Tweeter: Vifa D25AG
fs: 850 Hz
6 ohms
I've been looking into using a cap, but a value of 31 uf is needed to protect at fs, or 33 uf at 800 Hz which is actually below resonance but satisfies the 1.5 octaves away from the 2.4kHz xo point.
Is it acceptable to put caps in series to make up the value?
I can only find electrolytics of this value. Even if I could find polyester, they seem to be too expensive. I'd spend more on tweeter protection than on the tweeter itself.
I have considered using an electrolytic and just replacing it periodically as its value drops.
I would also like to use an active loudspeaker protection cirtuit, but I also want passive as a last line of defence.
The amp I plan to use for the tweeter will be a chip amp, which apparently has its own protection built in which has ugly performance when it starts to work. However, this chip amp has about 60w whereas I only need about 20w at most - hence it will NEVER be driven hard.
fs: 850 Hz
6 ohms
I've been looking into using a cap, but a value of 31 uf is needed to protect at fs, or 33 uf at 800 Hz which is actually below resonance but satisfies the 1.5 octaves away from the 2.4kHz xo point.
Is it acceptable to put caps in series to make up the value?
I can only find electrolytics of this value. Even if I could find polyester, they seem to be too expensive. I'd spend more on tweeter protection than on the tweeter itself.
I have considered using an electrolytic and just replacing it periodically as its value drops.
I would also like to use an active loudspeaker protection cirtuit, but I also want passive as a last line of defence.
The amp I plan to use for the tweeter will be a chip amp, which apparently has its own protection built in which has ugly performance when it starts to work. However, this chip amp has about 60w whereas I only need about 20w at most - hence it will NEVER be driven hard.
It's the switch-on glitches you need to worry about, not overdriving them/protection ccts kicking in. Monitor the output of your amp with a scope and see what comes out at switch on. If you have access to a chart recorder type device that would be the best thing to use.
I didn't risk it with my active system that used a chipamp to drive a tweeter; I went for a relay.
One easy way is to use a small AC relay driven from the transformer to switch in an RC network across the DC rails. This network drives a transistor that drives a relay that switches the speaker. Or see Doug Self's site for a way that uses a feed from the AC with just one relay but an extra few components.
I didn't risk it with my active system that used a chipamp to drive a tweeter; I went for a relay.
One easy way is to use a small AC relay driven from the transformer to switch in an RC network across the DC rails. This network drives a transistor that drives a relay that switches the speaker. Or see Doug Self's site for a way that uses a feed from the AC with just one relay but an extra few components.
I can get up to 10uf with good quality caps. They cost about the same as my tweeter though to get 3 of them 
I don't have access to a scope, will a software one do?
What is a chart recorder?
How about this for tweeter protection:
www.altronics.com.au/index.asp?area=item&id=K5052
I don't have access to a scope, will a software one do?
What is a chart recorder?
How about this for tweeter protection:
www.altronics.com.au/index.asp?area=item&id=K5052
An externally hosted image should be here but it was not working when we last tested it.
I'm tempted to take my chances with an electrolytic, see how it sounds, but also use the protection circuit. If I'm really unhappy with the electrolytic I could consider taking it out or more expensive caps ...
It does seem a bit of a compromise where I have to put it though. I have to go just slightly below fs to get 1.5 octaves below the xo point of 2.4k ... that's 33uf cap (800Hz)
fs is 850
Thoughts?
It does seem a bit of a compromise where I have to put it though. I have to go just slightly below fs to get 1.5 octaves below the xo point of 2.4k ... that's 33uf cap (800Hz)
fs is 850
Thoughts?
Hi Paul,
These guys http://www.speakerbits.com/net/catalogs/cat19.aspx (in Melbourne) have solen 33uF caps for about $30.... probably about the same as what you were looking at for 3 10Uf caps
I've been looking for caps today too.... have been playing around trying to find the best resistor/cap combo for a zobel (somewhere around 30uF is looking about right for me too .... I think I will go buy a bunch of BP electros in various values and use them for experimenting, then when I'm happy with a value, I'll consider forking out for the more expensive polys.....
