Bi-amping and temperature management.

[KEW]

I'm curious about thermal implications for the receiver (in my case).
I have a 7.1 receiver, but only have 5.1, so the amps are idle as is.
My thought is it is free. I have never read any disadvantage (aside from cost of amplification or wire - both of which I already have).
The advantages I have read about seem to be negligible, but I have never seen any comment on the thermal aspect of bi-amping.

I am no EE and hope those of you that know more about amps will comment on the accuracy of my reasoning (which may well be based on bogus ASSUMPTIONS):
1) If the same amount of power is used for bi-amping, is it advantageous to spread the heat out over the extra circuits, thereby reducing he highest temperature seen by any individual component.
2) Is heat output linear or logarithmic? IOW, if bi-amping reduces the load on the lower frequency amp by 5% would the resulting decrease in temperature (relative to room temperature) be 5% or would it be less or greater than 5%?

 

[PENG]

Interesting question, without deep thinking, I would say assuming the same load the overall heat produced should be roughly the same but heat dissipation will become more efficient though it won't be significant. You get to use the heat sinks of the other two amp channels.

 

[TheWarrior]

I'm curious about thermal implications for the receiver (in my case).
I have a 7.1 receiver, but only have 5.1, so the amps are idle as is.
My thought is it is free. I have never read any disadvantage (aside from cost of amplification or wire - both of which I already have).
The advantages I have read about seem to be negligible, but I have never seen any comment on the thermal aspect of bi-amping.

I am no EE and hope those of you that know more about amps will comment on the accuracy of my reasoning (which may well be based on bogus ASSUMPTIONS):
1) If the same amount of power is used for bi-amping, is it advantageous to spread the heat out over the extra circuits, thereby reducing he highest temperature seen by any individual component.
2) Is heat output linear or logarithmic? IOW, if bi-amping reduces the load on the lower frequency amp by 5% would the resulting decrease in temperature (relative to room temperature) be 5% or would it be less or greater than 5%?

That is not Bi-Amping, but Bi-Wiring(effectively, not literally). You must have a separate power supply to truly Bi-Amp. But if the perceived benefit is there, then it is worth something.

 

[Steve81]

Interesting question, without deep thinking, I would say assuming the same load the overall heat produced should be roughly the same but heat dissipation will become more efficient though it won't be significant. You get to use the heat sinks of the other two amp channels.

One other thought: since Class A/B amps tend to be more efficient as output rises, you might be generating more overall heat by splitting the load, albeit with a more efficient distribution for dissipation.

 

[Irvrobinson]

That is not Bi-Amping, but Bi-Wiring(effectively, not literally). You must have a separate power supply to truly Bi-Amp. But if the perceived benefit is there, then it is worth something.

Incorrect on both counts. What KEW is proposing is bi-amping. Sharing a transformer and caps do not invalidate the separateness of amplifier channels. It's all a matter of how much capacity the power supply has.

 

[PENG]

One other thought: since Class A/B amps tend to be more efficient as output rises, you might be generating more overall heat by splitting the load, albeit with a more efficient distribution for dissipation.

Yes, that's why I said the overall heat produced would be roughly the same, more specifically like you said the biamp version would generate slightly more. That would offset the effect of more efficient heat dissipation so the overall difference may not be significant. Following this line of logic, if you push the amps closer to their limits then the biamp version would have better heat management overall especially in 3 way speakers where the mid range driver is typically grouped with the tweeter. Without studying any specific cases, I would call all of these just educated speculation.

 

[Irvrobinson]

One other thought: since Class A/B amps tend to be more efficient as output rises, you might be generating more overall heat by splitting the load, albeit with a more efficient distribution for dissipation.

I agree with Steve. Class AB amps have a bias current that keeps the output devices in Class A mode up to a certain voltage. In most receivers that voltage is quite low, but you'd have twice as much current dissipated.

 

[PENG]

I agree with Steve. Class AB amps have a bias current that keeps the output devices in Class A mode up to a certain voltage. In most receivers that voltage is quite low, but you'd have twice as much current dissipated.

I also agreed with Steve's point. I just felt the efficiency part is not that significant either way, when you consider his 7.1 receiver has the idling amps on all the time with bias current flowing wether the amps are used or not. On top of that, you have all the overheads found on typical AVRs such as HDMI board, processors, DACS, preamps etc. Some AVRs such as the Denon 4520 has a preamp mode but apparently that does not even turn off the unused amps. My AV8801 gets quite hot without any input signals. I do think the additional heat sinks would result in marginal improvement on the thermal side of thing and only if the amps are driven hard.

 

[Irvrobinson]

My AV8801 gets quite hot without any input signals. I do think the additional heat sinks would result in marginal improvement on the thermal side of thing and only if the amps are driven hard.

