Thermistor L-Pad to correct dynamic compression

[Bill F.]

Woke up this morning with this odd idea on my mind :

How about making an L-pad using NTC and PTC thermistors tuned to the thermal properties of the loudspeaker's VC/motor system to compensate for loudspeaker voice-coil heating?

As VC temperature fluctuations cause a loudspeaker's Re to wander, dynamics are impacted by varying degrees of power compression, and the tuning of passive crossover networks becomes compromised. Perhaps this could fix both problems.

If you wanted to get a little crazy, you could even tune the thermistors to supply a touch of artificial dynamic expansion.

The obvious penalty, of course, would be the loss of sensitivity that comes with use of any L-pad, but if your application needs an L-pad anyway, this added functionality might make sense.

Anyone ever heard of this being done? A cursory search revealed nothing.

Bill

[Bill F.]

Here's a diagram.

[transphere]

So how are you going to find components with dynamic properties, time constants and stuff to match the voice coil heating process? I think in analog this will prove extremely difficult.

O.t.o.h, using current measurements on the speaker and some proper DSP, it should be possible to determine the change in impedance due to the temperature rise and compensate for it. I think the difficult part of it would be mapping the change in impedance to some kind of 'dynamic eq' or multiband compressor to get the proper power-compression-to-frequency curve.

[Brian Oshman]

That may adjust for crossover values but the voice coil resistance will still fluctuate. How about a water cooled magnet structrue?

Don't use any passive crossovers and use all fully variable active crossovers and muliple amp channels.

[Bill F.]

So how are you going to find components with dynamic properties, time constants and stuff to match the voice coil heating process?

Yeah, that probably wouldn't be trivial.

However, IIRC, some other threads here seemed to indicate that instantaneous heating effects were less important than longer-term temperatures. Maybe you could sink the thermistors to the motor structure.

[Bill F.]

That may adjust for crossover values but the voice coil resistance will still fluctuate.

And the thermistors compensate for that by raising the voltage on the voice coil as temps rise.

[Dr.EM]

Originally Posted by Brian Oshman

Don't use any passive crossovers and use all fully variable active crossovers and muliple amp channels.

An effective and quite practical solution

[Bill F.]

Don't use any passive crossovers and use all fully variable active crossovers and muliple amp channels.

Off topic, but you're preaching to the converted. I have a Rane RPM88 and an 8-channel amp.

[Brian Oshman]

Originally Posted by Bill F.

Off topic,

Ahhhh, but it is a viable solution.

[jcx]

the BeoLab 5 active loudspeaker has voice coil heating nonllinear compensation internally in dsp

good thermal modeling may be a little complicated:
http://www.klippel.de/pubs/Klippel%2...ransfer_03.pdf