Well, this seems to come up an awful lot so I figured I'd provide the best explanation I could. I'll do my very best to simplify this and put things in laymen's terms. Here goes. Bridging - this is when a load is connected across two channels of a stereo power amplifier. The term bridging itself refers to the act of connecting across the channels like a bridge connecting two points. Consider the following: Each channel (or half) of the power amplifier drives half of the load. Therefore, when connecting an 8 Ohm load, each channel of the amplifier sees a 4 Ohm load. Power is doubled when bridging to a given load - IE, an amplifier that can make 50 wpc into an 8 Ohm load will make 100 w into 8 Ohms when bridged. [This does not apply to all makes and models of amplifiers - explanation to follow.] Now, some amplifiers require you to throw a switch on the rear panel when bridging and others do not. This switch may have several functions, depending on the design and make of the power amplifier. The net result is that one channel of the amplifier will drive the positive half of the waveform and the other channel of the amplifier will drive the negative half of the waveform. McIntosh power amplifiers are somewhat different in their fundamental design from many other power amplifiers in that the channels have Common terminals. These Common terminals are always negative. In addition, the positive terminals are always HOT. Other amplifier designs invert one of the amplifier channels - RED is HOT on one channel (typically LEFT) and BLACK is HOT on the other channel (typically RIGHT). The signal is inverted at the output so that a bridged load can be connected across these terminals which are obviously both HOT without having to have a switch to do so. Because the one channel is inverted at the output, the signal is also inverted as it comes into the input of the amplifier on the same channel. This allows both channels to operate in phase with one another when using the amplifier in stereo and allows both channels to operate out of phase with one another when bridging the amplifier - all without having to flip any kind of switch. There are surely other methods, but these would be the two most common. For the following computations, consider: P = Power I = Current in Amperes E = Voltage in Volts R = Resistance in Ohms Now let's apply Ohm's Law (I = E / R), the Power formula (P = I x E), and combinations thereof to a few examples to see what's really going on when we bridge a power amplifier. Now keep in mind that when we talk "R" here, what we're really talking is a load resistor on a test bench - a constant, which is required to allow us to apply Ohm's Law and the Power formula in an effort to make comparisons. A 50 wpc power amplifier will make 20 Volts per channel into an 8 Ohm load. Let's first prove this: P = E^2 / R P = 20^2 / 8 P = 400 / 8 P = 50 Watts Bridging Example 1 - The power amplifier doubles the power of a single channel when bridging into an 8 Ohm load: P = E^2 / R P = 28.28^2 / 8 P = 800 / 8 P = 100 Watts Note that the voltage increases accordingly, but it is not doubled. Bridging Example 2 - The power amplifier doubles the voltage of a single channel when bridging. Therefore, it will quadruple the power of a single channel - (2^2 = 4) - when bridged into an 8 Ohm load: P = E^2 / R P = 40^2 / 8 P = 1,600 / 8 P = 200 Watts McIntosh amplifiers with autoformers follow the first example while some other brands of power amplifiers follow the second example (specifically those that permit bridging but do not have switches to do so). It is important to note that the designers may choose to current limit a power amplifier which can show up in the amplifiers specifications - IE: 200 wpc stereo into 8 Ohms, 300 wpc stereo in to 4 Ohms, 600 wpc bridged into 8 Ohms. The easiest way to limit current is to limit voltage to the output devices to prevent the amplifier from doubling power when impedance is halved. This can prolong the life of the output devices or allow an amplifier to provide the desired play time given the mass of heat sink the bean counters approve in the design phase (this is referred to as thermal time - time to thermal shut down). Don't let this throw you off when applying