I think this is the thread (link is to Ray's example)
https://www.audiosciencereview.com/...-needed-power-for-a-speaker.13779/post-420206
-
WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!
Bi-amping 101
- Thread starter DonH56
- Start date
https://www.audiosciencereview.com/...-needed-power-for-a-speaker.13779/post-420206
Bingo.
Actually it was @Blumlein 88 who discovered this in posting #6. But I think the higher peaks in the high pass filtered signals are due to the very low xover frequency.
Ross Naheedy
New Member
- Joined
- Jul 5, 2021
- Messages
- 3
- Likes
- 1
Thanks for the very informative posts Don. Regarding your last paragraph in the first post:
I can think of one good case to passive bi-amp. Since the power needed for the highs is nowhere near the power needed for the lows, it's very feasible to stick with a class A/B amp for the lows and upgrade to a class A for the highs. That is exactly what I'm planning on doing with my NHT 3.3 speakers. These speakers are very, very power hungry because of the 12" subs they house. Plus they only have a sensitivity of 87db. Currently they're being driven by an ATI-1502, which is a beast of an amplifier in my opinion and is doing a great job, but I suspect most of my listening is at class B level.
Prior to considering bi-amping, I had been looking into doing a pair DIY Pass Aleph 2, but I can't justify a 600W heater in my room. So I am trying to figure out if I can bias the Aleph 2 differently so that I end up with 50W/amp and a lot less heat, or if I go with another class A amp that'll give me 50W/channel. I think the end result may be worth the effort.
I can think of one good case to passive bi-amp. Since the power needed for the highs is nowhere near the power needed for the lows, it's very feasible to stick with a class A/B amp for the lows and upgrade to a class A for the highs. That is exactly what I'm planning on doing with my NHT 3.3 speakers. These speakers are very, very power hungry because of the 12" subs they house. Plus they only have a sensitivity of 87db. Currently they're being driven by an ATI-1502, which is a beast of an amplifier in my opinion and is doing a great job, but I suspect most of my listening is at class B level.
Prior to considering bi-amping, I had been looking into doing a pair DIY Pass Aleph 2, but I can't justify a 600W heater in my room. So I am trying to figure out if I can bias the Aleph 2 differently so that I end up with 50W/amp and a lot less heat, or if I go with another class A amp that'll give me 50W/channel. I think the end result may be worth the effort.
That is running under the assumption that a Class A amplifier will produce better sound quality than a A/B amp, or a modern high efficiency Class D amp. These are all misconceptions, a properly designed amp in any of the classes can be fully transparent to the source with the only basic difference being the amount of AC power they use..
No, because with passive bi-amping, the output voltage available from each amplifier has to be identical, as does the gain. Each amplifier gets the same input voltage, so the output voltage must be the same, or the lower output one will clip first. One way to destroy tweeters.
S.
OP
- Thread Starter
- #26
Not with passive bi-amping. See posts just above, and look at the article again. It does not matter that the treble amp actually needs (delivers) much less power; the problem is that the voltage swing is the same for both amplifiers, so you cannot use a low-powered treble amp in a passively-bi-amped system. That is a major drawback IMO. You need a crossover that only sends the desired frequency range to the amp.
Ross Naheedy
New Member
- Joined
- Jul 5, 2021
- Messages
- 3
- Likes
- 1
That's basically just matching the levels between the two amps, and that's something of concern for sure, but easily overcome with separate passive volume controls for each amp. The NHT 3.3s are very finicky in regard to their placement and room response to bass, so it wouldn't be out of the realm of possibilities to attenuate the lower signal more if the physical placement of the speakers don't get the desired bass output.
MakeMineVinyl
Major Contributor
The passive crossovers in the speaker are designed around the assumption that the same signal voltage will be fed to the input of the HF and LF crossover sections, so altering the gain of one of the amps will upset the balance that the speaker manufacturer designed in. It is far, far, far, far better to do the bandpass splitting before the respective amplifiers (aka active crossovers) and eliminating passive crossovers from the speakers entirely.
Passive crossovers somehow became the norm in the early days of HiFi and in my opinion are one of the stupidest possible ways to approach the problem. I'm frankly surprised that 'high end' audio never latched onto active crossovers since it would mean them selling twice as many amplifiers, speaker cables and interconnects - in addition to an insanely overpriced crossover box. Morons.
