Slight course correction. I was just able to get a screaming deal on a set of Infinity RS 1.5s: http://www.infinity-classics.de/technik/manuals/RS_1.5_technical_sheet.pdf My speaker friend and I are going to collaborate on a custom job. I'll build cabs and he'll handle electronics. They'll be upgraded based on a project he recently completed with some RSbs: Infinity RSb reincarnated So I'm not going to mod the AR-3as. This will be much better. They'll be set up for active bi-amping from the start. They have the 12" Watkins DVC woofers. And he's got a few pairs of higher-end EMITs on hand as well. The Phase Linear 400 should be arriving at the shop any time now. So I've got a real project shaping up here.
I think you are way better off with this choice. Building the speaker you want, with the components you choose. The other way was simply not going to be workable in any meaningful way. Now, you have the opportunity to create your own masterpiece. My friend set up a HT with five Altec Lansing A7's. He powers them with all Phase Linear amps. Do suggest you implement some output protection. These were good amps, high slew rate and all that. Most powerful amps for their time. But, they were very basic amps. They developed the nickname "Flame Linear" for a reason. If there is a failure in the output section, you will end up with 70V DC going directly to your speakers voice coils. There are videos on YouTube and information on the Internet.
The rebuild will include output protection and a few other safety/reliability oriented modifications. Definitely seems like those units earned that nickname!
Good to hear that you're going to keep the AR3's original. Also kudos to you that you're going to put a protection circuit in the PL and experiment with bi-amp and find out first hand for yourself. Having heard both the AR3a's and the Infinitys I'm thinking those 2 share almost nothing in common sonically other than they are both speakers. Have fun and good luck to you on your projects.
Now I'm off to wondering whether a system will be more energy efficient if the passive high-level crossovers at the speaker are replaced with passive low-level crossovers between the pre-amp and power amps. An often cited benefit of "active" bi-amping - or rather crossing the signal before the power amps instead of after - is the avoidance of significant heat-waste by passive speaker crossovers. A passive crossover at low level will obviously bleed off a lot less heat than a passive crossover at high-level. But it will also reduce the strength of the signal arriving at the power amps. Which means that the power amps will have to work harder to produce the same output voltage as before. So how does this actually work out? Is one approach more energy efficient than the other? In other words, does a low-level passive crossover bleed off a smaller percentage of the voltage running through it than a high-level passive crossover? (Lets assume all crossovers in question are well-designed and constructed with quality parts.) If a low-level crossover is more efficient than its high-level counterpart, then although the power amp would have to work harder with a passive crossover in front of it to produce the same output voltage, it still wouldn’t have to work as hard as it would with a passive crossover “behind” it at the speakers to produce the same decibels. The system would be more energy efficient, resulting in more amplifier headroom and lower energy consumption. I think what I'm saying is that it's not fair to simply say that the avoidance of a high-level passive crossover means that the amount of heat it would have bled off is the amount of energy saved. Maybe most of it is. But that depends on understanding relative energy efficiencies as mentioned above. I just don't know how to answer that question. But several of you probably have a direct understanding of this.
An active low-level crossover network, unless I'm missing something, achieves its purpose by adding to - not subtracting from - the signal. So the signal strength arriving at the power amplifier is a bit stronger. As an amplifier itself, the active crossover uses power and creates heat. But it also prevents the power amps from having to work quite as hard, which reduces heat and increases headroom. From an energy usage perspective, whether this produces a more efficient system depends on the amount of energy used by the active crossover to achieve a percent increase in signal strength vs that used by the power amplifiers to achieve the same. Either way, an active crossover ensures additional headroom, which - all else equal - is a good thing.
No, the power amp has a fixed gain. If it is producing the same output voltage into the same load, that is the same output power, so it isn't working any harder. The power amp connected directly to the driver instead of through a passive crossover will not have to work as hard for the same sound level since the losses are lower. Using passive or active crossovers before the power amp has no effect on the required amplifier power, the signal level is adjusted in the crossover to balance power delivery to all the drivers for the desired response. Same as using the main volume control, if you switch to a source that has a lower input level and have to turn up the volume control for the same sound level, the amplifier isn't working any harder.
I've no idea if the subjective performance of actives are grossly overestimated, as my anecdotal evidence of delighted customers and personal enjoyment of living with active systems in preference to passive systems suggest otherwise. All I have to say is that listening to music, either on a passive or active system and the 15 odd years spent demonstrating and installing mostly passive, but also many active systems, put me in a better position to make a judgement call on the performance of different components rather than the six years I spent studying electronics and working with test equipment. It was also massively more satisfying! I could say plenty more on the subject, but my opinion is totally irrelevant - people can of course either study data sheets or do as Neil Gader did back in 2017 - listen for themselves. I have never owned ATC speakers, but very much enjoyed both their passive and active speakers. @Soundgarden, here is some information about the benefits of active from ATC, who manufacturer their own drive units and their own passive and active crossovers. For anyone really interesting in comparing passive to active systems, the fact that ATC manufacturer both passive and active versions of some of their speakers, makes it is very easy to compare both affordable and very expensive power amps in a passive set up, compared to ATC's integrated active version.
