ProSoundWeb Community

Sound Reinforcement - Forums for Live Sound Professionals - Your Displayed Name Must Be Your Real Full Name To Post In The Live Sound Forums => LAB Subwoofer Forum => Topic started by: Andre Vergison on January 06, 2021, 03:20:57 pm

Title: Thermal limiter: why not based on RMS current?
Post by: Andre Vergison on January 06, 2021, 03:20:57 pm
Having read a lot about thermal limiters and RMS limiters , I've always wondered why manufacturers base their designs on the signal VOLTAGE?

The (simplified) electrical equivalent circuit of a speaker driver is a series of a resistive part and an inductive part (and an EMF). The resistive part is the voice coil resistance. So the voice coil heating will be proportional to the square of the CURRENT thru that circuit, not the voltage.
Yes - the voltage is okay as long as the speaker is resistive, but it never is.

For instance, the power fed to a speaker near the resonance frequency (or near the two impedance peaks of a bass reflex) is way lower than elsewhere, even on high voltages.

IMHO the CURRENT, and in particular the RMS value of it, would be better data to let the limiter work on.

Agree / don't agree?
Title: Re: Thermal limiter: why not based on RMS current?
Post by: David Sturzenbecher on January 06, 2021, 04:50:10 pm
Having read a lot about thermal limiters and RMS limiters , I've always wondered why manufacturers base their designs on the signal VOLTAGE?

The (simplified) electrical equivalent circuit of a speaker driver is a series of a resistive part and an inductive part (and an EMF). The resistive part is the voice coil resistance. So the voice coil heating will be proportional to the square of the CURRENT thru that circuit, not the voltage.
Yes - the voltage is okay as long as the speaker is resistive, but it never is.

For instance, the power fed to a speaker near the resonance frequency (or near the two impedance peaks of a bass reflex) is way lower than elsewhere, even on high voltages.

IMHO the CURRENT, and in particular the RMS value of it, would be better data to let the limiter work on.

Agree / don't agree?

This is a limiter type available to powersoft users...
The TruePower limiting is a Powersoft technology useful to avoid overheating of the voice coil and can be used to avoid power compression. The DSP provides the measurement of the real power delivered (and then dissipated) to the coil, not the apparent power handled by the line.
Title: Re: Thermal limiter: why not based on RMS current?
Post by: John Roberts {JR} on January 06, 2021, 06:01:42 pm
Having read a lot about thermal limiters and RMS limiters , I've always wondered why manufacturers base their designs on the signal VOLTAGE?
"manufacturers"?

Generally voltage is all the aftermarket has access to without placing thermal sensors inside the drivers.

That said there are smart driver protection circuits that can directly impute voice coil temperature from sensing voltage and current. I have been out of the trenches for 20 years but even back then we sold big dog amps for professional fixed install market that sensed both voltage and current at the amplifier speaker feed. Since the resistance of the voice coil increases with temperature you could predict temperature from comparing voltage and current draw to a cold reference resistance. Of course you have to learn the behavior for specific drivers. This amplifier current sensing is probably more commonly used to detect blown drivers in a large entertainment venue with tens of speakers (or more), so they can tell maintenance which boxes to repair.

I can also imagine a smart powered speaker that is made very hard to kill with DSP inside and some specific knowledge about the drivers. 

IMO impractical for stand alone aftermarket protection modules because the end users would need to know more about their speakers than most do. I though about this for a product but dismissed it because it would require so much hand holding (and claims for expensive replacement speakers if the customer didn't set up the protection properly).

I seem to recall one dsp based driver protection card that mounts right on the driver ( sold by eminence? IIRC), but probably cheaper to do that in software these days.

Quote
The (simplified) electrical equivalent circuit of a speaker driver is a series of a resistive part and an inductive part (and an EMF). The resistive part is the voice coil resistance. So the voice coil heating will be proportional to the square of the CURRENT thru that circuit, not the voltage.
Yes - the voltage is okay as long as the speaker is resistive, but it never is.

For instance, the power fed to a speaker near the resonance frequency (or near the two impedance peaks of a bass reflex) is way lower than elsewhere, even on high voltages.

