Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[Rolando Saenz]

Wow, the graph looks very nice, it is almost prefect flat

i found this on setting limiters on the Danley Web page

How should I set my limiter?

There are no limiter settings that can guarantee that you will not “blow up” a loudspeaker
unless it is set so low as to inhibit the capability of the loudspeaker to get loud.

There is a large “grey area” for the settings.
The actual settings for protection will depend on the a few things, the dynamic range of the material being played-the type of material and so forth.
Some musical styles are more abusive on loudspeakers than others Below is a basic guideline.

I prefer to actually use 2 limiters if possible; one as a “peak stop” and the other as a “heating” limiter.
The 2 things that cause damage to loudspeakers are over excursion (too much cone movement which causes physical damage)
and overheating-which can melt wires, glue etc. I prefer to talk in terms of voltage applied to the loudspeaker-because that is easy to measure.

The following formula should be used when converting power (wattage) to voltage. Take the square root of (The power rating X rated impedance).
This will give a voltage to be measured. This voltage should be measured at the output of the amplifier using a sine wave WITHOUT a loudspeaker hooked up.
If the loudspeaker is hooked up-it will probably be destroyed.

For the peak stop limiter-the shortest attack time should be used and the voltage would be equal to the program rating of the loudspeaker.
A ratio of 20:1 should be used.
For example, the voltage applied to a 4 ohm ( nominal) loudspeaker rated at 2000 watts program, would be the square root of (2000x4) or 126 volts.
But if you push the level to where this limiter is limiting hard (lots of reduction) then your average level will be a lot higher and could cause heating damage because the average level applied will be higher.
The peak stop is just to stop the occasional peak-not a long term "leveler".

The heating limiter is a different story. The actual setting could vary a good bit-depending on the parameters available
and the type of source material being played etc. A typical range would be from around ½ the continuous rating up to the continuous rating.
Attack times should be at a minimum 100ms and preferably around 2-3 seconds. As a general rule, longer attack times would result in lower voltage settings. The ratio should be at least 6:1 and up to 10:1 or higher. The voltage is calculated from the continuous rating
(not the program as for the peak limiter), and ½ the rating is a good start.
So for the same loudspeaker as above the continuous rating of 2000 watts program is 1000 watts.
So half of that would be 500 watts. So the square root of 500x4 is 45 volts.
The only time this limiter would be engaged is when the 45 volt settings is exceeded for more than the attack time (2-3 seconds).
The maximum setting would be equal to the continuous rating, or square root of 1000x4 or 63 volts.
So somewhere between 45 and 63 volts would be a good setting for the slow or “heating” limiter.Y
You may be able to set these higher and be just fine-especially with material that has a large dynamic range.
NOT with compressed MP3s or dance music that is mixed to be as loud as possible, because the dynamic range is not very large on those sources.

The best way to measure the voltage is to use a sine wave (at the correct freq range being used)
going into the compressor/DSP etc and turn the threshold all the way up, or to as high a value as possible.
Then using a volt meter hooked to the output, adjust the level (of the generator or the compressor input
NOT the compressor output or amplifier level control) until the voltmeter measures the correct value.
Then turn down the threshold level until the limit meter just starts to register.
This is usually 1dB. You should also notice that the voltage has gone down just a little bit.
This would be the correct setting.

DO NOT change the amplifier gain or compressor output levels-as this will throw off the actual threshold set.

[Art Welter]

Hi Rolando,

What Art said, re amps and high pass filter.

But ease into that.  My design was to dig low, 30Hz f3, and has big ports to do so.
Excursion is greatest around 40Hz.
The bms i used has a 19mm xmax. The TBX100 9mm.
So excursion is going to be needed to approached a little more
carefully.
It's not a big deal, and frankly, given the $ difference, if i build any more it will be with the TBX100.

Mark,
Cutting Xmax in half is a 6dB LF output reduction, while the driver cost difference is about 3dB, not to mention the weight difference while suffering with a healing shoulder, hmmm- do I hear some pain meds talking  ?

Cheers,
Art

[Mark Wilkinson]

Mark,
Cutting Xmax in half is a 6dB LF output reduction, while the driver cost difference is about 3dB, not to mention the weight difference while suffering with a healing shoulder, hmmm- do I hear some pain meds talking  ?

Cheers,
Art

Good stuff Art, made me laugh.  Hey, I'm feeling proud...ditched the pain meds within 48 hours !

Ok, the weight diff...yeah, that sucks... but everybody can read that spec without any help
The $ dB ?   i dunno...guess it depends on how many $ you have.  Me? Getting 2 TBX100 for less than Assistance will let go of one bms...uhm, that's a 6 dB gain to my wallet..

But more seriously....the xmax stuff....below are two hornresp sims that have been displacement limited at 40Hz...the peak displacement area.
First is the bms using 19mm. 2nd is tbx100 using 11mm xvar.
I'd call it about a 4 dB difference.  (The Equalizer filter is a single out-of-band PEQ to kill the plenum resonance)

[Rolando Saenz]

Good stuff Art, made me laugh.  Hey, I'm feeling proud...ditched the pain meds within 48 hours !

Ok, the weight diff...yeah, that sucks... but everybody can read that spec without any help
The $ dB ?   i dunno...guess it depends on how many $ you have.  Me? Getting 2 TBX100 for less than Assistance will let go of one bms...uhm, that's a 6 dB gain to my wallet..