Have you tried it without any loads connected? On my ATI 3005 I find if there isn't a load connected to a given channel it doesn't dissipate heat.

 

[TheWarrior]

Incorrect on both counts. What KEW is proposing is bi-amping. Sharing a transformer and caps do not invalidate the separateness of amplifier channels. It's all a matter of how much capacity the power supply has.

I stand corrected... I recalled learning that passive bi-amping was only effective when using mono blocks, for example.

To that end, have you tried it yet KEW? Any improvements?

 

[AcuDefTechGuy]

You are using a quiet PC fan for your AVR, right?

 

[KEW]

I stand corrected... I recalled learning that passive bi-amping was only effective when using mono blocks, for example.

To that end, have you tried it yet KEW? Any improvements?

I have an infrared thermometer somewhere. Whenever I stumble across it, it will remind me I want to experiment with this.

As far as SQ (if that is what you are asking about) I am convinced that any differences would be far too minuscule for me to detect them - especially since there will be a substantial delay between hearing one way vs the other.

 

[PENG]

Have you tried it without any loads connected? On my ATI 3005 I find if there isn't a load connected to a given channel it doesn't dissipate heat.

Even with the AVRs, AVPs, Preamps or power amps just sitting there with power on they still get quite warm. Only when on standby mode they would stay cool. You ATI must have some power saving feature on it that mine don't. Still, I can't understand why with or without a load, assuming you meant the speakers connected should make any difference. If there is no input to the power amp, the output voltage would measure just a few mV.

 

[Irvrobinson]

Even with the AVRs, AVPs, Preamps or power amps just sitting there with power on they still get quite warm. Only when on standby mode they would stay cool. You ATI must have some power saving feature on it that mine don't. Still, I can't understand why with or without a load, assuming you meant the speakers connected should make any difference. If there is no input to the power amp, the output voltage would measure just a few mV.

I meant with inputs and speakers disconnected. I agree that preamps and prepros get hot just sitting there cooking with nothing connected. (The Benchmark Media preamp I use runs especially warm, as does the Comcast set-top box.) With power amps it has dawned on me that with the ATI this is the first time I've had unused amplifier channels, at least that I paid attention to. (I did have a Sony five-channel AVR of which I used only two channels, but I never paid any attention to its heat output.) I've switched which two channels of the ATI that I connect occasionally, and sure enough the heated positions in the chassis move around. I also suspect that a lack of input or output loads are detected and the channel circuitry is powered down after turn-on. The ATI is huge, and its internal architecture makes this effect obvious and easy to notice; it may not be so easily detectable on an AVR.

 

[highfigh]

As long as current is conducted by the output devices in the formerly unused channels, heat will be wasted. How much depends on the bias current and input level to each channel (since input determines output level). If the AVR's efficiency was known, it would be possible to calculate the theoretical heat loss by measuring the current demand, line voltage and power output because 1 Watt is roughly 3.4 BTU/hour. The difference between power in and power out is the wasted heat.

Stereo making your place too hot during Summer?

 

[Irvrobinson]

As long as current is conducted by the output devices in the formerly unused channels, heat will be wasted. How much depends on the bias current and input level to each channel (since input determines output level). If the AVR's efficiency was known, it would be possible to calculate the theoretical heat loss by measuring the current demand, line voltage and power output because 1 Watt is roughly 3.4 BTU/hour. The difference between power in and power out is the wasted heat.

I think both PENG and I can safely say we know this. The question is, are modern amp designers turning off unused channels, where unused is defined as open input and/or output circuits? (Not just low signal levels on complete circuits.)

In the ASIC world it is common to power down unused portions of chips, or even vary clock speeds to save power. There is an Energy Star proposal to affect audio amplifiers, but I have no idea if anyone implements the recommendations.

https://www.energystar.gov/ia/partners/prod_development/revisions/downloads/audio_video/AV_Draft2_V2_Audio_Amplification_Proposal.pdf

 

[highfigh]

I think both PENG and I can safely say we know this. The question is, are modern amp designers turning off unused channels, where unused is defined as open input and/or output circuits? (Not just low signal levels on complete circuits.)

In the ASIC world it is common to power down unused portions of chips, or even vary clock speeds to save power. There is an Energy Star proposal to affect audio amplifiers, but I have no idea if anyone implements the recommendations.

https://www.energystar.gov/ia/partners/prod_development/revisions/downloads/audio_video/AV_Draft2_V2_Audio_Amplification_Proposal.pdf

It would make sense to shut down the B+ to unused channels but I don't know that it's done. From the manufacturer's view, it would probably just end up being another way for them to cheapen the power supply.