If you think standard polys are expensive just have a look at the hovland music caps on that site!!!! I think I'll stick to the solens or possibly even the generic ones from WES. BTW WES have solen caps up to 15uF and they aren't as expensive as the site I posted above (whether they are the same series or not, I don't know).
Dare I suggest that you consider my preferred form of speaker protection the polyswitch??? I know a lot of people think badly of them, but I would think they have less of an effect (if any) than a reactive component like a cap.... My current amp (soon to be replaced/complimented) has relay protection (dc sensing) and also a speaker turn on delay. The relays don't connect the speakers until about 2 or three seconds after turn on, stops the clicks and thumps nicely, also disconnects the speakers the instant the power is switched off..... probably something similar to the Altronics kit you linked to, except it was integrated into the amp (Playmaster series 200).
Tony.
These guys http://www.speakerbits.com/net/catalogs/cat19.aspx (in Melbourne) have solen 33uF caps for about $30.... probably about the same as what you were looking at for 3 10Uf caps
I've been looking for caps today too.... have been playing around trying to find the best resistor/cap combo for a zobel (somewhere around 30uF is looking about right for me too
.... I think I will go buy a bunch of BP electros in various values and use them for experimenting, then when I'm happy with a value, I'll consider forking out for the more expensive polys..... If you think standard polys are expensive just have a look at the hovland music caps on that site!!!! I think I'll stick to the solens or possibly even the generic ones from WES. BTW WES have solen caps up to 15uF and they aren't as expensive as the site I posted above (whether they are the same series or not, I don't know).
Dare I suggest that you consider my preferred form of speaker protection the polyswitch??? I know a lot of people think badly of them, but I would think they have less of an effect (if any) than a reactive component like a cap.... My current amp (soon to be replaced/complimented) has relay protection (dc sensing) and also a speaker turn on delay. The relays don't connect the speakers until about 2 or three seconds after turn on, stops the clicks and thumps nicely, also disconnects the speakers the instant the power is switched off..... probably something similar to the Altronics kit you linked to, except it was integrated into the amp (Playmaster series 200).
Tony.
Thanks for the tips Tony. I have actually taken polyswitches out of my passive xo's as I considered them unecessary!
Different story now though ...
How do you work out the values? What exactly are polyswitches? Aren't they resistors that act at a certain temperature or current? Do they protect against DC?
Different story now though ...
How do you work out the values? What exactly are polyswitches? Aren't they resistors that act at a certain temperature or current? Do they protect against DC?
As UrSv said, caps in parallel add the value. If you can only afford electros though, you need to use polarised ones of double the value, connected back to back, i.e. -ve to -ve (connecting in caps series is like paralleling resistors). You could also bypass each cap or the pair with a better quality one of smaller value.
A chart recorder is a pen that draws along and plots voltage against time. Think of those earthquake recorders in films A computer based scope would be OK. The trouble with using a scope is that you may miss the transient or it will be very, very difficult to judge its spectral content.
That module looks OK but it's about 5x the cost and complexity as what I described.
NFB = negative feedback.
A chart recorder is a pen that draws along and plots voltage against time. Think of those earthquake recorders in films
A computer based scope would be OK. The trouble with using a scope is that you may miss the transient or it will be very, very difficult to judge its spectral content.That module looks OK but it's about 5x the cost and complexity as what I described.
NFB = negative feedback.
Hi Paul,
polyswitches are made from a conductive polimer which has very low resistance. When the current goes higher than their trip rating they melt and have a very high resistance wich is maintained until the current drops, at wich time they resolidify and work as normal again.....
I don't know about the protecting against dc, My guess would be only if the dc had enough current to trip the polyswitch....
working out the values is kinda tricky... I just read the info on WES catalog page for them and my head now hurts basically you need to work out what the max current is going to be for your rated power (and impedance of the driver) and choose a polyswitch that trips at that current or less. Interestingly the really low trip value ones have quite high resistances (something the jaycar catalog doesn't show from memory). .2A trip has resistance of 2.5 Ohms!! only once you get to a trip current of 2.2 amps is the resistance actually what I would call low, at .15 ohms....
Tony.
polyswitches are made from a conductive polimer which has very low resistance. When the current goes higher than their trip rating they melt and have a very high resistance wich is maintained until the current drops, at wich time they resolidify and work as normal again.....
