Hi I have a pair of passive PMC FB1+ transmission line speakers and am thinking of convertering them to actives using amplifiers i have and building a linkwitz riley 2 way crossover from elliot sounds. But I am not sure if it is straight forward as this to get decent flat response and sound that is better than the original. Would I need some sort of eq circuit too for the speakers???.
Has anyone had any exerience converting existing passive speakers to actives with any success.
Cheers
Steevo
Has anyone had any exerience converting existing passive speakers to actives with any success.
Cheers
Steevo
It would be easy to make some equalizing in a active filter, but to you that is a question this means your not into creating your own circuit.
But no worry here many use mini-dsp. With basic knowledge of filter and speakers you can create your best solution with the program.
Mini dsp film
http://www.youtube.com/watch?v=6uptQHSU9TY
And a good amp with four outputs.
http://cgi.ebay.com/4-100W-4O-Tripath-TK2050-D-class-Audio-Amplifier-Board-/370293320597
Last edited:
You are right it is not that easy, unless you have good data of the driver responses and know what you are doing, then it is unlikely you will get a result that is equal to let alone better than the dedicated crossover.
A classic mistake a lot of people make is to just take the published crossover frequency and make that in an active version.
A classic mistake a lot of people make is to just take the published crossover frequency and make that in an active version.
I do have a UV box and etching tank so i can etch my own PCBs and have done so in the past but not for years, I have tended to build and buy circuit boards from the likes of Signal transfer designed by doug self as these amps give fantastic peformance, although i have started to design my own high performance feed forward current dumping amplifier, this will take me a while to do. I find that PCB layout has been difficult in the past to give the best SNR performance although i was never able to really measure properly so that is partly why i have opted to buy other peoples designs and circuit boards. I have only recently been able to measure amplifier performance after buying the excellent TRUE RTA sprectrum analyzser from TRUE AUDIO.
Anyway yes i think without knowing the driver characterisitcs it is going to be diffucult to come up with something better than the passive version.
The bass unit is a VIFA driver but the tweeter is bespoke soft dome PMC
unit i think, so getting the data sheet for the tweeter my prove difficult.
I did read an article about the possibility of sweeping throught frequency rage
and measuring the voltage at the passive crossover outputs and then
reproducing the response using active network, but i am not sure if this would
be successfull. Any comment on this idea??.
The origianl FB1 could be made into active speaker using a Bryston power amplifier pack, not sure if this included the EQ and active crossover circuits, it
must have done, but they removed the active option with the FB1+ and FB1i version, so i am not sure how to approch it now.
cheers
steevo,
Assuming you want to create the same transfer function as the passive filters, you could take a differential measurement from crossover input to each of the driver terminals and save those as "target" filter responses. You can then try to realize/match those with some sort of active design (either analog or digital.)
No acoustic measurements are required.
Cheers,
Dave.
Assuming you want to create the same transfer function as the passive filters, you could take a differential measurement from crossover input to each of the driver terminals and save those as "target" filter responses. You can then try to realize/match those with some sort of active design (either analog or digital.)
No acoustic measurements are required.
Cheers,
Dave.
I did think of that but I think the enclosure is tuned to the response of the drivers when combined with passive crossover, when using a active crossover
the transient response of the drivers will be improved so their response will
change drastically. It will sound a bit crap i think. The only option is to use a
mic to measure the response and correct it with eq. What do you think.
Cheers
Steevo
Not sure if it is pertinent to this model but PMC used to offer active and what they called 'activated' speakers.
The 'activated' speakers got a single Bryston amp in form of a so called power pac. This did not include a line level crossover and they thus retained the passive one.
No, I don't agree with that. If you duplicate the passive crossover electrical transfer function then the tonal balance, summation of woofer/tweeter, low-end response (box tuning), etc, etc, will all be unchanged. I've performed this type of passive/active conversion (using LspCAD) on a number of speakers and it will work.....IF duplication is the objective.
Of course the question all this raises is, if it's even worth duplicating the original passive crossover function when a possibly better result could be achieved starting with a clean sheet of paper and implementing with an active crossover.
Cheers,
Dave.
Sorry for the small OT...
That seems like a GREAT amp choice. Is there a link to builds for it? I have zero skills, zero knowledge on amp building but I have alway wanted to build one. I would need instructions (dummy version if possible). That looks like an incredible start. EDIT: I now remember these on Parts express and there was a video. I also remember some discussions on them having SQ issues...hmmm. Still would love to build one and find, out...power supply,board are no PE (Buy from them instead of ebay!!!) Just need to find a quality case, proper RCA connectors??
On topic, I would suggest measuring the drivers with HOLM and build new XOs for it. The MiniDSP is a great device (I own 3 now and Im buying a 4th).
Last edited:
Looking at their website they do say the following,
4. What is the difference between passive and active speakers?
Passive speakers have a non-powered crossover between the power amp and the drive units, whereas an active speaker has a powered crossover between the preamp and power-amp. This means that a single stereo amplifier can control a two or three way speaker pair, while an active design requires one channel of amplification per drive unit. So a two way active design would require 2 stereo amplifiers and a three way active design would require 3 stereo amplifiers to power the speaker pair. PMC active designs are given the ‘A” suffix in the name e.g. BB5-A.
Activated speakers use passive crossovers, but have a separate amplifier for each speaker. This eliminates cross-talk between the channels giving a clearer sound. The DB1S-A and TB2S-A are both activated designs.
