Power amplifier with remote sensing (speaker cable compensation)

[pma]

The Hypex NC500 OEM by the same designer also has remote sensing and, if memory serves, it's documentation states that sensing at the speaker just doesn't work.

Exactly. FBH and FBC must be connected to OUTH and OUTC not more than a few cm away from the amplifier.

 

[restorer-john]

I'm with @SIY here, no net benefit. But it's a nice engineering challenge notably when it has to be bullet-proof (sense wires not attached, partly attached, reversed polarity, etc).

Yep, the Sigma Drive was a total fiasco for Kenwood. They killed it after less than 2 years.

Their LO (LO-5/7m etc) series from the late 1970s was, all in all, a better concept and pretty much the first to push ultra linear, high powered monoblocks to be placed right at the speakers with very short speaker cables.

 

[Colin James Wonfor]

I use Remote sense on my big PSU's and on the M100TS Power Amp no problem with any cable and no screaming Parrot Sick Oscillations (yes spelling naff). But the Bulk caps are in each channel 6 off 47,000uF and the transformers are 1KVA each dual mono design, with 16 power transistor matched and in the output only tiny ones.

 

[MC_RME]

Why not check the theoretical effect in a practical real-world approach? Add 1 or 2 Ohms 50 Watt resistors (very cheap to get) at the speaker output and try to hear the difference. Then you know if all that trouble is worth it...

 

[antcollinet]

Every stand of the wire is insulated individually with a thin layer of varnish like in a transformator. This gives you lower impedance at high frequencies due to reduced skinn effekt.
Last year I have been developing a small inductive heater running at 100kHz and we got by far the lowest impedance using HF litz wire. And 100kHz is not that for of the audio band. Using some leftover wire was rather cost effictive

Capacitance and inductance is not specifed, but I could measure that - well not today, as the measurement equipment is at my workplace and I'm at home.

Skin depth at 20KHz is around 0.5mm.

My (quite capable) 4mm sq speaker cable (aprox. AWG 11 to 12), has a diameter of 2.25mm.

If you take the simplification that skin effect can be approximated as a tube with wall thickness equal to the skin depth, then in my cables there is a 1.25mm diameter centre core that isn't being used for 20KHz. However, this only reduces the area to about 2.6mmsq. (Effectively the next wire size down).

Since this results in a resistance increase from around 0.005ohm/m to 0.008ohm/m Assuming 8ohm speakers and 3m cables, this represents an additional 0.009db (was 0.006 before edit) reduction at the speaker at 20KHz compared with low frequencies.

Given that this effect only just starts above 4.3KHz (with my 4mmsq cable), and that musical power spectral density falls off dramatically with frequency (power many 10s of dBs down at 20KHz - then I'm gonna suggest your Litz wire impedance improvement is inaudible.


This is audio science - see how it works? You can do a little basic maths, and work out if what you are doing makes any sense.

EDITED: to correct additional dB reduction - see above.

 

[Stefan1024]

BTW what exactly is the problem you're trying to solve? I get that you want to reduce the output impedance further, but why?? The ahb2 is likely transparent without further modification, particularly since you're supposedly running low DCR and low inductance cable.

I don't have a problem with the AHB2, just doing some thougth experiment on possible improvement.

I'm biased (no pun intended) toward Class D for reasons of efficiency but the AHB2 is a very good amp. Others have pointed out the very real dangers of stability issues so stick with what you have.

We can argue the improvements the litz wire makes will be inaudible but, it seems the litz wire did not cost you much. It's a great conversation piece.

For best overall efficiency, you need high efficient speakers. With my 99dB/W speakers I'm basically allways below 1 Watt and only the idle power consumption of the amplifier is important. And the AHB2 is rather good in this regard, ca. 12 watts.
That said, I was runnign a Pass XA-25 at one point for about a month, but didn't like the sound and sold it on. That monster had about 240 watts idle power consumption...

Pardon me if I am missing the big picture, but isn't the loudspeaker impedance the big fish here?
[snip]

Yes, I rased exactly this question in my original post to have a discussion if it is reasonable to reduce the output impedance. And the answer is overwelmingly that it's not really beneficial bejond a certain point.

That would be a good reason to use it.
Since it was developed do you have info on the cables tested ?
What length are you talking about ?

100kHz is over 2 octaves above the audible range (LF radio frequency). How much effect will there be at a more important 10kHz ?

