PA speakers for brothers band
Posted 30th November 2011 at 10:42 PM by googlyone
My brother was complaining about his vocals not coming through clearly in his (rather modest / amateur) band. It turned out that he was trying to use some old hifi speakers we had as kids!!!
I had a bunch of Beyma 12G125 bass drivers and some P.Audio PA-D26 compression drivers, so offered to throw some proper speakers together for him.
There were a couple of illustrations of design principles in here that are worth reiterating:
- The horn driver has a massive (and by this I mean both acoustically and electrically) resonant peakj just below 2KHz.
To use this driver and avoid horrible the resultant upper midrange peak, it is essental to put a trap on the tweeter to flatten out the impedance.
this at least gives the crossover a hope of working...
- The horn driver still has a big fat peak at 2KHz.
- The horn was CD - and needed compensation to bring the top end up to "flat".
- The bass driver (well at least the ones I had - I suspect that these were prototypes, as I bought them off a South Australian manufacturer of cinema and HIFI speakers who has close links with Beyma) was:
- 4 Ohms (never seen this advertised for this driver)
- More efficient than published
- Measued to have a peak of 6-7dB, about 500Hz wide centred at about 850Hz. this is a particularly painful spot for this sort of misbehaviour - unless you are willing to go 3 way.
- After some angst I decided to put a resonant trap in series witht he woofer to flatten this peak out. Given the impedance I needed to have a DCR way - way below 1 Ohm - these are PA speakers! I ended up with an LCR trap of 1MH, 60uF, 4 Ohms (ish) with the inductor wond from 1.6mm wire, and four chunky 15uF poly caps. The resistor was a 75W WW and went on a heatsink...
So in the end I had a very efficient bass driver, with zobel and a resonant trap, and "fairly flat" response. Also a compression driver with again a complex zobel trap, lightbulb protection (2*18W 12V bulbs) and about 4dB of attentaution.
With everything being nice and contrlled, the crossover "just worked" - with the minor tweak of pushing the compression driver XO frequency up a bit to account for the rising gain of the horn at the low end.
Result?
My brother was blown away. His initial comment was "Hey these are actually HIFI speakers".
If he had seen the frequency response plots he would never have said that!
But the lesson illustrated here (to me) is that if you remove the major flaws in a speaker - in this case two really significant peaks / resonances - and get the balance between drivers right, then it is really - really surprising how good things can sound.
Is the contemplative, "simulation centric" stuff worth the effort? I think it has a place - but in the end of the day speakers are things you listen to. To my mind the models are interesting to explain why something sounds crap, or does not work - but don't get lost in the contemplation.
I had a bunch of Beyma 12G125 bass drivers and some P.Audio PA-D26 compression drivers, so offered to throw some proper speakers together for him.
There were a couple of illustrations of design principles in here that are worth reiterating:
- The horn driver has a massive (and by this I mean both acoustically and electrically) resonant peakj just below 2KHz.
To use this driver and avoid horrible the resultant upper midrange peak, it is essental to put a trap on the tweeter to flatten out the impedance.
this at least gives the crossover a hope of working...
- The horn driver still has a big fat peak at 2KHz.
- The horn was CD - and needed compensation to bring the top end up to "flat".
- The bass driver (well at least the ones I had - I suspect that these were prototypes, as I bought them off a South Australian manufacturer of cinema and HIFI speakers who has close links with Beyma) was:
- 4 Ohms (never seen this advertised for this driver)
- More efficient than published
- Measued to have a peak of 6-7dB, about 500Hz wide centred at about 850Hz. this is a particularly painful spot for this sort of misbehaviour - unless you are willing to go 3 way.
- After some angst I decided to put a resonant trap in series witht he woofer to flatten this peak out. Given the impedance I needed to have a DCR way - way below 1 Ohm - these are PA speakers! I ended up with an LCR trap of 1MH, 60uF, 4 Ohms (ish) with the inductor wond from 1.6mm wire, and four chunky 15uF poly caps. The resistor was a 75W WW and went on a heatsink...
So in the end I had a very efficient bass driver, with zobel and a resonant trap, and "fairly flat" response. Also a compression driver with again a complex zobel trap, lightbulb protection (2*18W 12V bulbs) and about 4dB of attentaution.
With everything being nice and contrlled, the crossover "just worked" - with the minor tweak of pushing the compression driver XO frequency up a bit to account for the rising gain of the horn at the low end.
Result?
My brother was blown away. His initial comment was "Hey these are actually HIFI speakers".
If he had seen the frequency response plots he would never have said that!
But the lesson illustrated here (to me) is that if you remove the major flaws in a speaker - in this case two really significant peaks / resonances - and get the balance between drivers right, then it is really - really surprising how good things can sound.
Is the contemplative, "simulation centric" stuff worth the effort? I think it has a place - but in the end of the day speakers are things you listen to. To my mind the models are interesting to explain why something sounds crap, or does not work - but don't get lost in the contemplation.
Total Comments 0