamps is, and the parallel resistor. As the lamps get hot, their resistance increases, so the amount of attenuation increases to protect the tweeter, limited by the parallel resistor. Attenuating the tweeter isn't to bring out more bass, but to reduce the amount of treble, which is not quite the same thing.

The bass driver doesn't go that low, its size is more to increase efficiency and reduce excursion than for depth in bass, that's why I think it can take a fair amount of EQ.

S.

Link to comment
Share on other sites

If it is a CD horn, which I'm not sure it it, It will have compensation for it in the crossover, this is more than just a high pass. Active is by far the best suggestion. When changing a crossover you really have to ask yourself, do I know better than a JBL engineer? I certainly don't, I find it highly doubtful the crossover was compromised in any meaningful way, even by the protection circuit.

Link to comment
Share on other sites

Thanks for the continued input guys. Iteration number 3 now, have paralelled a 2.2uf with the 3uf for the tweeter to reduce the high pass frequency and get more crossover in the 1Kish region. This was successful, but too harsh. I then unsoldered the bypass on the resistor so the tweeter was attenuated as original. Very subjective I know, and I'll measure them soon.. but this is the best iteration so far. My guess is the bass driver doesnt go that low really, so the tweeter is attenuated to bring out more bass for a given volume.

learning.. and a better sounding stereo.. ace!

ps. shocked at the big differences little changes can make. Also the perceived increase in detail can be just a volume increase in those frequency bands. My guess is a lot of work goes into voicing speakers!

That's a 1" compression driver and probably wont be giving it's best below 2.2khz, also you are also dropping below the 2xhorn cut off freq zone too so would be very surprised at any improved sq. The T&S for the bass driver have the fs at 46hz, xmax of 5.1mm, qts 0.42 and vas of 230l. It looks to be a ported cab so go easy with the lf eq below 55hz.

Link to comment
Share on other sites

If it is a CD horn, which I'm not sure it it, It will have compensation for it in the crossover, this is more than just a high pass. Active is by far the best suggestion. When changing a crossover you really have to ask yourself, do I know better than a JBL engineer? I certainly don't, I find it highly doubtful the crossover was compromised in any meaningful way, even by the protection circuit.

The brochure for the loudspeaker describes the horn as Constant Directivity, with a 90 degree 'beamwidth'. A 15" driver will beam considerably long before the likely crossover frequency, so the loudspeaker won't be wide dispersion at anything much above LF. My experience of these types of loudspeaker is that they throw forward and their output drops considerably when out of the 'beam'. They also don't have much at extreme LF as their primary function is to go loud for PA purposes, which they do.

Whether I would ever bother to try and redesign one of these for home HiFi use is doubtful, but as a learning exercise it can be quite fun to do.

S.

Link to comment
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

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

×   Your previous content has been restored.   Clear editor

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

 Share

  • Recently Browsing   0 members

    No registered users viewing this page.

×
×
  • Create New...