Actually we basically allways worked with litz wire of diffent cress section and diameter of individual wires. Only in the very beginning a made a coil using a speaker cable and it was about 3 times worse than the litz wire. To be fair, the insulation of ca. 0.5mm on the speaker wire increased the diameter unnessessarily and prevented thight wrapping of the coil, what gave it a large disadvantage.
Total length of the wire for the coil was 1.5 meter.

Source: https://www.elektrisola.com/en-us/Litz-Wire/Info#dimensions
There is also much more information about litz wires.

Also even when buying new, the litz wire costs about 80 Euro per killogram. At my rather high 8mm^2 thats ca. 16 Euro per meter and scaless down linearly with cross section.

The Hypex NC500 OEM by the same designer also has remote sensing and, if memory serves, it's documentation states that sensing at the speaker just doesn't work.

Thank you, good to know.

 

[Stefan1024]

Exactly [Snip]

Good to know, thank you.

Why not check the theoretical effect in a practical real-world approach? Add 1 or 2 Ohms 50 Watt resistors (very cheap to get) at the speaker output and try to hear the difference. Then you know if all that trouble is worth it...

Not sure how this simulates a lower or higher output impedance. For pure resisistive loads the output impedance is completly irrelevant as long as it is static (despite form a sligth decrease in signal level).

[snip]

As your reach the skinn depth limit within the audio band, one can argue that the cable acts as a sligth low pass filter within this band. So I like to avoid that if I can and there are no mayor drawbacks, e.g. price.
Is it audiable? Most likely not. As you rigthly pointed out the effect is very small.

 

[antcollinet]

As your reach the skinn depth limit within the audio band, one can argue that the cable acts as a sligth low pass filter within this band. So I like to avoid that if I can and there are no mayor drawbacks, e.g. price.
Is it audiable? Most likely not. As you rigthly pointed out the effect is very small.

A low pass filter with 0.009 db of attenuation at 20KHz???

Pretty much any piece of kit is gonna have an effect that small if you so much as look at it funny.

 

[preload]

I don't have a problem with the AHB2, just doing some thougth experiment on possible improvement.

I see. So you're hoping to possibly improve a measurement even though there isn't really a problem? Sounds like a worthwhile exercise.

 

[RMD]

The output impedance of valve amplifiers is large as compared to solid state amplifiers. In fact its large enought to be comparable to the loudspeaker impedance. So with valve amplifiers, the effect of impedance variation of the loudspeaker becomes small as compared to the output impedance of the amplifier itself. This makes power delivered into the loudspeaker more proportional to input signal than is the case with solid state amplifiers. This is one of the reasons why treble is silky smooth and mids are not shouty with valve amplifiers and also the reason why its easier to drive single full range speaker with valve amplifiers than solid state amplifiers.
I am new to this forum and so I am unaware if the topic of measuring current through the loudspeaker was discussed. But if we do that, then the load current feedback can be used make output power proportional to input signal only and in which case speaker cable impedance, connector joint resistance can all be taken out of the equation.
The topic of load current feedback is not new and has been taken up by a lot of big names in the industry and irrespective of their findings, practically, the load current feedback improves the tonal quality of your system tremendously making low cost driver sound like more expensive ones. If done correctly, semiconductor amplifiers can be given valve amplifier tonal signature. Parameters like cable impedance and damping factor, then, don't make much sense in.

 

[kchap]

The output impedance of valve amplifiers is large as compared to solid state amplifiers. In fact its large enought to be comparable to the loudspeaker impedance. So with valve amplifiers, the effect of impedance variation of the loudspeaker becomes small as compared to the output impedance of the amplifier itself. This makes power delivered into the loudspeaker more proportional to input signal than is the case with solid state amplifiers. This is one of the reasons why treble is silky smooth and mids are not shouty with valve amplifiers and also the reason why its easier to drive single full range speaker with valve amplifiers than solid state amplifiers.

My understanding is that speakers are designed on the assumption that they are being driven by a constant voltage source. The ideal signal source wants to have a zero source impedance. Your silky smooth appears to be a tapering frequency response due a high source impedance.

 

[RMD]

My understanding is that speakers are designed on the assumption that they are being driven by a constant voltage source. The ideal signal source wants to have a zero source impedance.

Agreed. That said, we still need to account for the impedance curve of the speaker, unless of-course we can come up with a perfectly designed constant impedance, passive, crossover network since the constant voltage power amplifier is blind to impedance variations of the load.

