PDA

View Full Version : HF protection capacitor compensation in active setup?



sebackman
02-06-2011, 02:45 AM
Protection capacitor compensation?
Most of us are using a protection capacitor in front of our tweeters or mid-drivers to protect from amp spikes or other dangerous signals when we use active crossovers.
As a capacitor always will shift the signal 90 degrees there will be a difference in phase between the woofer and the protected mid/high driver.
Now, is that audible or measurable?
The newer digital filters (BSS, JBL, Behringer, DBX) gives the option to apply shift phase for each driver in an active setup so my question is;
Is it any good to apply 90 degrees phase shift on the drivers without the capacitor to restore phase coherence?
Any thoughts?

Best regards
//RoB

JBL 4645
02-06-2011, 03:37 AM
I just use the (audio limiter) on the DCX2496 for the LF/HF LCR fronts.

ratitifb
02-06-2011, 07:00 AM
Protection capacitor is a first order filter "only" compared to the impedance ot the driver and according to JBL litterature "the optimum value of the protection capacitor assumes that the capacitor is active at approximately one octave below the X-over frequency, assuring minimum acoustical interaction with the X-over region performance while maintaining a good degree of protection"

Ruediger
02-06-2011, 07:16 AM
The capacitor is part of a network, and the phase shift is not constant but rather depends on the frequency. The simplest network You can consider for this application is an RC highpass.

Your protection capacitor is the series capacitor C , and Your speaker is the load R. Say You wanted to crossover a midrange unit at 1 kHz (Your active xover does the 1 kHz). You would design the protection capacitor such, that the crossover frequency for the protection network would be 100 Hz or less.

In the frequency range of interest the phase shift would be very small.

Ruediger

Lee in Montreal
02-06-2011, 09:59 AM
My three way Kef 105/2 have four 125F capacitors at the entry of the crossover. That is 500F who's only task is to cut any DC signal, and way out of the audible range.

I am tempted to fit such DC filter on each of my tweeters, mids and woofers, between amps and drivers. Using highly efficient Solen caps will not alter efficiency of drivers.

Mr. Widget
02-06-2011, 10:25 AM
For my TAD TD-4003s with $2300 replacement diaphragms, I have always used protection caps. Amplifiers do die spectacularly at times and people make mistakes rather stupidly at times. Burning up $2300 or $4600 if both channels are affected isn't my idea of fun.

I use moderately priced Metallized Polypropylene 67uF caps on them which is a theoretical 150Hz crossover point on a 16 ohm driver. I have done numerous tests and can hear and measure no change with them in or out of the circuit. The actual active crossover point has been anywhere from 600Hz to 900Hz depending on the system.


Widget

Earl K
02-06-2011, 11:57 AM
,,,,snip,,,,As a capacitor always will shift the signal 90 degrees there will be a difference in phase between the woofer and the protected mid/high driver.
Now, is that audible or measurable?,,,,,

I fear you are labouring under a common misconception about phase shift .

The truth is that the phase shift caused by a single passive device happens in the "stop-band" not the "pass-band" range of frequencies .
- What we listen to is primarily the pass-band ( of frequency content ) .
- The stop-band is where the attenuation caused by the passive device takes place ( a capacitor in your case ) .


If you set the value of the protection cap for a 3 db down point that is 2 to 3 octaves from your active crossover point / the phase-shift contributed by the protection cap will be very minor ( and will eventually be statistically zero as you get further into the pass-band ) .




:)

sebackman
02-08-2011, 08:01 AM
Dear all,

Thank you for valuable input.

It does make sense to ignore phase shifts out side the pass-band. Point taken .

Does the fact that any passive filter component will shift current and voltage 90 degrees make it audible? Is a protection capacitor (good quality) audible in an active setup?

Best regards
//RoB

Earl K
02-08-2011, 08:47 AM
snip , ,,,,,,,, Is a protection capacitor (good quality) audible in an active setup?

Best regards
//RoB

Not in my experience . A blown diaphragm ( caused by not using one if an amp goes DC ) is way more audible .




It does make sense to ignore phase shifts out side the pass-band. Point taken .


- One still needs to pay attention to phase shifts in the stop-band if the passive component adds any significant phase rotation ( say over 30 degrees ) in the pass-band .