IMHO the CURRENT, and in particular the RMS value of it, would be better data to let the limiter work on.

Agree / don't agree?
You are under thinking this... This is more complex than just performing RMS math, but not a lot more. Without knowing exactly how fast the driver dissipates heat you are still left guessing even if you perfectly compute the power going into the drivers. That is only one piece of the puzzle.

JR
Title: Re: Thermal limiter: why not based on RMS current?
Post by: Brian Jojade on January 06, 2021, 06:11:43 pm
While it may be better data, voltage data is usually close enough to get the job done and far less complex and expensive to implement.
Title: Re: Thermal limiter: why not based on RMS current?
Post by: Scott Holtzman on January 07, 2021, 02:59:52 am
"manufacturers"?

Generally voltage is all the aftermarket has access to without placing thermal sensors inside the drivers.

That said there are smart driver protection circuits that can directly impute voice coil temperature from sensing voltage and current. I have been out of the trenches for 20 years but even back then we sold big dog amps for professional fixed install market that sensed both voltage and current at the amplifier speaker feed. Since the resistance of the voice coil increases with temperature you could predict temperature from comparing voltage and current draw to a cold reference resistance. Of course you have to learn the behavior for specific drivers. This amplifier current sensing is probably more commonly used to detect blown drivers in a large entertainment venue with tens of speakers (or more), so they can tell maintenance which boxes to repair.

I can also imagine a smart powered speaker that is made very hard to kill with DSP inside and some specific knowledge about the drivers. 

IMO impractical for stand alone aftermarket protection modules because the end users would need to know more about their speakers than most do. I though about this for a product but dismissed it because it would require so much hand holding (and claims for expensive replacement speakers if the customer didn't set up the protection properly).

I seem to recall one dsp based driver protection card that mounts right on the driver ( sold by eminence? IIRC), but probably cheaper to do that in software these days.
You are under thinking this... This is more complex than just performing RMS math, but not a lot more. Without knowing exactly how fast the driver dissipates heat you are still left guessing even if you perfectly compute the power going into the drivers. That is only one piece of the puzzle.

JR


John do you know how the Meyer processors worked that had an input to the speaker?  I know it did more than limit it changed the crossover frequency too IIRC.
Title: Re: Thermal limiter: why not based on RMS current?
Post by: Doug Fowler on January 07, 2021, 03:52:53 am

John do you know how the Meyer processors worked that had an input to the speaker?  I know it did more than limit it changed the crossover frequency too IIRC.

I was told by a "guy" in a position to know it's actually Apogee controllers that used the sliding xover, not Meyer. I used to have UPAs and USWs pre-measurement days.

I would love to see this settled. Anyone with a biamp Meyer analog controller could easily verify this with a simple measurement.
Title: Re: Thermal limiter: why not based on RMS current?
Post by: Helge A Bentsen on January 07, 2021, 05:16:11 am
I was told by a "guy" in a position to know it's actually Apogee controllers that used the sliding xover, not Meyer. I used to have UPAs and USWs pre-measurement days.

I would love to see this settled. Anyone with a biamp Meyer analog controller could easily verify this with a simple measurement.

According to a measurement friend, Meyer didn’t use sliding xovers, but once the controller started limiting in one band, the audible effect was similar to sliding crossovers. When measured with one band in limit, one of the bands would stay at it’s level while the other would increase level leading to the audible xover moving. The electronic crossover stayed put.
Title: Re: Thermal limiter: why not based on RMS current?
Post by: John Roberts {JR} on January 07, 2021, 09:42:33 am

John do you know how the Meyer processors worked that had an input to the speaker?  I know it did more than limit it changed the crossover frequency too IIRC.
I don't have any specific knowledge about that but vaguely recall some pre-DSP speaker processors that used a sliding HPF to reduce LF bass energy at high output levels.

I also suspect some audiophools played with 4 wire Kelvin sensing (2 speaker wires and 2 sense wires) to provide numerically high damping factor.

I momentarily considered an after market SKU that performed the the current sensing between amp and loudspeaker to provide the data to impute voice coil temperature, but almost as quickly abandoned it after thinking about the nightmare of customer service for users that in fact need such a product, and speaker specific set-up assistance. 

