Hello Everybody
I have been reading tons of stuff about active speakers as I want to build two three-way boxes and one of the questions I have not been able to solve so far is how to effectively protect the costly tweeters from DC-voltages in case of problems.The amps for both midrange and tweeters deliver more than enought output voltage to never go into clipping and are working well below their power ratings so no problems to be expected there.
Everything I have been able to find and read points to no circuit truly being able to protect directly connected tweeters against DC. be it because of time-constant or whatever else any existing circuit is too slow or has other drawbacks.
So my question is what would happen if I simply connect a large coil of say 5 mH across the tweeter and an admittedly slow fuse between the tweeter and the output of the amp which is driven by an active filter of 24 dB and a crossover frequency of 3 KHz.. There should be very little signal ,if any , getting there below 500 HZ so the extra load on the amp caused by this should not be any problem ?
The tweeter-amp runs on the full supply of the bass-amp but needs a fraction of the power so one could use a 4 ohm resistor in series with the tweeter to limit the current required of the amp throught the coil as frequency decreases and adjust the gain accordingly to compensate the lesser volume of the tweeter bringing it back to the correct level. Of course there will be some power lost in the series resistor but I think acceptable if the tweeter is effectively protected.
I would be grateful for any info / explanation as to why or why not.
I have been reading tons of stuff about active speakers as I want to build two three-way boxes and one of the questions I have not been able to solve so far is how to effectively protect the costly tweeters from DC-voltages in case of problems.The amps for both midrange and tweeters deliver more than enought output voltage to never go into clipping and are working well below their power ratings so no problems to be expected there.
Everything I have been able to find and read points to no circuit truly being able to protect directly connected tweeters against DC. be it because of time-constant or whatever else any existing circuit is too slow or has other drawbacks.
So my question is what would happen if I simply connect a large coil of say 5 mH across the tweeter and an admittedly slow fuse between the tweeter and the output of the amp which is driven by an active filter of 24 dB and a crossover frequency of 3 KHz.. There should be very little signal ,if any , getting there below 500 HZ so the extra load on the amp caused by this should not be any problem ?
The tweeter-amp runs on the full supply of the bass-amp but needs a fraction of the power so one could use a 4 ohm resistor in series with the tweeter to limit the current required of the amp throught the coil as frequency decreases and adjust the gain accordingly to compensate the lesser volume of the tweeter bringing it back to the correct level. Of course there will be some power lost in the series resistor but I think acceptable if the tweeter is effectively protected.
I would be grateful for any info / explanation as to why or why not.
Last edited:
Why not a big C in series with the tweeter? Seems a lot safer to me and the highpass can be accounted for in the active X-over.
An ideal capacitors DC resistance is infinite, an ideal inductors DC resistance is zero.
Just pick a capacitor chosen so it does not interfere with your chosen xover frequency and put it in line with your tweeter.
Just pick a capacitor chosen so it does not interfere with your chosen xover frequency and put it in line with your tweeter.
Chosen xover is 3000 hz so for 4 ohm tweeter what would that have to be as it should not be electrolytic? What about phase-delay ?
As others have already said, simply put a capacitor in series with the tweeter. Something bigger that 33uF won't interfere with your desired crossover point, for a smaller value you need to account its transfer function into the active crossover. If you want no change whatsoever in the phase you need probably a 100uF cap.
Ralf
Ralf
Your original idea is creative, but also a common way of doing things (lesser known in regards to speakers). I have no data though, you'd have to make your own.
You could use a small value of capacitor. No need to make it large to hide it from the filter band of frequencies.
It would be part of the filter you are creating that has that phase variation anyway.
If you do not want to use one capacitor, like reasonable people would, you can do what bose people do, use incandescent bulb in series.
The shunt certainly works. As does the power dissipated in the resistor. And the (little) change in the transfer function. So I'd stick to the cap. Cheaper, power efficient and more reliable (think of a loose lead in the shunt...).
Hello everybody , I like the approach with the capacitor . A 100uf 100V foil capacitor can be had for about 12 euros or 15 $ which is not a lot compared to the price of good tweeters and will definitely block any DC from reaching the tweeter.
If this does not introduce any phaseshift at the used band of frequencies with a xover of 3000Hz and being a foil capacitor it should be absolutely neutral at any frequencies up from there. The construction inside it avoids it being a coil internally so if interested look at the spec-sheet (Pdf) on Mousers page.
Modifying the the 24dB Linkwitz-Riley-filters which are already built is not an option for me.
R60EW61005000K KEMET | Mouser Espana
If this does not introduce any phaseshift at the used band of frequencies with a xover of 3000Hz and being a foil capacitor it should be absolutely neutral at any frequencies up from there. The construction inside it avoids it being a coil internally so if interested look at the spec-sheet (Pdf) on Mousers page.
Modifying the the 24dB Linkwitz-Riley-filters which are already built is not an option for me.
R60EW61005000K KEMET | Mouser Espana
I am here to learn. Would you trust that tweeter is protected if classAB amp would fail and 59 volts would be dumped on what you posted?