Does the fact that any passive filter component will shift current and voltage 90 degrees make it audible?

Again. these shifts are generally confined to the stop-band .

Phase shifts in the pass-band are generally regarded as being inaudible .

Fwiw, inductors and capacitors induce their respective phase shifts in opposite directions of rotation ( this is why at the crossover point or Fc , a 2 pole crossover is 180 out of phase because each pair of passives has moved their induced 90 degree phase shift in opposite directions .


<. :)

jcrobso
02-08-2011, 01:01 PM
DC blocking caps are a MUST if the speaker are hooked directly to the amp!;)

Baron030
02-08-2011, 03:08 PM
For those that want to know the phase angle of their protection cap circuit, I found a free calculator.

Here is the link: http://www.thetaeng.com/calc_RC.htm (http://www.thetaeng.com/calc_RC.htm)

Pictured is an example of own setup. R1: = 2446H driver impedance, C1: = my protection cap value and Frequency is my lower active crossover point. As you can see by using a fairly large cap value, I have minimized the phase shift to only about 14 degrees and that the corner frequency falls fairly close to the recommend value of 2 octaves lower then the active crossover point. While not pictured, at my upper active crossover point, the phase shift drops to about 2 degrees.

Baron 030:)

49815

louped garouv
02-09-2011, 12:54 PM
http://alteclansingunofficial.nlenet.net/publications/techletters/TL_205.pdf

is this chart applicable?

Baron030
02-09-2011, 02:14 PM
courtesy Altec Lansing unofficial page...
http://alteclansingunofficial.nlenet...ers/TL_205.pdf (http://alteclansingunofficial.nlenet.net/publications/techletters/TL_205.pdf)

is this chart applicable?


Absolutely! :yes:

Realize that there is always going to be a trade off between protection and phase shift.
In the Altec example, the chart is uses a crossover point that is 1/2 the frequency of the active crossover point.
That is 1 octave lower then the active crossover point and as a result it offers more protection to the driver.
The trade off is more phase shift. Since, most people dont have all of drivers physically time-aligned.
The protection cap phase shift may or may not present any noticeable difference to the overall systems sound.
I would recommend the Altec chart as a very good starting point.
If you amplifiers dont have any loud thumps at start up or shut down then it might be safe to use larger cap values as I have. Pictured below is the Altec Chart Example:

Baron030 :)
49831

Allanvh5150
02-10-2011, 10:40 AM
I have never tested the blocking cap idea but I would have thought if a large amplifier was being used, if the out put stage failed and went DC, say 100v, as the blocking cap charged, would the driver not see the DC?

Allan.

Eaulive
02-10-2011, 03:54 PM
I have never tested the blocking cap idea but I would have thought if a large amplifier was being used, if the out put stage failed and went DC, say 100v, as the blocking cap charged, would the driver not see the DC?

Allan.

Yes of course, but the few milliseconds it takes to charge the cap is better than the few seconds it takes for you to realize it and hit the kill switch :D

Baron030
02-11-2011, 07:22 AM
Hi Allen

Assuming your absolute worst case scenario where an amplifier could instantaneously produce a pure 100 volt DC current with no source series resistance. Such an amplifier with + & - 100 volt power supply rail voltages is far more powerfully when what would recommended for any high frequency compression driver.

Without a protection cap, the only thing that would limit the flow of current would be the DC resistance of the drivers voice coil. So, assuming that the driver voice coil has DC resistance is 5 ohms. Then the voice coil would start conducting 20 amps of current and generate 2,000 watts of heat. This would heat the thin aluminum wire up to a molten state very quickly, and for a brief moment it would go into a super heated plasma state, just before vaporizing. Poof no more voice coil.