Right, but I feel like I'm setting up a different hypothetical. Three cases, brutally simplistic: Case 1: Pre-Amp spit out 1V. Passive Crossover reduces signal to .5V. Amount of power required to boost signal to 10V = X Case 2: Pre-Amp spits out 1V. No passive Crossover. Amount of power required to boost signal to 10V < X. Case 3: Pre-Amp spits out 1V. Active crossover increases signal to 1.5V. Amount of power required to boost signal to 10V? Less than in case 2. Right? Maybe that's too simple.
An active line level will have amplification circuitry to bring the output equal to the input. The circuitry may, or may not, have a sound signature. It has a power supply so it is inherently more complex than a passive line level. A passive line level will have an insertion loss. Years ago I owned a pair of Magneplanar Tympani 1Ds. It came with a passive line level as well as built in passive crossovers. Initially, I drove them with 2 Crown DC300As, strapped in mono, using the built in passives. I scored a McIntosh 2300 amp at an auction. I then used the PLLXO strapping the Crowns for the tweeter panel, and using each channel of the Mac for the bass panels. The internal passive were rated as an 8 ohm load. Using the PLLXO, each panel was rated as a 4 ohm load, thus drawing more power from the amps. This more than made up for the 3db insertion loss of the PLLXO. Using the rated output of the amps into 4 ohm loads, I had a 2.1kw system. That system was glorious!
"It was at this point in the thread (Post #237) that the other members of the Steve Hoffman Forum collectively began to wonder whether @Soundgarden, despite his tenacious attitude and thirst for knowledge, was truly over his skis."
I really wouldn't worry about heat from crossovers or overworking power amps. Both of these components are designed for this. You will loose some of the energy using passive crossovers at line level or speaker level. In either case, you compensate for this loss by adjusting the volume of the preamp so that you are sending more signal to the power amp. You might be getting to the point where you are over thinking this a bit? Any passive component in the signal path will incur a loss. That does not make it any big deal. Lose a bit of signal, turn up power, problem solved. In normal systems, these are small losses and are easily compensated for. Where these losses become critical is where you are working with high sensitivity speakers and flea watt amps. Non-horn speakers that achieve high sensitivity, like Zu's, involve using a single full range driver and no crossover network. Mentioning that some have dual 10" drivers. What is sometimes referred to as a crossover, is simply a capacitor in series with the tweeter, to allow frequencies over 10-kHz. to pass through to the tweeter, while blocking out lower frequencies. This design relies on the 10" full range driver rolling off by 10-kHz. This means that no actual "crossover" is necessary for the single driver. Take the example of a speaker level passive crossover. Is there heat being generated due to signal loss, yes. But these losses are minimal because the crossover, unlike the speaker driver unit, is not being tasked to convert electrical energy into mechanical energy. The crossover in theory is not there to use the available signal, just to split it into different frequency bands and pass them along to the drivers. This is what a crossover would do, in a perfect world. Consider that you are cranking twenty Watts into your speaker. The drivers themselves are very inefficient. Only about 10% of that power is converted into sound and the other 90% is dissipated as heat. Compared to the drivers, the crossover loss is fairly minimal. The main problem with speaker passive crossovers are that they usually are of poor design and built with inferior parts. Crossovers tend not to be on anyone's radar. Consider all of the threads in the audio forum. Audiophiles argue over things like cables, speaker placement and many other things. Just about nobody concerns themselves with the crossovers. As a comment, you are to be congratulated for paying attention to this very important of audio reproduction.
Well, it's just that in your example, the amplifier has a gain of 10, so 1V at input = 10V at the speakers. The gain is fixed at 10, it is always 10, if the input is only .5V then it will only output 5V to the speakers, so if you want to listen at the equivalent of 10V to the speakers, you turn up the volume control to get 1V at the amp input. The amp isn't doing any extra work, and neither is the preamp, which has a fixed gain as well, typically around 5x. The volume control just throws away most of the input signal until you get to the listening level you want. At full volume it has the same gain as when it is turned down all the way, but you aren't throwing away any of the signal.
What @Davey is saying is spot on. When a power amplifier is designed, the amount of gain that amplifier has is fixed and is determined by the design of the amp. The gain that the amp has when it leaves the factory, is the gain that it will forever have. Most power amps are just large boxes with an off/on switch. How loud they play is determined by the level of the voltage going into the power amp. This is determined by the volume control on the preamp. Some pro-sound power amps do have a volume control. What this control is actually doing is turning down the volume of the amp by limiting the input signal. It is not actually turning "up" the amplifier.
so, am i to believe all the rhetoric about how passive crossovers should not be used in line behind an active because of these loses. thats what i get from this thread. but iono much.
If I am understanding your post, you are talking about placing a passive line level crossover directly before an active line level passover? If that is the case, what purpose would be served placing two crossovers in line with each other? Does not make sense.
If you don't go full complementary and new White Oak pcb's you will be missing out on a huge gap from stock/upgraded original pcb in these units. Hope this is in your plan.