By now, decades later this should be mature technology and premium brands should be pretty well managed. Smart power amps with inboard current sensing, and modest DSP "can" do this but must be programmed for specific loudspeakers. 

Caveat, I have been out of the trenches for a couple decades so not up on latest technology.

JR
Title: Re: Thermal limiter: why not based on RMS current?
Post by: Lee Douglas on January 07, 2021, 11:13:11 am

I also suspect some audiophools played with 4 wire Kelvin sensing (2 speaker wires and 2 sense wires) to provide numerically high damping factor.


I had a Peavey product called Dynamic System Controller that did that.  I used it on a 3020ht/415SUB system.  Output to amplifiers and low/mid/high sense wires from the speaker output of each amplifier.  Or was that something different?
Title: Re: Thermal limiter: why not based on RMS current?
Post by: Jim McKeveny on January 07, 2021, 12:45:19 pm
Apogee controllers ... used the sliding xover, not Meyer. 

Renkus-Heinz had "Smart" loudspeaker line that moved crossover points upward for driver protection. Boxes were 2-15's handing off to a 2 inch HF. The whole crossover region polars adjusted dynamically with material. I gigged outdoors in San Jose on one of these rigs, way undersized for the crowd. The sonic effect was, umm, novel.

Title: Re: Thermal limiter: why not based on RMS current?
Post by: John Roberts {JR} on January 07, 2021, 01:14:02 pm
I had a Peavey product called Dynamic System Controller that did that.  I used it on a 3020ht/415SUB system.  Output to amplifiers and low/mid/high sense wires from the speaker output of each amplifier.  Or was that something different?
That was perhaps a little before my time, and definitely from a different engineering group.

My guess is that it was a Peavey version, of some other companies respected product. The transducer group didn't do much of their own electronic circuit design so that was probably inspired by some other companies existing design. I don't know what was inside.

JR 

PS: Speaking of damping factor Jack Sondermeyer (RIP) invented a circuit to allow amplifiers to deliver negative output impedance. Not only could he drive the amplifier's output impedance to zero ohms, be he could go further to compensate for 100% of the speaker wire losses, and even inside the box to compensate for crossover and voice coil impedance. The ability to alter driver behavior is obvious, less obvious how to use it productively for Peavey customers. This technology was clearly too cool for the room (Peavey distribution) and the high damping factor capability was only built into one amp model (the CS800s). 
Title: Re: Thermal limiter: why not based on RMS current?
Post by: Andre Vergison on January 07, 2021, 04:06:17 pm
The (simplified) electrical equivalent circuit of a speaker driver is a series of a resistive part and an inductive part (and an EMF). The resistive part is the voice coil resistance. So the voice coil heating will be proportional to the square of the CURRENT thru

My original simplified 'model' did not elaborate the EMF part, which, for a closed box or a driver in open air, can be modeled as a parallel LCR circuit. So the total load is 'complex', it has a resistive part and a reactive part. The reactive part cannot dissipate or produce any power.

The 'real' power is the power dissipated by the resistive part, which in case of a speaker is the sum of the voice coil resistance and the resistive part of the LCR circuit (EMF). We know that the total resistance varies by frequency between the voice coil resistance and a number of times that value.

The voice coil resistance is a constant for a given temperature. It rises as the coil temperature increases.

The point is that the heating of the voice coil is purely caused by the value of the current, and also by the temperature dependent value of the voice coil resistance (the product).

That means that an amplifier can safely feed a speaker with a voltage higher than normal at cone resonance frequency (leaving aside Xmax), simply because the resistive part of the EMF is many times higher than the voice coil resistance, e.g. 100 ohms for a 8 ohms speaker. Hence the current is low, and hence little voice coil heating will occur.

Take a typical bass reflex enclosure (see pic). Again leaving Xmax aside, the actual impedance for played bass notes will be way higher than 8 ohms. Why would we then thermal-limit the input voltage if the current, and heating, is lower?

That's exactly my point: RMS limiters don't take current into account, they work on voltage only, so they may limit needlessly (or not enough).