the above math but this is reality versus theory. Strapping - This is when a load is connected to both the left and right channels simultaneously. This is often referred to as Parallel-Mono as the stereo channels are wired in parallel with one another - Hot to Hot and Common to Common - and the load is then tied to them both. The term strapping itself refers to using short straps of wire to connect the channels together. The idea here is that when both channels are doing identical work and the signals each channel is amplifying are identical one can tie them together. [NEVER try this on an amplifier that was not specifically designed to do this! Especially when the Red on one channel is the HOT and the Black on the other channel is the HOT as described above.] When it comes to HiFi power amplifiers, McIntosh is the only company I'm aware of that permits this and this applies only to McIntosh amplifiers with autoformers. In the pro industry they all do it - Crown being I believe the earliest adopter - and they're all direct coupled designs. Many McIntosh and Crown power amplifiers allow either Bridged-Mono or Parallel-Mono wiring schemes. This makes a given power amplifier incredibly versatile. Now, the common perception is that when you strap channels you double the current. Is it true? Consider the following which I've copied directly from the McIntosh MC2300 owners' manual - a McIntosh SS stereo power amplifier with autoformers that can only be strapped to mono and cannot be bridged: Power Output / Stereo: 49.0 Volts across 8 Ohms (49^2 / 8 = 300) Power Output / Mono: 69.3 Volts across 8 Ohms (69.3^2 / 8 = 600) Recall, it's specifications are 300 wpc into 8 Ohms stereo and 600 Watts into 8 Ohms mono. The math is just as when bridging - note the difference in voltage. Hmmm . . . But is current doubled? Stereo: I = SQRT(P/R) I = SQRT (300/8) I = SQRT (37.5) I = 6.12A Mono: I = SQRT(P/R) I = SQRT (600/8) I = SQRT (75) I = 8.66A It isn't. The reason for this is because power isn't quadrupled when connecting the amp in mono with the McIntosh design. If it were, then yes current would be doubled. Plug 1200 Watts into the above formula and do the math . . . Let's look at a Crown design. This is copied directly from the Crown Macro Tech amplifier brochure: Bridge-Mono mode provides double the output voltage and Parallel-Mono mode provides double output current from a single channel. OK, so it's easy to see where the perception comes from. It's printed right there in the manual. Is it true for the Crown MacroTech 2400? Let's find out. Here are the printed specs on the Crown MacroTech 2400: 800 Watts into 4 Ohms per channel - let's look at both voltage and current: I = SQRT(P/R) I = SQRT (800/4) I = SQRT (200) I = 14.14A E = SQRT(RP) E = SQRT(4 x 800) E = SQRT(3,200) E = 56.57 Volts [Checking our math, P = I * E, P = 14.14 * 56.57, P = 800W] 2,070 Watts into 4 Ohms bridged-mono E = SQRT(RP) E = SQRT(4 x 2,070) E = SQRT (8,280) E= 90.99 Volts This is a case where the designers implement current limiting for some reason (prolong the life, only so much heat sink mass to dissipate heat, etc.). If the amp was not current limited and it did actually double the output voltage, it would quadruple power at 3,200 Watts into 4 Ohms bridged mono. [This is an incredibly powerful amplifier nonetheless.] 1605 Watts into 2 Ohms parallel-mono I = SQRT(P/R) I = SQRT (1,605/2) I = SQRT (802.5) I = 28.32A Interesting. The amplifier does double it's current BUT this results in only a doubling of output power as impedance is halved as well. Note that if we do a comparison in 4 Ohms parallel-mono, this is not comparing apples to apples as the amplifier only makes 1,035 watts when operated into 4 Ohms parallel-mono. [This amounts to only 16.1A at 4 Ohms in parallel-mono.] We really need to look at its 1 Ohm parallel-mono ratings to compare apples to apples, which is 2,080 Watts: I = SQRT(P/R) I = SQRT (2,080/2) I = SQRT (1,040) I = 32.24A So, there are no hard fast rules in regards to doubling of voltage when bridging or doubling of current when strapping that apply to all amplifiers. As you can see, this is greatly dependent on the design. I hope this helps. I'm sure some Crown fans will chime in on strapping.