Ross Naheedy
New Member
- Joined
- Jul 5, 2021
- Messages
- 3
- Likes
- 1
I fail to see that being an issue. When you decouple the highs and lows on the 3.3 by removing the the plate that joins them at the connector on the back of the speaker, you're essentially left with 2 satellites and 2 subwoofers, each retaining their own passive crossover. Powering them would be no different than powering satellites with one amp and having a powered subwoofer. The only issue I see beside the one sergeauckland mentioned (signal matching) would be possible phase difference between the bass and treble, but reversing the cables on either would resolve that issue. The following is from the manual for the NHT 3.3:
OP
- Thread Starter
- #30
Please look at the first post again and see the pictures and explanation under "Passive Bi-amplification". The signal to the sub and the signals to the speaker terminals are not the same. The difference is that there is a crossover inside the AVR before the signal is sent to the subwoofer's amplifier so the subwoofer only sees a low-pass filtered signal (only the lows are sent to the sub). With passive bi-amping as implemented by an AVR there is no crossover to the main speaker channels' high and low sections. The same signal is sent to both (high and low) amplifiers, and that signal includes everything (bass and treble). Power is the product of voltage and current. Even if no power were delivered to the tweeter terminals, the voltage is exactly the same as sent to the bass terminals, and thus voltage clipping will occur at the same point no matter how much (or little) power is delivered. That is the fundamental problem with passive bi-amping; both amps provide the same (voltage) signal, and thus both will clip at the same point (voltage), even though they are not delivering the same power.
If you have or add a crossover before the power amplifiers to separate the high and low signals to the amp then you have conventional ("active') bi-amping and the amps need not be the same. That is what is shown in the "Conventional Bi-amping" section of the article.
Last edited:
There are exceptions:Passive crossovers somehow became the norm in the early days of HiFi and in my opinion are one of the stupidest possible ways to approach the problem. I'm frankly surprised that 'high end' audio never latched onto active crossovers since it would mean them selling twice as many amplifiers, speaker cables and interconnects - in addition to an insanely overpriced crossover box. Morons.
- In the 1990s Linn offered the Kaber in passive and active version, with the idea to activate the passive version later by getting more power amps with additional matching crossover modules. The Isobarik speaker may have been always active.
- Backes & Müller produces active speakers since more than 40 years now. Prices are highend.
MakeMineVinyl
Major Contributor
I was referring to the conventional idea of 'active crossovers' and in that context the crossover is an outboard box inserted before the power amplifiers. Of course active speakers use active crossovers (I sure as hell hope so!), but the wider acceptance of active speakers is a more recent development although I'm sure somebody could be found who did this much earlier - I had Mackie active speakers in the late 1990s. As early as the 1950s, passive 'active crossovers' were being inserted before the power amplifiers which were then directly connected to the speaker drivers, so this is not a new concept. This was limited to hobbyists, unfortunately, and did not become the norm.
OP
- Thread Starter
- #33
Passive bi-amping, when I first read about it a decade or so ago, confused the heck out of me because what people were telling me made no sense. To me, passive bi-amping meant a passive line-level crossover (RC and/or RLC) before the power amps. Active bi-amping meant the crossover used active devices (tubes or transistors). When I was finally set straight, it still made no sense, as there was no crossover as I understood it save the one already in the speakers, and I could not understand the benefit. The only benefits I could see were a reduction in current-related and thermal distortion, via isolating back-emf and intermodulating speaker (driver) currents by separating the amplifiers, which meant the single amplifier and wiring was not up to the task. The benefits, then and now, seemed so slight as to not be worth the effort and additional wasted power.
Last edited:
I had 4 mono blocks., each 2 just behind the speakers for years and switched back to a single one per speaker.
Cannot differ the pleasure other than my brain wishes me to believe in
And the savings on your electric bill.
I really like the way this material is presented. A bit of a crash course but not entirely impenetrable for a layman/hobbyist such as myself. It seems like a wasted opportunity that this thread isn't more active, but I'm happy to take advantage.
So, given the best way to find the right answer to a Q in the 21st century is to post the wrong answer on the internet, let me know if my proposed bi-amp scenario (based on equipment I own) accurately reflects your recommendation?
Monolith 11ch amp (3x200W, 8x100W), Kef R3's, Denon 8500.
Kef lists the crossover points at 400Hz and 2.9KHz, but both Amir and Erin's reviews show response far exceeding that. How much is usable/relevant during normal operation, I'm not certain.
To give myself wiggle room, I'd get a 500Hz LPF,a 500Hz HPF (both @12dB/octave, HarrisonLabs) and a Y-cable for each L/R/C channel. The LPF go in-line toward the 200W channels and the HPF on the 100W channels. I would also use a standard 80Hz crossover In The Denon.
Do I have that right?
if so, lingering questions in my mind:
will the Denon pre-outs push more V to achieve same level as if I did NOT use y-adapter?
typically it's accepted that doubling W = 3dB of usable volume. I'm only increasing by 50%, but I'm "steering" it toward mid/tweeter so would I still only expect 1.5dB of gain altogether?
the most common argument against AVR bi-amping is "there's only one power supply.". You've gone into more detail here which is helpful, but in my scenario the Monolith has 2 transformers. Does that matter at all? Would I be better served by trying to split the load across the 2 transformers?