I don't know about the protecting against dc, My guess would be only if the dc had enough current to trip the polyswitch....
working out the values is kinda tricky... I just read the info on WES catalog page for them and my head now hurts
basically you need to work out what the max current is going to be for your rated power (and impedance of the driver) and choose a polyswitch that trips at that current or less. Interestingly the really low trip value ones have quite high resistances (something the jaycar catalog doesn't show from memory). .2A trip has resistance of 2.5 Ohms!! only once you get to a trip current of 2.2 amps is the resistance actually what I would call low, at .15 ohms....Tony.
Konnichiwa,
Well, there are a few issues here.
1) If you are sure you want to add a cap, why not design it in as part of the crossover? Just adjust the overall response to match the desired driver & X-over response (probably best simple solution in terms of implementation). It will protect against DC and "thumps", but not against overload...
2) You could use a simple relais across the tweeter (assuming your Amp is short circuit proof) which requires an external voltage to be opened, this voltage can be generated based on the presence of a signal not exceeding "safe limits" of voltage, free from DC and finally requiring an external "unmute" trigger.
Ideally such circuitry would however exist within the Amplifier itself and should not be needed externally. Suitable "speaker protection & delay" kit's are widely available and only need to be changed to replace the "series" mode of the relais with "shunt" (usually means changing the contact assignments).
It will protect against DC and "thumps", but not against overload...
3) Another option often found in Pro-Audio stuff is a incandescent lamp of suitable voltage and power rating (24V/12W seemed to be common IIRC) in series with the tweeter. This pretty reliably protects against most failure modes but introduces some compression due to the lamps PTC Termistor behaviour.
Ideally one would design the amplifier specifically for the driver in question, include a circuit that protects the Amp against clipping (reduce level automatically on clipping - this is usually best done for the whole active speaker - if one channel clips drop the level 3db permanently untill reset) and includes a circuit accounting of the RMS power applied to the Driver and limits the RMS power as well as accounting for excursion and limits the power with frequency to safe values. DC protection and switch on delay are considered included too.
In the 80's and 90's so-called "Processor PA" systems first featured such technology extensively. They allowed you to drive your PA-Speakers to the ragged edge but any mishap (singer drops the Mike, someone accidentally knocks the master volume sliders.... etc) at that edge would not blow your expensive drivers out.
Such systems can be operated "looped", meaning you sense directly at the drivers terminals or open loop meaning you work out the clipping input voltage for the Amps (worst case), the relation between input voltage and output power of the Amp and also between the Amp's input voltage vs. excursion and provide suitable limiters.
If you have something like the Behring DCX Digital Crossover or the like you usually have limiters and dynamic EQ's that can perform most if not all of these functions and then add amplifiers that have an output protetcion system that elminates thumps and protects against DC in case of failure and all your drivers should be safe.
In the 80's I designed such an open-loop system into our active crossovers (for really big PA systems), using LED/Photoresistor style limiters and discrete peak & RMS rectifiers etc... Took ages to work out but paid in less than 1/2 year for the work in drivers not blown up....
Sayonara
paulspencer said:
Well, there are a few issues here.
1) If you are sure you want to add a cap, why not design it in as part of the crossover? Just adjust the overall response to match the desired driver & X-over response (probably best simple solution in terms of implementation). It will protect against DC and "thumps", but not against overload...
2) You could use a simple relais across the tweeter (assuming your Amp is short circuit proof) which requires an external voltage to be opened, this voltage can be generated based on the presence of a signal not exceeding "safe limits" of voltage, free from DC and finally requiring an external "unmute" trigger.
Ideally such circuitry would however exist within the Amplifier itself and should not be needed externally. Suitable "speaker protection & delay" kit's are widely available and only need to be changed to replace the "series" mode of the relais with "shunt" (usually means changing the contact assignments).
It will protect against DC and "thumps", but not against overload...
3) Another option often found in Pro-Audio stuff is a incandescent lamp of suitable voltage and power rating (24V/12W seemed to be common IIRC) in series with the tweeter. This pretty reliably protects against most failure modes but introduces some compression due to the lamps PTC Termistor behaviour.