So activated speakers are not active speakers, they still use passive crossovers with 2 amps and 2 seperate passive crossovers. The bottom of their true active range is the AML2 which is about £5000
4. What is the difference between passive and active speakers?
Passive speakers have a non-powered crossover between the power amp and the drive units, whereas an active speaker has a powered crossover between the preamp and power-amp. This means that a single stereo amplifier can control a two or three way speaker pair, while an active design requires one channel of amplification per drive unit. So a two way active design would require 2 stereo amplifiers and a three way active design would require 3 stereo amplifiers to power the speaker pair. PMC active designs are given the ‘A” suffix in the name e.g. BB5-A.
Activated speakers use passive crossovers, but have a separate amplifier for each speaker. This eliminates cross-talk between the channels giving a clearer sound. The DB1S-A and TB2S-A are both activated designs.
So activated speakers are not active speakers, they still use passive crossovers with 2 amps and 2 seperate passive crossovers. The bottom of their true active range is the AML2 which is about £5000
No, I don't agree with that. If you duplicate the passive crossover electrical transfer function then the tonal balance, summation of woofer/tweeter, low-end response (box tuning), etc, etc, will all be unchanged. I've performed this type of passive/active conversion (using LspCAD) on a number of speakers and it will work.....IF duplication is the objective.
Of course the question all this raises is, if it's even worth duplicating the original passive crossover function when a possibly better result could be achieved starting with a clean sheet of paper and implementing with an active crossover.
Cheers,
Dave.
I might try that option replicating the passive crossover, it should imrpove the speakers performance, faster transient response, lower distortion and cross talk. How did your speakers perform when you carried out this conversion. Was it a significant improvment and did you consider it worth while??.
It's a noticeable improvement in performance, but not night and day.
If part of your objective is learning about crossover design and become familiar with active and passive networks then it's worth doing. If you're just interested in trying to improve the performance of your speakers and don't give a hoot about the technical aspects then keep the status-quo and enjoy.
Or, as I said, you could start from scratch and design an active crossover network with a clean sheet of paper.
Cheers,
Dave.
Learning about active crossover design is partly my objective so I think it is worth while doing. I plan to buy doug selfs new book on active crossover design when it comes out. Can you recomend any good books.
Wouldnt starting from scratch and designing a active crossover involve designing new speakers as well??. If it does then this is a bigger undertaking and i am not quite ready for that yet.
Starting a crossover design from scratch is exactly the same as designing a new speaker minus the box tuning.
You will need to measure accurately. It's certainly not impossible, but there is a learning curve. You'll invest a bit in a calibrated microphone and preamp, but once you do it once you'll be doing it over and over again.
Like many projects the hardest part is defining what you want. Here the hard part is determining the transfer functions you need, and that requires accurate measurements and a bit of understanding what they mean.
Designing filters to meet a transfer function is relatively straight forward. Once you get transfer functions that is reasonably flat on and off axis with good phase transitions, then comes the art part, tweaking slightly to maximize performance.
See Linkwitz Lab - Loudspeaker Design for an active crossover design tutorial. He doesn't call it that, but he walks you through the design process. It's fairly deep, but after a few times through it the math will begin to make sense. He offers formulae for each of the filter building blocks. He also offers boards on which to build crossovers with many of the blocks he describes.
You're at the tipping point. There isn't much difference between trying to replicate the existing transfer functions and finding ones that work on your own. Dive in, it's fun.
Edit: You might try simplifying the process a bit by sticking with on axis measurements for your first attempt or two. As you gain confidence in your ability to manipulate the on axis response then look at off axis. The goal is to avoid sharp changes in off axis response, and sometimes doing so requires that you give up some on axis flatness. There are snap apart sticks of sockets that you can use to try different component values without having to re-solder if you were to use a board like Linkwitz', in the same way that D. Self is proposing DIP sockets for prototyping resistor values on his boards.
You will need to measure accurately. It's certainly not impossible, but there is a learning curve. You'll invest a bit in a calibrated microphone and preamp, but once you do it once you'll be doing it over and over again.
Like many projects the hardest part is defining what you want. Here the hard part is determining the transfer functions you need, and that requires accurate measurements and a bit of understanding what they mean.
Designing filters to meet a transfer function is relatively straight forward. Once you get transfer functions that is reasonably flat on and off axis with good phase transitions, then comes the art part, tweaking slightly to maximize performance.
See Linkwitz Lab - Loudspeaker Design for an active crossover design tutorial. He doesn't call it that, but he walks you through the design process. It's fairly deep, but after a few times through it the math will begin to make sense. He offers formulae for each of the filter building blocks. He also offers boards on which to build crossovers with many of the blocks he describes.
You're at the tipping point. There isn't much difference between trying to replicate the existing transfer functions and finding ones that work on your own. Dive in, it's fun.
Edit: You might try simplifying the process a bit by sticking with on axis measurements for your first attempt or two. As you gain confidence in your ability to manipulate the on axis response then look at off axis. The goal is to avoid sharp changes in off axis response, and sometimes doing so requires that you give up some on axis flatness. There are snap apart sticks of sockets that you can use to try different component values without having to re-solder if you were to use a board like Linkwitz', in the same way that D. Self is proposing DIP sockets for prototyping resistor values on his boards.
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.