ARe the 500Hz FMODs the right choice given there's a lot more extension from the individual drivers?
thank you again for the write-up
OP
- Thread Starter
- #37
Thanks.
A lot to cover... I suggest opening a thread to deal with your specific situation. The result of this whole effort would be to gain a little headroom in each of the amplifier channels, and to isolate the high and low sections of your speakers. Whether that provides any audible benefits I could not say, and of course the passive crossovers within your speakers would remain with their impact on performance (good or bad). From here on the answers are more guesses about your situation since there are too many unknowns about the various parameters (e.g. all the component and speaker impedances, how loudly you listen, room effects, etc.) Don't bother trying to find all of that as this is not meant to be a tutorial for preferences. I personally would not bother to bi-amp in this situation as it seems like a lot of effort for something unlikely to provide audible improvement, at least in a blind test.
Short answers: The connections seem OK and 500 Hz is a reasonable choice for the filter corners given your experiment. The Denon's preamp outputs will probably be unaffected by the splitter and I would not be concerned. The maximum power gain available in dB will be 10*log10(Pamp/Pavr) for you amps and AVR, and different for each amp since they have different power ratings. I doubt the two transformers is a significant factor. The additional filters will change the effective crossover response and you must use the AVR's room correction program to compensate, if it can, otherwise you will likely have severe frequency response problems.
IME most people who try even passive bi-amping report improvement, and most (perhaps all) would be unable to prove those improvements in a controlled blind listening test. The theoretical benefits are small but as always overridden by our innate bias (something we cannot do anything about).
A 500 Hz crossover (HPF+LPF) is barely above the speaker's crossover so will contribute significantly to the response, both amplitude and phase. Hopefully the Denon's room correction program will compensate for that so the speaker still sounds good in the end. It is probably as good a choice as any. True bi-amping eliminates the internal speaker crossovers but that is not practical for most of us, and more pain than it is worth for a home system.
One thing to consider is how loud each frequency band may be and how much power each will require, bearing in mind you are still using a sub (inferred from your 80 Hz crossover for LFE and bass management). 500 Hz is still in the midrange, so chances are the LF amp will benefit from greater power than the HF amp, though the HF amp covers a much wider frequency range.
How much the Y-adapter, and filters, decrease the Denon's output depends upon its preamp output impedance, the impedance of the filters, and the input input impedance of the amplifier. The usual assumption is that the preamp outs are low-impedance and amp inputs high-impedance so a Y splitter does not significantly decrease the signal level. I don't know the impedances of any of those components but suspect the answer is that the Y splitter does not change things enough to worry about, and in any event your Denon+amp combination should be able to nudge the volume a hair to compensate.
Doubling the power is a 3 dB increase and corresponding increase in SPL. You are offloading the power demands from the AVR so it should run cooler and draw less power from the wall outlet, and perhaps the external amplifier provides other benefits like lower output impedance and the ability to handle longer peak (transient) demands due to more robust power supply and output stages. Whether or not the extra power improves performance I could not say (too many unknown variables); perception bias says you will hear an improvement even if it is not "real".
The power supply is one of many factors, and whether a single transformer or multiple transformers is the better choice depends upon their capability. Personally I wouldn't worry about it in this scenario. The overall power supply capacity is probably much greater for the external amp.
HTH - Don
Last edited:
Sorry Don. I don't think @tifune's idea is a good one.
We need to careful about correct summing at the cross-over region. Linkwitz-Riley was invented for that reason. Below is my simple simulation of the (electrical responses of the) filters.
The top graphs are the responses using the standard LR4 with cross-over at 400 Hz. The bottom graphs are the responses when both the low-pass and high-pass filters are cascaded with 2nd order Butterworth filters with corner frequencies at 500 Hz.
The phase mismatch from adding the Butterworth filters resulted in a deep suck out.
We need to careful about correct summing at the cross-over region. Linkwitz-Riley was invented for that reason. Below is my simple simulation of the (electrical responses of the) filters.
The top graphs are the responses using the standard LR4 with cross-over at 400 Hz. The bottom graphs are the responses when both the low-pass and high-pass filters are cascaded with 2nd order Butterworth filters with corner frequencies at 500 Hz.
The phase mismatch from adding the Butterworth filters resulted in a deep suck out.
bo_knows
Addicted to Fun and Learning
OP
- Thread Starter
- #40
I do not think so either, but he can try. The crossover will be completely mucked up unless the AVR compensates. I said that in my post but your pictures provide an excellent explanation of the "why" behind my statement. Basically it all depends on the AVR's room correction program (Audyssey, I think, for Denon?) to compensate the additional amplitude roll-off and phase change through the crossover region. I've no idea how well (or even if) that works and probably did not emphasize the problem enough in my reply. My experience is all with conventional bi-amping.
I added another line to help highlight the issue.
Last edited:
Similar threads