Generally speaking, the easiest way to know the bridged power for conventional (aka Crown, et al) amps is simply to sum the watts per channel and multiply the impedance x2. For example, if the amplifier in stereo mode is rated: 100wpc @ 8 150wpc @ 4 200wpc @ 2 Then in bridged mode the power of the example amp would be: 200W @ 16 300W @ 8 400W @ 4 Notice in bridged mode there is no 2 ohm rating. That's because it would be like operating with 1 ohm load in stereo mode, and the example amp doesn't carry 1 ohm stereo rating. Rather than look at bridging as doubling the voltage swing, it's better to look at it in terms of summing the stereo operation voltage swings. That works always since very few amps do the theoretical doubling as impedance drops. Referring back to the example amp, let's look at stereo 4 ohms / bridged 8 ohms. Stereo at 4 ohms per channel (150wpc @ 4 ohms) the example amp will swing 24.5 volts on each channel. Add those per channel voltage swings to end up with 49V bridged. Now apply that 49V swing back to the bridged load (8 ohms) and you see where the 300W @ 8 ohms comes from. With this type of amp (aka Crown, et al). Parallel mono/strapped math works similar to bridging math, except load is 1/2 and it's the sum of current instead of voltage. Again going back to the example amp, in a strapped configuration it would be 200W @ 4 300W @ 2 400W @ 1 How do we get there? Well, the stereo 8 ohm rating was 100wpc. 100wpc @ 8 ohms is 3.54A. Since in parallel current sums, it's now the equivalent of 7.08A at 4 ohms, which is our 200W @ 4 ohms parallel/strapped rating.
I don't understand all the math I have 2 of the new MC275 Mk VI, said to have 90wpc. Witch way should I go? Witch way is the safest for my new equipment? What would be the pros and cons for both ways?
As long as amp offset voltages are zero, on both channels and amps can operate in bridged mono or parallel there should not be a safer. If offsets between channels are mismatched in strapped mode one channel will work harder or run hotter by being strapped to the other attempting to establish a balance.
I've personally never tried parallel-mono and bridge-mono on the same power amplifier with the same speakers. As I normally run one pair of 8 Ohm speakers, this is typically not an option to do so. It is an option for lower impedance loads. My MC2600s are bridged - simply because that's the only way that I can run them mono into 8 Ohm speakers. Before, my MC2300s were strapped - again, simply because this was the only way to connect them in mono. Velocityboat has the MC2500s and he's running 4 Ohm loads - so he can actually do this comparison and report back to us on any perceived SQ differences between strapped and bridged. twiiii reports better bass in club installs when using parallel-mono. It is interesting to note that McIntosh abandoned parallel-mono wiring schemes altogether and quite some time ago. My MC2600s make 1,200 Watts parallel-mono or bridge-mono. Personally, I'd be hesitant to say that 1,200 Watts would sound better than 1,200 Watts based on the math that I presented above and how it relates to current and voltage. That's not to say that there won't be an audible difference - I've just never been in a position to experiment with that. I did once use my Crown MT2400 to drive (16) 8 Ohm 12" woofers which were series-paralleled to a single 2 Ohm load via 8 AWG cable. I used the amp in parallel-mono and the bass was pretty epic. Of course, that may have to do more with minimal power loss, cone coverage, and mutual coupling than the parallel-mono operation.
I specifically bought my MT-1200s to try parallel mono/strapped operation with a pair of Thiel CS3.6 (~2.7 ohms). Didn't do an exhaustive evaluation but in that time nothing jumped out at me as a big improvement over other amp delivering similar power.
