EPDR, speaker phase and DSP

[bud947]

Hello,

I recently learned about how EPDR and other power dissipation metrics may affect my system.
For reference, I have a pair of R700 which are drived by 2 x LA90 (in mono). The LA90 are fed with a pioneer LX-805 and I also have a minidsp flex at my disposal.
If my understanding is correct the LA90 should have little to no issues driving the minimum impedance of 3.6-ohm (very close to the amp 4-ohm minimum) as EPDR doesn't change the speaker impedance and thus the current required (roughly 9A). BUT the speaker phase will require extra power that will increase heat in the amplifier (class A/B) and trigger any protection circuit.

This lead me to think that one solution to EPDR may be correcting the anechoic phase of my speakers with DSP to reduce (or even suppress) power dissipation effects. Would that work ?

 

[DVDdoug]

I dont know what EPDR is but you can't put DSP between the amplifier & speaker. You can't change the phase relationship between the amp & speaker.

 

[Keith_W]

I dont know what EPDR is but you can't put DSP between the amplifier & speaker. You can't change the phase relationship between the amp & speaker.

Equivalent Phase Dissipation Resistance. ASR thread here.

I suppose it would be possible to convolve the EPDR response with the phase response of the speakers under anechoic conditions to generate some kind of transfer function that you could correct via DSP, but my reaction would be - why? In what way would this be superior to measuring the acoustic output of the driver under anechoic conditions with the amplifier you are using, and then DSP that instead?

 

[bmc0]

This lead me to think that one solution to EPDR may be correcting the anechoic phase of my speakers with DSP to reduce (or even suppress) power dissipation effects. Would that work ?

No. The extra power dissipation is due to the voltage and current not being exactly in phase. You cannot change this relationship with signal processing placed before the amplifier.

 

[bud947]

So DSP only works (to some extents) for room correction and frequency response. This means that the only solution in my case is to use an amplifier rated into 2 ohms and therefore that can match the EPDR requirement at least for peak current of the maximum voltage/power I need.

 

[bmc0]

This means that the only solution in my case is to use an amplifier rated into 2 ohms

I wouldn't be so hasty with this conclusion. Have you ever triggered the LA90's protection circuitry? Unless you're running the amps at the limits of their power capability, I wouldn't expect problems.

 

[RayDunzl]

If my understanding is correct the LA90 should have little to no issues driving the minimum impedance of 3.6-ohm (very close to the amp 4-ohm minimum) as EPDR doesn't change the speaker impedance and thus the current required (roughly 9A). BUT the speaker phase will require extra power that will increase heat in the amplifier (class A/B) and trigger any protection circuit.

Has the protection been triggered in the past?

If not... Well...

My crude power graph shows 70W into 3.6 ohms and 87dB sensitivity gives 102dB for one speaker at one meter.

Do you push it to its limits? 102dB SPL is pretty loud.

 

[amirm]

Please ignore EPDR. As @RayDunzl correctly states, leave it to the designer to have put in protection circuits that kick in should it think the output stage is being overloaded. If the amp is shutting down, you need a more powerful one. Otherwise, there is no reason whatsoever to be concerned about an academic assumption about output stage of an amplifier.

 

[bud947]

Indeed I have triggered LA90's protection circuit after running it through the LX-805 with some DSP (Dolby Surround mode). This is the reason I was looking for a DSP solution to keep my amp while having its full 200w capacity. I listen at around 3m and was playing at loud level.

I was refering to EPDR as a indicator of the minimum stable impedance I should look for a specific speaker. In my case, for polk r700 my understanding is that I would be better with an amplifier that is stable (and ideally voltage linear) at 4/2 ohm load.

 

[amirm]

There is no certainty that what the protection circuit in LA90 thinks is an overload and what EPDR says. With the DSP, you need even more power than indicated so the solution is to get a more powerful amplifier.

 

[dfuller]

Wrong kind of phase, I'm afraid - acoustic phase rotation around crossovers and electrical phase shift are not the same thing, and so you can't correct for the latter by unwrapping the former.

 

[Old_School_Brad]

Hello,

I recently learned about how EPDR and other power dissipation metrics may affect my system.
For reference, I have a pair of R700 which are drived by 2 x LA90 (in mono). The LA90 are fed with a pioneer LX-805 and I also have a minidsp flex at my disposal.
If my understanding is correct the LA90 should have little to no issues driving the minimum impedance of 3.6-ohm (very close to the amp 4-ohm minimum) as EPDR doesn't change the speaker impedance and thus the current required (roughly 9A). BUT the speaker phase will require extra power that will increase heat in the amplifier (class A/B) and trigger any protection circuit.

This lead me to think that one solution to EPDR may be correcting the anechoic phase of my speakers with DSP to reduce (or even suppress) power dissipation effects. Would that work ?

The Polk R700 should not be an issue to most amplifiers.

Spoiler: Polk R700 Phase/Impedance trace

If you're triggering the protection circuitry, that's a clear sign you need a more powerful amplifier -ideally one with at least twice the output of the LA90 in 4-6 ohm to see any real-world improvement.