This is a new thread that picks up on the discussion of active crossovers in the 4345 - 3 way project thread over on the Lansing DIY forums.
https://www.audioheritage.org/vbulle...-with-JBL-2447
https://www.audioheritage.org/vbulle...l=1#post442634
Given this new thread is a technical reference l have posted it in the Technical Help Forum.
Looking over my shoulder.
To the thousands of lurkers watching this thread try and interpret each post from your own perspective.
Everyone has has their own level of hands on experience, insights and accumulated technical understanding when it comes to active crossovers.
Context is a powerful thing in the face of ambiguity and rhetoric.
Putting the above paragraph into context it’s about where you yourself are starting from (1), your goals, your expectations and what information (2) you have available to implement your bi amp or fully active system correctly. (See (1) and (2) is the following posts below for why this is important).
Getting started.
Try and avoid aiming for perfection at the outset.
The best approach is to take it step by step starting off with bi amp mode of operation. This will prepare you and pave the way for tri amp mode or full active operation should you wish to do that. Taking a step by step approach will provide you with experience, new insights and a more realistic view of what you want to achieve with your system.
Everyone is at a unique station in life.
Reading on the forums about how someone was able to set up their bi amp or active system is only really valid from their own perspective.
In this post Rob discussed his experience with active crossovers:
https://www.audioheritage.org/vbulle...l=1#post442639
(1) At the end of this post Rob comments on ease of which you can create filters using DSP is impressive so you can try and make changes almost on the fly. But that is an individual perspective and we all exist in a different station in life.
Rob has years, decades of diy loudspeaker experience behind him and number of important skills he has picked up along the way. So Rob’s comments are valid in his situation.
Yes you can make changes quickly with a DSP active crossover but unless you have an in-depth understanding of crossover networks and how dsp active crossovers work you most likely will be shooting into the darkness of the black hole.
If your curious with what this is all about and new to the scene it’s not going to a walk up start. I have personally met someone who spent a fortune on state of art jbl drivers and attempted an active system with a DEQX active crossover believing it would be easy based on DEQX marketing at the time.
DEQX active crossovers have employed an auto wizzard that promise according to DEQX marketing to work the magic for you. Please do not believe the hype.
This person ended up selling everything because he gave up trying to make it work.
(2) Obtaining complete and accurate information to make your bi amp or fully active JBL system work the way it should.
Let’s first look at the pre determined JBL passive loudspeaker systems
Unfortunately grabbing complete and accurate information to bi amp or convert a passive system to full active mode is not as straightforward as we would like.
If your starting from scratch looking at a graph and some technical numbers such as shown in the attachments this is not enough for you to grasp exactly how to realise the true potential of a bi amp mode or a fully active system.
JBL contemporary passive systems
JBL do provide some guidelines in the owners manuals on connections to the binding posts for bi amp mode but the technical information for the crossover characteristics is sketchy. Unless the user has a detailed familiarity with loudspeaker crossover technical “jargon” the information provided is like reading a foreign language. Refer to the attachments.
The work around
In the absence of complete information on the crossover characteristics the work around is to obtain detailed frequency response measurements of the passive system and that of the individual drivers. This can be done with REW and some technical guidance. REW have a very good online Q&A and there are tutorials on the www. These measurements are the templates for obtaining the optimal crossover characteristics with an active crossover. In some cases the acoustic response measurements will closely match a 4th order Linkwitz Riley crossover characteristic. But this is not always the case.
Be aware however that the crossover characteristic that you program in the DSP active crossover will most certainly NOT be a perfect 4th order Linkwitz Riley crossover curve. This is due to the fact that most loudspeaker drivers do not have flat text book response curves in the crossover region. Many people have not understood this. You may also need to account for some equalisation incorporated in the passive crossover characteristics. A DSP active crossover may have PEQ filters to obtain the correct acoustic crossover curves.
As a recommendation shifting the crossover point +- 1/3 octave is one approach to obtaining a reasonably close match to your acoustic driver measurements with the passive crossover. If this does not seem to work on the horn high pass curve try a 2nd order Linkwitz crossover character and repeat adjustment of the crossover frequency. Failing this you may need to raise or lower the filter Q of each filter section. This will create a peaking effect at the crossover point.
If you are able to match the passive acoustic crossover characteristics it is not necessary to apply a dsp delay in an attempt to time align the woofer and the horn compression drivers. This is because the crossover frequencies and the acoustic response in the crossover region accounts for the driver phase alignment.
A test to confirm or validate the crossover characteristics is to reverse the electrical connections of the mid or high frequency driver. Then measure the system response in the crossover region. A uniform null of 20 db should occur at the acoustic crossover frequency. Re adjust the crossover points of this is not the case.
If you want to try a dsp time alignment first obtain a very close match to a 4th order Linkwitz crossover characteristic for the woofer and the horn at identical crossover frequencies. Then with the horn compression driver wired out of phase take a frequency response measurement. Adjust the delay on the woofer until a uniform null of 20 db is obtained. Then wire the compression driver in phase with the woofer. Some adjustment of the crossover frequency +- 1/3 octave maybe required to optimise the subjective performance.