Now, I will try to explain the physics of how a protection cap would prevent this from happening. For a very brief moment there would be 100 volts applied across the voice coil and the resulting current would start to charge the capacitor. As the capacitor charges, voltage that is stored in the capacitor will cause the voltage across the voice coil drop. Pictured below are charts showing of the voltages across the cap and voice coil relative to time. The value T for time can be calculated using a RC time constant formula (T = R x C) where T = seconds, R = to resistance in ohms and C is in farads. To give you an example, assuming the DC resistance is 5 ohms and that the protection cap is 50 micro farads then the time constant is 0.0025 seconds (T = 5 ohms x 0.0005 farads). So, during the first 1/400th of seconds the voltage across the voice coil will drop from 100 volts to 36.8 volts and then it will continue to drop to nearly zero by the fifth time constant, which would be 1/80th of a second. This gives very little time for heat to build up in the voice coil. So, with a protection cap, I would think that most drivers could survive a 100 volt DC voltage spike. But, I really would not want to put that to the test. Considering the high efficiency of JBL drivers, it would be the loudest, most ear splitting Click that you would ever heard in your entire life.
Baron030:)

49862

badman
02-11-2011, 07:38 AM
Hi Allen

Assuming your absolute worst case scenario where an amplifier could instantaneously produce a pure 100 volt DC current with no source series resistance. Such an amplifier with + & - 100 volt power supply rail voltages is far more powerfully when what would recommended for any high frequency compression driver.

Without a protection cap, the only thing that would limit the flow of current would be the DC resistance of the drivers voice coil. So, assuming that the driver voice coil has DC resistance is 5 ohms. Then the voice coil would start conducting 20 amps of current and generate 2,000 watts of heat. This would heat the thin aluminum wire up to a molten state very quickly, and for a brief moment it would go into a super heated plasma state, just before vaporizing. Poof no more voice coil.

Now, I will try to explain the physics of how a protection cap would prevent this from happening. For a very brief moment there would be 100 volts applied across the voice coil and the resulting current would start to charge the capacitor. As the capacitor charges, voltage that is stored in the capacitor will cause the voltage across the voice coil drop. Pictured below are charts showing of the voltages across the cap and voice coil relative to time. The value T for time can be calculated using a RC time constant formula (T = R x C) where T = seconds, R = to resistance in ohms and C is in farads. To give you an example, assuming the DC resistance is 5 ohms and that the protection cap is 50 micro farads then the time constant is 0.0025 seconds (T = 5 ohms x 0.0005 farads). So, during the first 1/400th of seconds the voltage across the voice coil will drop from 100 volts to 36.8 volts and then it will continue to drop to nearly zero by the fifth time constant, which would be 1/80th of a second. This gives very little time for heat to build up in the voice coil. So, with a protection cap, I would think that most drivers could survive a 100 volt DC voltage spike. But, I really would not want to put that to the test. Considering the high efficiency of JBL drivers, it would be the loudest, most ear splitting Click that you would ever heard in your entire life.
Baron030:)

49862

Cept it's not power that's a concern with compression drivers, but excursion. Might as well throw an atlatl at your phragm if you toss 100V on the driver even for a split second.

badman
02-11-2011, 08:34 AM
Because while you can use a large-valued cap, that's a full-on extraneous component. Instead, one might choose to make the CD EQ align with the protection cap. This, of course requires a flat impedance curve, so you wind up adding complexity there... and you're halfway to a passive XO, but you DO get a smaller cap(and more protection), and your EQ....


I always find it funny that so many active devotees (and I've used active extensively) refuse to acknowledge that some things are done well/better passively.

Baron030
02-11-2011, 08:56 AM
Ok Badman, I would have to agree.
Diaphragm damage do to over excursion would most likely to occur with a 100 volt spike.

But, realistically, I cant image someone hooking up a high current amplifier with 100 volt power supply rail voltages to a high frequency driver. Thats got to be an amplifier packing literally kilowatts of power.

The point of my posting was to show just how fast a protection cap can prevent a melt down.

Baron030:)

JeffW
02-11-2011, 09:49 AM
Might as well throw an atlatl at your phragm

While you could throw an atlatl at your phragm, it's actually used to throw phragms (or spears). :bouncy:

I think that's the first time since Jr. High that I've used "atlatl" in a conversation!

badman
02-11-2011, 10:02 AM
While you could throw an atlatl at your phragm, it's actually used to throw phragms (or spears). :bouncy:

I think that's the first time since Jr. High that I've used "atlatl" in a conversation!

Yeah, but it sounds better my way) Re: Jr. High.... shows ya how mature my thinking is :)