 

[RMD]

My understanding is that speakers are designed on the assumption that they are being driven by a constant voltage source.

To drive my point, kindly take a look at the impedance curve of FaitalPro 3in (attached) and 4in full-range drivers. Its almost constant impedance (Flat). These speaker, I believe, were designed and tweaked for constant voltage amplifiers and they sound the same or almost the same with an amplifier with load current feedback
Normally

 

[antcollinet]

The output impedance of valve amplifiers is large as compared to solid state amplifiers. In fact its large enought to be comparable to the loudspeaker impedance. So with valve amplifiers, the effect of impedance variation of the loudspeaker becomes small as compared to the output impedance of the amplifier itself. This makes power delivered into the loudspeaker more proportional to input signal than is the case with solid state amplifiers.

This is the exact opposite of reality.

High output impedance means as the speaker impedance varies through the frequency range, so the output of the amplifier will vary as the output voltage is divided between the output impedance, and speaker impedance.

High output impedance will create a very non flat FR at the output of the amp.

 

[RMD]

This is the exact opposite of reality.

High output impedance means as the speaker impedance varies through the frequency range, so the output of the amplifier will vary as the output voltage is divided between the output impedance, and speaker impedance.

High output impedance will create a very non flat FR at the output of the amp.

Agreed.
However, its not the voltage but current that drives the speaker. So its worth taking a look at the power response instead of FR (frequency response).

In the attachment, in table 1, the output voltage of the amplifier is the voltage appearing across the speaker. Power delivered into the speaker is calculated as speaker impedance varies from, say, 5ohms to 12ohms which is not uncommon for an 8ohm speaker. It is seen that power decreases as speaker impedance increases. The variation in output power is from 0.2W (@5ohms) to 0.083W (@12ohms)

Now consider, say, a 1ohm resistor in series with the loudspeaker. This arrangement will create variable voltage across the speaker as the speaker impedance varies, just like what you said in your post above.
Attached table 2 calculates power delivered into the loudspeaker with this new arrangement for the same speaker impedance variation from 5ohms to 12ohms. The variation in output power is from 0.1388W (@5ohms) to 0.071W (@12ohms)

To make interpreting the results easy, allow me to increase the volume (and hence the amplifier output voltage from 1V to 1.2V) to compensate for the loss in the series resistor . This is shown in table 3. The variation in output power is from 0.2W (@5ohms) to 0.102W (@12ohms)

Variation in power delivered into the loudspeaker has reduced as compared to a pure voltage amplifier with zero output impedance. This is the point I was driving: To make power delivered into the speaker depend on input signal and input signal only and not on loudspeaker impedance variation. This should be the goal and this does not happen in reality.

 

[SIY]

To make power delivered into the speaker depend on input signal and input signal only and not on loudspeaker impedance variation. This should be the goal...

The data behind this assertion are very skimpy and ambiguous. If it takes less power at a constant voltage for a given SPL at resonance, then who cares? Flat (or at least, controlled) SPL is the goal.
]
There's a guy in NZ (I think) who makes a very big noise about having flat impedance so the speaker draws constant power with frequency- of course, the extra power to make up for it is merely dissipated in passive components, contributing nothing to the acoustic output. And indeed, if you remove the impedance compensation from his speakers and use an engineered amplifier (i.e., low source impedance), the frequency response changes not a bit.

 

[RMD]

Flat (or at least, controlled) SPL is the goal.

Agreed!

 

[audio2design]

@Stefan1024. What do you think the inductive reactance is of your cables? I expect by 3-4Khz it exceeds DC resistance possibly at a lower frequency, but it's so far below speaker impedance and voice coil resistance to be inconsequential.

 

[Head_Unit]

Why worry about 0.1dB variances in the FR caused by varying speaker impedances ?

Have you asked Purify if remote sensing is a good idea and what length they would specifiy as max ?

Well, why not? But yes, a hindsight-obvious great suggestion to ask the manufacturer what they think. Kenwood seems to have buried Sigma Drive, though I have no idea if due to technical problems or because the concept didn't really successfully get across to customers.

 

[Head_Unit]

The Hypex NC500 OEM by the same designer also has remote sensing and, if memory serves, it's documentation states that sensing at the speaker just doesn't work.

But...but...I really WANT to believe!!!