Revised but still brutally simplistic scenarios: Case 1: Pre-amp spits out 1V. Active crossover outputs 1V. Power amp boosts 1V input signal to 10V. No passive crossover. Case 2: Pre-amp spit out 2V. Low-level passive crossover reduces signal to 1V. Power amp boosts 1V input signal to 10V. No high-level passive crossover. Case 3: Pre-amp spits out 2V. No low-level passive crossover. Power amp boosts 2V input signal to 20V. High-level passive crossover reduces signal to 10V. These scenarios assume a power amp with a fixed gain of 10. They assume the passive crossovers burn off half the voltage as heat. (I know that's high but it happens) They assume the same of a low-level crossover. (I have ZERO IDEA if that's reasonable) In case 1, an active crossover uses its power power source to ensure no crossover loss. In case 2, the pre-amp output is boosted to offset the inefficiency of the low-level passive crossover. In case 3, the pre-amp is boosted to offset the inefficiency in the high-level passive crossover. In all case, the speakers see 10V. Am I inching closer to reality here?
Yes, though a line level passive crossover doesn't normally have much insertion loss unless you try to do higher order crossovers without using inductors (ie cascaded RC sections). Also, high level passive crossovers aren't nearly that inefficient either, though often the high frequency section needs to be padded down significantly to match the efficiency of the woofer. An active crossover is just that, you can design it to have as much (or as little) gain as you want. Often the low pass is set up with some gain along with an output adjustment for balancing with the oftentimes more sensitive highs.
What this illustrates is that there are different approaches to address the same problem and can yield the results that you are looking for. The key points here are what you are intending to accomplish, how to best go about it and at what cost? The lowest cost is by using cheap a$$ speaker level passive crossovers. Usually not your best option. The easiest and most cost effective next step up is to re-cap the original capacitors with newer and better quality ones. The next step up might be using an active crossover and bi-amping. Active line level crossovers are generally not very expensive and afford you an almost infinite amount of adjustments at the touch of a button. There is a cost for the additional amplifiers that will be needed. Next might be line level passive crossovers. These can vary quite a bit in cost. Many factors here, circuit design, parts quality and whether or not the crossover network is fixed or adjustable and what those adjustments will allow. An additional power amp is still required. Finally, custom designed passive speaker level crossovers. In theory (at least), these will be better engineered and utilize higher quality parts than the manufacturer's crossovers. Usually, you can use either a single power amp or you will have the option for bi-amping. Custom crossover upgrades are not inexpensive. For example, if you have the Klipsch K-Horn's which crossover at 400-Hz. and opt for the top of the line ALK crossovers, it will run you: ES400 + ES5800 (ES-Khorn set) ............ $1,819.00 This is for a pair of extreme slope crossovers, the ES400 and a pair of the squeaker to tweeter ES5800 crossovers (this amount includes a 5% discount for ordering the package which includes the HF tweeter crossover. This is the main crossover (ES400) from the woofer to the midrange squaker horn. It comes in two parts that are mounted on either side of the squaker horn, to prevent crosstalk. It can be configured for either 2-way or 3-way operation. It can also be configured for bi-amping. This is the ES5800 tweeter extreme slope crossover. I am just explaining that building quality passive speaker level crossover networks is neither easy nor is it inexpensive. As of the beginning of the year, Al has a backlog of about seventy crossovers of various types and the turnaround times are currently averaging about three months. This is still subject to parts availability, which can be difficult in these times.
Okay, now I'm more or less back on track. I know the pre-amp is where the adjustment occurs, but in my efforts to simplify somehow I flew right past that... So now that I'm back to a proper frame of reference, I'm also back to the question of what tends to be more efficient - low-level or high-level passive crossovers. I'm sure there's quite a bit of variability here that's dependent on complexity of design, quality of parts used, dollars spent, etc. I'm wondering though if it's possible to generalize with some degree of confidence. If I'm trying to cobble together an understanding from what's been written here and what I've read (so far) elsewhere, I'm going to guess that low-level passives tend to be more efficient than high-level passives. Agree or disagree?
Thanks for laying this out in such detail. It's really helpful. Let's assume that freshly restored speaker-level crossovers are my baseline. And that expensive, custom speaker-level crossovers are out of my price range. (Also, there's clearly some disagreement among the contributors on this thread about where these might fall in the hierarchy you've offered. But I'm certainly in no position to challenge you on this.) And let's assume the second amp is already bought and paid for. So if I'm looking to move up your food chain here, that leaves me weighing active vs passive low-level crossovers. My general sense is that a passive is best if the network topology is simple and the insertion loss is low. I'm not interested in fiddling with endless crossover knobs and dials. Simpler is better for me. And honestly I'm still thinking that maybe there's an advantage to starting with a variable passive crossover and moving to one with fixed values - assuming I can get it confidently dialed in - with the notion that a fixed-value crossover will be more efficient and less likely to color the sound. I'm not interested in digital at all for my all-analog, vintage system. So it's the pros and cons of passive vs active analog that I'm trying to understand better.