Yes, most of the current would go through the coil in parallel, but still, it is fully protected?
If the coil is 0.3 ohm, and tweeter is 4 ohm, hmmm, you get almost one tenth of the current still going through the tweeter.
Cap is better protection for sure. But my impression from first post was, thread starter did not want to use cap. Now he does. Ok.
I know not just bose used small bulbs in series with the tweeter, and even mids. The resistance the the filament increases with current/heat going through the filament. In a sense its a compressor. Limiting the peaks. Plus it glows, so you get cheap indication that you are running it too hard. Once it goes over certain threshold, filament burns, it acks like fuse, and tweeter is saved. So it protects from excessive ac and well as dc. If one insists that no cap, than my suggestion would work.
I would not trust your suggestion, but its ok, as long as you trust it. I always run tweeters with caps.
Crowbar protection has been around longer than I have. MalleMike observed it correctly because he included a fuse. I'm not taking sides here, but this is a very basic electronics techique.
Attachments
Hello everybody ,
My goal is simply a failsafe! protection for expensive tweeters.
Crowbar protection is not failsafe on one side, it can be triggered erroneously under certain circumstances and it is very destructive if that happens.mMechanical relays are far too slow, triacs can be triggered by fast rising voltage etc. etc. Time constants have to be short to effectively protect a tweeter and long to avoid false triggering . I do not like these compromises at all.
My initial idea was a large coil of say 10 mH shunting the tweeter but after reading and considering all that you have written and contributed i am leaning more to the capacitor provided I can exclude any phase shift introduced by it being in series with the tweeter.
my active filters are Linkwitz-Riley following a circuit published by Elliot Sound Products but modified somewhat see link:
"Linkwitz-Riley Electronic Crossover"
The cross over frequencies are as published though because they happen to go well with the speakers I am going to use.
The amps on rails of plus/minus 50 volts stabilised with 10,000uF decoupling before stabilizing and 4700uF at each amps PCB so the amps can deliver a very powerful pulse of DC at 50 volts if something fails...........and even on power on or off , worst case being one of the rail voltages failing or short in output
My goal is simply a failsafe! protection for expensive tweeters.
Crowbar protection is not failsafe on one side, it can be triggered erroneously under certain circumstances and it is very destructive if that happens.mMechanical relays are far too slow, triacs can be triggered by fast rising voltage etc. etc. Time constants have to be short to effectively protect a tweeter and long to avoid false triggering . I do not like these compromises at all.
My initial idea was a large coil of say 10 mH shunting the tweeter but after reading and considering all that you have written and contributed i am leaning more to the capacitor provided I can exclude any phase shift introduced by it being in series with the tweeter.
my active filters are Linkwitz-Riley following a circuit published by Elliot Sound Products but modified somewhat see link:
"Linkwitz-Riley Electronic Crossover"
The cross over frequencies are as published though because they happen to go well with the speakers I am going to use.
The amps on rails of plus/minus 50 volts stabilised with 10,000uF decoupling before stabilizing and 4700uF at each amps PCB so the amps can deliver a very powerful pulse of DC at 50 volts if something fails...........and even on power on or off , worst case being one of the rail voltages failing or short in output
The triacs would not be the triggering device in the case of a speaker.MalleMike said:
- I was only trying to open eyes and allay fears to keep this thread on an even keel 😉 All that needs to be said has been said.
^ i agree bulbs are pretty good as tweeter protection ( learned it when a pair of Jbl control one i used 'flashed' after a stupid things i did 🙂 ).
Why not a cap then a bulb ( or the inverse)?
For the cap not to influence phase you need to make them two decade lower that your target xover freq ( one decade should be ok with minimal artefacts and more doable in practice if you want to stay away from electrolytics).
Why not a cap then a bulb ( or the inverse)?
For the cap not to influence phase you need to make them two decade lower that your target xover freq ( one decade should be ok with minimal artefacts and more doable in practice if you want to stay away from electrolytics).
I vote for a simple series capacitor. I put such caps in my active speakers too because an extra failsafe never hurts.
Instead of paying extra $$$ for huge fancy capacitors a just as good and much cheaper option is to use a normal 100 uF electolytic in parallel with a smaller, say 1 uF higher quality cap. MKP is great here. This is one of the tricks that Nelson Pass uses in his capacitor output coupled amplifiers.
Instead of paying extra $$$ for huge fancy capacitors a just as good and much cheaper option is to use a normal 100 uF electolytic in parallel with a smaller, say 1 uF higher quality cap. MKP is great here. This is one of the tricks that Nelson Pass uses in his capacitor output coupled amplifiers.
Yes I want to stay away from electrolytics !
Considering my use of active filters 24dB octave , xover of 3000Hz , the signal to the tweeter is down by 40 dB at 1500 Hz which is down to 1% of the signal said tweeter has to reproduce. so what would the value of the series capacitor have to be for this frequency , using 8Ohm tweeters ? Any effect inphase should then be negligible ?