Analogue active crossovers
If this is something you don’t feel confident in doing consider using a custom analogue active crossover. Consult with the manufacturer to arrange the correct crossover characteristics. Some measurements of the passive crossover maybe required.
Ian,
This could be a cool thread. I like and agree with your suggested approach of doing things in a very methodical, systematic way. Baseline measurements are super important.
I think there are lots of us on the LH forums that have pursued a bi-amp setup with a two-output (high-pass and low-pass) external crossover, but probably fewer who have tried a full active tri-amp or quad-amp setup with the big JBL 3-way/4-way systems. This is something I've been wanting to try for a long time, and it might be fun to chronicle it here on the LH forums (if anyone is still around)!
I Matt,Originally Posted by mattking52
Thanks for your comments,
A thread without something visual is a bit bland. I plan to put together a case study that illustrates the key points l made in my lengthy posts.
One thing l missed was that an active crossover is not like buying or using a preamp.
A preamplifier selects from a known source then amplifies the signal with a defined gain. There is only one control and that’s the volume knob.
You don’t need to program it or find out and interpret information on how to program it. Then figure out how to adjust it.
Having analysed the process of using an active crossover there are two roads:
1. You need to have a step by step approach from the very outset and be prepared to do some
research and work.
If that is not something you are interested in refer to 2 below.
2. You purchase an active crossover that can be programmed for pre determined loudspeakers that
have facilities for bi amp mode of operation.
About 15 years ago l got involved in active crossover and built a diy clone of the Pass Labs XVR1.
What l soon discovered and then validated with an LHS forum member in CA is that most pro and consumer active crossover significantly impact on the transparency of the original passive system. You get firmer bass but the rest sucks. This is more noticeable with good amplifiers like Bryston, Macintosh, Pass Labs and other high fidelity power amplifiers and preamps.
My other earlier post on the myths of active crossovers is to highlight what you can and can’t change.
The signal processing capability of a dsp crossover might look impressive and make you think it will help you make a perfect loudspeaker and is easy to set up. Neither claim is true.
As many have discovered if you buy a dsp crossover on the basis of those two claims it’s a road to hell.
This is illustrated by the number of posts and detailed explanations by Pos and others on how to use a dsp crossover. Secondly people realise it takes an eternity of adjustment and Re adjustments before they start to feel remotely confident they have got it right.
It ends up being a very bad customer experience. But humans do like to believe in miracles once they are sold the notion of a miracle.
The forthcoming case study will reveal why option 2 with a high quality design is the pragmatic option.
The whole idea behind sound reproduction is to listen to the music.
Not the system because you have no idea if you have got it right.
Case study
Bi amp of the JBL Pro 4343.
System sensitivity 93-94 db 1 watt one metre
Power rating 75 watts continuous
This particular system is now part of JBL’s journey in multi way studio monitors.
Back then these systems were empirically developed using trial and error measurements. If you can find it JBL comments in an article that the 4343 system exhibited a 2 db bump in its bass mid crossover region due to mutual coupling of the 2231A bass driver and the 2121A mid cone driver. JBL determined the crossover frequencies in bi amp mode using the JBL 5234 active frequency dividing network should be spread from a 230 hertz low pass filter and 280 hertz high pass filter.
Nice work Ian.
Possibly how a rank amateur would configure a system, as improperly as possible. Even leaving the excess gain amp argument alone the system described is:
Every output single ended.
Every input differential drive.
All the gain at the end of the signal chain.
Every rule of signal chain gain stage broken.
Poor SNR will undoubtedly be but one of the results.
What would be the results if we fixed all the above?
Barry.
If we knew what the hell we were doing, we wouldn't call it research would we.
-96.6 db SNR at the 1.4 volt signal level. Add +16 db of noise at the listening levels l referred to above. That assumes blameless equivalent.Nice work Ian.
Possibly how a rank amateur would configure a system, as improperly as possible. Even leaving the excess gain amp argument alone the system described is:
Every output single ended.
Every input differential drive.
All the gain at the end of the signal chain.
Every rule of signal chain gain stage broken.
Poor SNR will undoubtedly be but one of the results.
What would be the results if we fixed all the above?
Barry.
There are ways of improving on the noise level up to 20 db when bi amping at real (normal levels). This will feature in my forthcoming analogue active crossover using some elaborate technology.
Edit. It took a long time find and understand the problem. Fixing the problem was about technology and good design. I mean why come out with something what isn’t a real improvement on the current state?
The other way around this is it sit 8-10 metres back from your system and just use more power.
But that’s not doing to work for most users in a domestic situation.
I've actually been running my 4343s quad-amped via DSP for the last few months. I'll see if I can write up a thread/guide on my approach to it over the weekend. Quite the learning curve, but very rewarding when you finally get everything in sync.I can attest to everything Ian is talking about. You have to be very methodical and take a step-by-step approach to implementing it.
There are currently 1 users browsing this thread. (0 members and 1 guests)
Posting Permissions
Reply With Quote