Biggest perceived difference with my line array stack. This is a true hybrid so point source may sound different. Bridged More open highs plus more amp noise. Not sure if this is because there is no common hooked to speakers or one MC2500 has not had Rich Andrew's speaker control relay mod performed. I did have a hum in bridged I cured by moving amps to same ground. Percussion pops like live. Strings very airy. Sound Stage immersion width and depth goes to infinity on live recordings. Closer to sound my Yamaha Pro Amp gives. Except the highs float way more with Mcintosh sound. It gives me goose bumps and it is electrifying. Mono parallel More Bass, way quieter, highs still excellent. Highs sound as they do running one amp stereo. Bass more out front the highs become second fiddle in what you notice. Where bridged always hanging there. Old amp has fan start sooner maybe because I am running new one with bottom and top covers off. The power guard peak lights not timed exactly though same source channel either amp. Did mot notice in bridged. To be fair I need to rerun Room Perfect for both. If I can bring up bass with bridged and have those highs, and do the RA mod to quiet the right MC2500. I think I have a clear winner. Bit with balanced Room Perfect maybe I won't heard such a difference. However yet to test each speaker off its own 8 ohm tap with just a mono signal fed to each channel. My wife likes mono parallel more. Then again she thinks boomy bass is great so it maybe all about the bass reaction.
In order for such a comparison to be objective, it would have to be done without any room correction or equalization applied.
So far I have kept room correction constant as well as equalization. Since my listening area is not an anechoic chamber that is as objective as I can go. Besides hitting bypass on RP which would favor the less bass solution since my speakers can overpower room in bass peaks at certain frequencies without RP and moderate volume.
Nice write up Tony. Just a couple of notes. The most common amplifier design are such that the negative speaker terminal is common/earth. The only amplifiers which I am aware if that are not are floating bridge designs, and I believe quad balanced McIntosh amplifiers fall into this category as the negative speaker terminal is also above ground. As far as bridging and what happens to the outputs, on all amplifiers I have had experience with bridging which are QSC, Crown, BGW, Carver, Lab Gruppen, ALL these amplifiers do this when they are bridged. One channel pushes positive, and the other channel pulls negative as you have stated, the inverting of the channel which pulls negative always happens in the front end of the amplifier. You never use the negative terminal in a bridging situation, the amplifier becomes floating and therefore cannot use anything common to the other amplifier channel for its output. There is also gain compensation built into it as well and the gain of the amplifier will increase in gain with bridging. So, the channel 2 positive output becomes Positive, and the channel 1 positive output becomes negative. They always in all the amps I have bridged work this way. The only one which is different is the Macro-tech, they use channel 2 as Neg and channel 1 as positive. Carry on men......great thread...
Lack of common ground and increased gain. Definitely explains increase in amp noise I notice in bridging. The push pull sure seems to change sonic high character. Somewhat more of a tube sound perhaps.
All my QSCs are inverted on Ch 2 so Ch 2 (+) is actually the negative speaker connection in bridged mode. Doesn't really matter though as bridged mode connections are usually indicated near the speaker connections. QSC has a nice illustration showing the normal and inverted signals acting on the speaker.
I'll have to check my system tonight and see if I can notice any more amp noise since I've bridged. Nothing obvious so far, but perhaps I'll notice it if I go looking for it. I'll just crank the volume with no music playing to see how much hissing there is.
That's not true Tony, my new mc275s directions say to do Parallel mono and no other option. The mc 402 is the same way in the instructions.
Thanks for taking the time to type this up. Eventually I would like to run a pair of 2105's. This is an excellent reference.
Yes damacman thank you for the write up. There's a lot there for me to digest. Now where is that 2nd mc2500 so I can make my own decisions on which way sounds best
No, a bridged amplifier rejects more noise than a normal amplifier. Most amplifiers will have compensation in the switching for gain so they are the same as when they are on normal mode. I am not sure if McIntosh do this or not, I would have to check out the schematic. At the same gain, a bridged amp will be quieter, this is the same basic theory behind differential amplification, this is the theory which is behind the whole "quad balancing" thing McIntosh does. The amp is class AB push pull in both normal, bridged mono, or parallel mono...