Ideally one would design the amplifier specifically for the driver in question, include a circuit that protects the Amp against clipping (reduce level automatically on clipping - this is usually best done for the whole active speaker - if one channel clips drop the level 3db permanently untill reset) and includes a circuit accounting of the RMS power applied to the Driver and limits the RMS power as well as accounting for excursion and limits the power with frequency to safe values. DC protection and switch on delay are considered included too.
In the 80's and 90's so-called "Processor PA" systems first featured such technology extensively. They allowed you to drive your PA-Speakers to the ragged edge but any mishap (singer drops the Mike, someone accidentally knocks the master volume sliders.... etc) at that edge would not blow your expensive drivers out.
Such systems can be operated "looped", meaning you sense directly at the drivers terminals or open loop meaning you work out the clipping input voltage for the Amps (worst case), the relation between input voltage and output power of the Amp and also between the Amp's input voltage vs. excursion and provide suitable limiters.
If you have something like the Behring DCX Digital Crossover or the like you usually have limiters and dynamic EQ's that can perform most if not all of these functions and then add amplifiers that have an output protetcion system that elminates thumps and protects against DC in case of failure and all your drivers should be safe.
In the 80's I designed such an open-loop system into our active crossovers (for really big PA systems), using LED/Photoresistor style limiters and discrete peak & RMS rectifiers etc... Took ages to work out but paid in less than 1/2 year for the work in drivers not blown up....
Sayonara
The caps would be in parallel, but yeah you could do that. i dunno about $30 in caps on a $50 tweeter though.
Chip amps generally have pretty low dc offset, so if you left it on all the time, and never had blackouts etc you could be right. ie plug in tweeters after turning everything on.
pretty risky, but hey...
alternatively try a bipolar electro from jaycar vs nothing ( having checked offset ) and see if you can hear a difference
Chip amps generally have pretty low dc offset, so if you left it on all the time, and never had blackouts etc you could be right. ie plug in tweeters after turning everything on.
pretty risky, but hey...
alternatively try a bipolar electro from jaycar vs nothing ( having checked offset ) and see if you can hear a difference
Sounds like the polys will only work for overdrive, which is not what I'm concerned about. The filters are 4LR at 2.4k. I'm only concerned about protecting the tweeters against the kind of thing that will get past the polys most likely.
I already have the active xo - 4LR - it was a kit. My first venture into active xos. Is it simple to modify it to work as you have suggested? What type of 3rd order would I then use? Would it no longer be effectively 4LR combined with the passive pole?
Sounds like the speaker protection circuit I linked earlier, does that look like it's what you are talking about? The kit is very cheap, probably cheaper than just buying the parts, and a PCB and instructions are included.
Since the amp is diy, it won't have the relay delay unless I add it with this kit, which I intend to do. I'm told the chip amp has extensive protection built in to the chip, and apparently when pushed, it gets ugly due to the heavy duty protection. This is why I want to use it for a tweeter - the amp will never be driven hard.
With about 60w and only driving a tweeter, clipping should not be an issue. Even when paired with the Europower for the subs to 80 Hz and an Alesis 250w for 80Hz - 2.4KHz, it should be idling to drive the tweeters.
I now have Ultracurve DEQ, and have experimented with the limiters and dynamic eq - the problem is it tends to crackle when either is in operation. I had initially thought I could use the dynamic limiters to design the system to avoid possible overexcursion for the subs, but I haven't dared push either the amp to the point of clipping or the subs close to max excursion! With the rumble filter my sims say I should not be able to get more than 16mm one way excursion (well short of 23mm xmax or 30mm xsus). The biggest limit on my system is the amp driving the main speakers with only 40w (active xo not implemented right now).
I already have the active xo - 4LR - it was a kit. My first venture into active xos. Is it simple to modify it to work as you have suggested? What type of 3rd order would I then use? Would it no longer be effectively 4LR combined with the passive pole?
Sounds like the speaker protection circuit I linked earlier, does that look like it's what you are talking about? The kit is very cheap, probably cheaper than just buying the parts, and a PCB and instructions are included.