This is a limiter type available to powersoft users...
The TruePower limiting is a Powersoft technology useful to avoid overheating of the voice coil and can be used to avoid power compression. The DSP provides the measurement of the real power delivered (and then dissipated) to the coil, not the apparent power handled by the line.

Powersoft's TruePower seems to be the right thing indeed, according to https://www.soundpro.com/catalog/documents/22802_2.pdf , see 6:10.6.2:

TruePower™: the amplifier’s active output power is estimated by measuring the load current. The Truepower limiter is a powersoft patent technology useful to avoid overheating of the voice coil; it can however also be used to avoid power compression. The DSP provides the measurement of the real power delivered (and then dissipated) to the coil, ignoring the apparent power handled by the line.


Looking good. They also allow empirical parameters to be entered about type of speaker, thermal inertia, number of speakers etc.
Glad to see they did it this way.

More about complex (resistive and reactive) power (you can skip the formulas) :
https://electrical-engineering-portal.com/complex-power-analysis

Interesting analysis of speaker back-EMF (4 parts):
https://audioxpress.com/files/attachment/2701
https://audioxpress.com/files/attachment/2702
https://audioxpress.com/files/attachment/2703
https://audioxpress.com/files/attachment/2704
Title: Re: Thermal limiter: why not based on RMS current?
Post by: Ivan Beaver on January 07, 2021, 04:57:56 pm
I was told by a "guy" in a position to know it's actually Apogee controllers that used the sliding xover, not Meyer. I used to have UPAs and USWs pre-measurement days.

I would love to see this settled. Anyone with a biamp Meyer analog controller could easily verify this with a simple measurement.
In the first or second edition of sound system engineering, there was a mention specifically of Meyer and the sensing and sliding xover.  I loaned my copy out and have never seen it since :(
Title: Re: Thermal limiter: why not based on RMS current?
Post by: Ivan Beaver on January 07, 2021, 05:01:50 pm


PS: Speaking of damping factor Jack Sondermeyer (RIP) invented a circuit to allow amplifiers to deliver negative output impedance. Not only could he drive the amplifier's output impedance to zero ohms, be he could go further to compensate for 100% of the speaker wire losses, and even inside the box to compensate for crossover and voice coil impedance. The ability to alter driver behavior is obvious, less obvious how to use it productively for Peavey customers. This technology was clearly too cool for the room (Peavey distribution) and the high damping factor capability was only built into one amp model (the CS800s).
That sounds a lot like what powersoft is doing in their amps.

I believe an old Crown (Delta ohmega?) that would do the same sort of thing, or so I remember, but it has been a few decades.
Title: Re: Thermal limiter: why not based on RMS current?
Post by: Andre Vergison on January 08, 2021, 12:27:36 pm
PS: Speaking of damping factor Jack Sondermeyer (RIP) invented a circuit to allow amplifiers to deliver negative output impedance. Not only could he drive the amplifier's output impedance to zero ohms, be he could go further to compensate for 100% of the speaker wire losses, and even inside the box to compensate for crossover and voice coil impedance.

https://sound-au.com/project56.htm#s4
Title: Re: Thermal limiter: why not based on RMS current?
Post by: Andre Vergison on January 08, 2021, 01:21:45 pm
This is a limiter type available to powersoft users...

Apparently to ElectroVoice users as well:
https://electrovoice.com/media/downloads/thermal_energy_management_and_protection_limiter_technology_v1-0.pdf

" Another  method  would  be  a  combined  approach  of  monitoring  the  voltage  and  current  delivered  by  the  amplifier.  Unfortunately,  this  would  make  the  parameters  depend  upon  how  many  speakers  the  user  had  paralleled  on  each  channel  and  would  be  limited  to  amplifiers  with  DSPs  on  board. "

That's what PowerSoft does. So EV doesn't like it. What they do is the following:

" Sample  by  sample,  the  TEMP  limiter  tracks  the  voltage  applied  to  the  loudspeaker  terminals  and  uses  this  to  calculate  the  instantaneous  temperature  of  the  voice  coil.  "

As discussed above, voice coil temperature computed from voltage won't be correct, due to the complex load.