But more seriously....the xmax stuff....below are two hornresp sims that have been displacement limited at 40Hz...the peak displacement area.
First is the bms using 19mm. 2nd is tbx100 using 11mm xvar.
I'd call it about a 4 dB difference.  (The Equalizer filter is a single out-of-band PEQ to kill the plenum resonance)

uh ??
Mark,
those graphs are not the sensitivity difference of the drivers right ?
you are using more power to the BMS ?
i mean a bigger amps or in other words more watts?

BMS is more than double the price of the 18TBX100 , so 2 18tbx100 cabs instead of one cab loaded with the BMS
only thing is that we will need 2 amps
or at least a hugeee one that can drive 1 cab per channel at 4ohms

the powersoft K20 can deliver 5200 watts/ch in 2 ch mode@4 ohms ,that will feed 2600 watts to each driver on  one cab
so you can drive 2 cabs, one per channel on one K20, altough maybe is too much, i am not that "watts savy" lol

what is better to implement
1 2ch bridged amp @4 ohms per cab
or 1 2ch amp driving one cab/channel@4 ohms
 

[Tim McCulloch]

Volts move cones when met with electrical resistance, the potential that is not converted to motion becomes heat. Think about that for a minute.

[Art Welter]

Getting 2 TBX100 for less than Assistance will let go of one bms...uhm, that's a 6 dB gain to my wallet..

But more seriously....the xmax stuff....below are two hornresp sims that have been displacement limited at 40Hz...the peak displacement area.
First is the bms using 19mm. 2nd is tbx100 using 11mm xvar.
I'd call it about a 4 dB difference. 

Doubling of watts (or $$) is 3dB, 6 dB is quadrupling...

[Mark Wilkinson]

uh ??
Mark,
those graphs are not the sensitivity difference of the drivers right ?
you are using more power to the BMS ?
i mean a bigger amps or in other words more watts?

BMS is more than double the price of the 18TBX100 , so 2 18tbx100 cabs instead of one cab loaded with the BMS
only thing is that we will need 2 amps
or at least a hugeee one that can drive 1 cab per channel at 4ohms

the powersoft K20 can deliver 5200 watts/ch in 2 ch mode@4 ohms ,that will feed 2600 watts to each driver on  one cab
so you can drive 2 cabs, one per channel on one K20, altough maybe is too much, i am not that "watts savy" lol

what is better to implement
1 2ch bridged amp @4 ohms per cab
or 1 2ch amp driving one cab/channel@4 ohms

Rolando, you are correct, those graphs were not sensitivity graphs.  They are the SPL that can be achieved without the drivers over excursion.
Any yes, the bms has more power, or in more appropriate term as Tim keeps trying to help you think in, voltage.
The thing is, power to the driver varies so much by frequency, it tends to become a ball park spec that loses meaning the more you understand.

Anyway, Art in his reply to me,was kindly reminding me, that as well as the tbx100 does in place of a drop-in replacement for the bms,
once displacement kicks in, the bms will keep trucking past. 
He's helped a boatload of folks with alternate drivers to many of his designs, and has learned how to give salient comparisons quickly. 
I'm a first timer with this.  btw, did you look at his Keystone sub?

Ok, back to this sub with the tbx100.
Below are two approaches I'd take. Both use a 120 BW4 lpf. And have a PEQ for killing the plenum resonance 178Hz, -9.6dB, 1.5Q.

The first is designed to go as low as reasonable without worry.  It is based on a 63V drive voltage. It has a 25Hz BW4 hpf.
The second trades a little low end extension for a little more overall SPL.  It it based on 77V. It has a 31Hz LR2 hpf.

Graphs go like this this:
Acoustic power, then driver excursion, then amplifier power at the voltages above..

See how the power swings around so much.  Due to changing impedance, which i figure you no doubt know...it's just i don't think most folks know how much power actaully swings around by frequency.

Anyway, a semi-legitimate formula for determining amp wattage is V^2/nominal impedance.
Which for case 1, 63V: is about 1000w @ 4 ohms, and for case 2, 77V:  is about 1500w @ 4 ohms. Probably less than you imagined based on driver power spec...

Driver specs vs the reality of excursion through the entire passband, along with thermal compression from driving any close to driver specs, is why you the wise experienced heads say keep it to 1/3 to 1/2 AES rating.

Hope this helps..good luck !

[Rolando Saenz]

Ok, back to this sub with the tbx100.
Below are two approaches I'd take. Both use a 120 BW4 lpf. And have a PEQ for killing the plenum resonance 178Hz, -9.6dB, 1.5Q.

Mark,
but if we implement a 120Hz LP, why you need the 178hz PEQ
anything above 120Hz will be cut
so we do not need the PEQ
do we ??

[Mark Wilkinson]

Mark,
but if we implement a 120Hz LP, why you need the 178hz PEQ
anything above 120Hz will be cut
so we do not need the PEQ
do we ??

You can skip it if you want, you might even like the extra power that would provide below crossover
Personally, i like to tune to flat..

Response without the PEQ will look more like..

[Tim McCulloch]

Mark -

"The first is designed to go as low as reasonable without worry.  It is based on a 63V drive voltage. It has a 25Hz BW4 hpf.
The second trades a little low end extension for a little more overall SPL.  It it based on 77V. It has a 31Hz LR2 hpf."

You mean you're not "driving with Watts"? Huh?  /sarcasm

Rolando - the above is why I've been obliquely trying to get you to do some homework.  WATTS are relevant as an expression of potential (volts) turned into work (motion or heat), and as a way of referencing the amount of work that can be produced.  It's important that you understand the difference.

[ProSoundWeb Community]