Since the amp is diy, it won't have the relay delay unless I add it with this kit, which I intend to do. I'm told the chip amp has extensive protection built in to the chip, and apparently when pushed, it gets ugly due to the heavy duty protection. This is why I want to use it for a tweeter - the amp will never be driven hard.
With about 60w and only driving a tweeter, clipping should not be an issue. Even when paired with the Europower for the subs to 80 Hz and an Alesis 250w for 80Hz - 2.4KHz, it should be idling to drive the tweeters.
I now have Ultracurve DEQ, and have experimented with the limiters and dynamic eq - the problem is it tends to crackle when either is in operation. I had initially thought I could use the dynamic limiters to design the system to avoid possible overexcursion for the subs, but I haven't dared push either the amp to the point of clipping or the subs close to max excursion! With the rumble filter my sims say I should not be able to get more than 16mm one way excursion (well short of 23mm xmax or 30mm xsus). The biggest limit on my system is the amp driving the main speakers with only 40w (active xo not implemented right now).
Konnichiwa,
If all you are concerend about are switch-on and off transients, add a suitable speaker protection/delay module from a Kit to you chipamp, done.
Questions, questions. First, textbook LR4 filters do really work with real speakers (I know you want to use a DEQ to fix the fallout, I wouldn't).
Secondly, you would need to look exactly as what kind of response the tweeter cap produces and then simply design a 3rd order filter that adds the neccesary complementary curve so the net afterwards matches your target curve (LR4 FWIW).
Yes, i would modify the circuit to increase it's sensitivity to DC and to have a higher cutoff frequency in the DC Detection, so it will be quicker to trip matching the behaviour of the tweeter....
Then you have set something wrong.
Sayonara
paulspencer said:
If all you are concerend about are switch-on and off transients, add a suitable speaker protection/delay module from a Kit to you chipamp, done.
paulspencer said:
Questions, questions. First, textbook LR4 filters do really work with real speakers (I know you want to use a DEQ to fix the fallout, I wouldn't).
Secondly, you would need to look exactly as what kind of response the tweeter cap produces and then simply design a 3rd order filter that adds the neccesary complementary curve so the net afterwards matches your target curve (LR4 FWIW).
paulspencer said:
Yes, i would modify the circuit to increase it's sensitivity to DC and to have a higher cutoff frequency in the DC Detection, so it will be quicker to trip matching the behaviour of the tweeter....
paulspencer said:
Then you have set something wrong.
Sayonara
PPTC - sound?
I found this article about the usage of a PPTC (Polymeric positive temperature coefficient thermistor) for speaker protection.
I want to confirm - from people that actually used this - if it have any effect on the x-over and sound? Also, exactly how do you calculate the right value to use?
PPTC
I found this article about the usage of a PPTC (Polymeric positive temperature coefficient thermistor) for speaker protection.
I want to confirm - from people that actually used this - if it have any effect on the x-over and sound? Also, exactly how do you calculate the right value to use?
PPTC
Hi JDeV,
the link that you posted is broken... I've used the polyswitches on my 3 ways (admittedly not high end) and couldn't hear any difference. However they do have finite resistance (which can be quite high for the lower current ones) so you probably do need to take that into consideration.
If you know the actual rms wattage of the driver you want to protect, you should be able to calculate the right value polyswitch. you can do it by trial and error. for instance if you though a 2 amp polyswitch was a possibility and you have 8 ohm nominal speakers then the voltage would be 16V V=IR and the power would be 32W P=VI if this is less than your driver power rating try the next value up, if more try the next value down probably better to go on the lower side in any case.
Tony.
the link that you posted is broken... I've used the polyswitches on my 3 ways (admittedly not high end) and couldn't hear any difference. However they do have finite resistance (which can be quite high for the lower current ones) so you probably do need to take that into consideration.
If you know the actual rms wattage of the driver you want to protect, you should be able to calculate the right value polyswitch. you can do it by trial and error. for instance if you though a 2 amp polyswitch was a possibility and you have 8 ohm nominal speakers then the voltage would be 16V V=IR and the power would be 32W P=VI if this is less than your driver power rating try the next value up, if more try the next value down
probably better to go on the lower side in any case.Tony.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.