Some striking quote though:

" The difference between a coil temperature of 250C and 200C is 1dB. "

Shiver.
Title: Re: Thermal limiter: why not based on RMS current?
Post by: John Roberts {JR} on January 08, 2021, 02:21:14 pm
That sounds a lot like what powersoft is doing in their amps.

I believe an old Crown (Delta ohmega?) that would do the same sort of thing, or so I remember, but it has been a few decades.
Jack got a patent on his circuit. At the time Crown was "advertising" amplifiers with silly high damping factor, but as anyone with a grasp of math understands a modest length of speaker wire will dominate damping factor even in a combination using a zero ohm output impedance audio amplifier. 

JR

PS : I am not aware that this is some unresolved problem.

PPS: The Peavey variable amplifier source impedance technology was also used in solid state guitar amps, in the opposite direction, to mimic the loudspeaker/amp interactions common with higher output impedance tube guitar amps. Peavey even tweaked the guitar amp source impedance differently for high and low frequency for different sonic impact.
Title: Re: Thermal limiter: why not based on RMS current?
Post by: Art Welter on January 08, 2021, 03:40:58 pm
Some striking quote though:

" The difference between a coil temperature of 250C and 200C is 1dB. "

Shiver.
At about 232C (451F) paper starts to burn, a lot bigger problem than 1dB loss ;^).

Title: Re: Thermal limiter: why not based on RMS current?
Post by: John Roberts {JR} on January 09, 2021, 10:25:51 am
At about 232C (451F) paper starts to burn, a lot bigger problem than 1dB loss ;^).
One of the advances in loudspeaker technology over the years is development of higher temperature adhesives and high temperature wire insulation... resulting in power compression instead of outright failure, when things get really hot inside. When the VC gets hot and it's resistance increases and it makes less output from the same drive voltage

JR

PS: (232 x 9/5) +32 = 449.6  close enough to 451 for government work. The formers and components in immediate contact with the voice coils are probably not paper these days. 

 

 
Title: Re: Thermal limiter: why not based on RMS current?
Post by: Riley Casey on January 09, 2021, 03:54:59 pm
Ray Bradbury would be aghast

...
JR

PS: (232 x 9/5) +32 = 449.6  close enough to 451 for government work. The formers and components in immediate contact with the voice coils are probably not paper these days.
Title: Re: Thermal limiter: why not based on RMS current?
Post by: Steve M Smith on January 14, 2021, 07:16:37 am
That was perhaps a little before my time, and definitely from a different engineering group.

My guess is that it was a Peavey version, of some other companies respected product. The transducer group didn't do much of their own electronic circuit design so that was probably inspired by some other companies existing design. I don't know what was inside.

Did the Peavey HDH system incorporate any thermal limiting? Or was that more of a servo feedback sort of thing? I remember it being likened to a Meyer system a the time.


Steve.
Title: Re: Thermal limiter: why not based on RMS current?
Post by: Douglas R. Allen on February 15, 2021, 06:10:29 pm
Did the Peavey HDH system incorporate any thermal limiting? Or was that more of a servo feedback sort of thing? I remember it being likened to a Meyer system a the time.


Steve.

     I had 2 types of those for a very short time. One ( again way back cloudy memory ) was a generic Sub ,Mid and Top type. DSC Sub or something like that.  Like others have mentioned there was a sense line that measured , I'm guessing , the voltage coming from the speaker and moved the crossover point and I think some limiting. Any 1/4 inch speaker cable would work. I had my SP2's and subs on it.   The other one was for a Horn,Mid-High,Sub system. I remember 4 22A horn drivers in the top. A 12 inch or 10 inch midrange and 2-18 sub using a slight Manifold arrangement. There was pre made cables that went with the system.  I know those in my area who had them and drove them hard still burned a lot of speakers but if you just flickered the lights on them I could run the system for days without an issue.  I think they really increased the RMS value if you were out pushing hard for a long time.

   Douglas R. Allen

A quick look and I found the DSC Series 23 and HDH I had.

https://peavey.com/manuals/80300823.pdf

https://manualzz.com/doc/4628752/peavey-dynamic-system-controller-series-hdh-owner-manual