Talk:Bridged and paralleled amplifiers

Created the paralleled amplifier article, merged the paralleled & bridged amplifier articles into this one. Rohitbd 09:44, 18 September 2005 (UTC)

H-Bridge
H-Bridge refers to a fairly specific circuit. As such, it is not the correct term in general. Transformer coupled amplifiers, resistively loaded, and certian other amplifiers may not fit this discription very well. I don't think I've ever seen "h-bridge" used outside of a switchmode context. Mag-amp based designs likely don't fall well into this category. —Preceding unsigned comment added by 71.32.46.111 (talk) 09:32, 28 September 2008 (UTC)

The 4x power myth
"For example, if an amplifier can deliver a maximum of 100 W each into two 8 O loads in stereo mode, the same amplifier when bridged will deliver 400 W into a single 8 O load."

This is wrong, despite being often said. An amplifier can not double its current and power ratings simply by being connected to a different load. Each amplifier still has the same p, i and v ratings it had before. If you feed a single load with 2 amplifiers each of power rating P, connected in whatever fashion, it is not possible for the amps to suddenly have an output rating of 2P each.

A bridged output enables twice the voltage swing on the load. Thus a given power rail (or rails) can deliver upto 4x the power _using a different amplifier of different design & ratings_. This is why bridged amps are popular in car audio, they avoid the need to convert the low 14v supply to a higher voltage. Tabby 17:36, 3 December 2007 (UTC)


 * Well, look at it conversely...if an amp is designed to be operable into a 4 ohms load, then two such amps can be easily bridged into a single 8 ohms load. If one amp can drive 10W into 4 ohms then while bridging into 8 ohms, the total power delivered is 20W (same as two amps pusing 10W into 4 ohms each). Now consider a single amp driving 8 ohms (but capable of driving 4 ohms too): The power delivered is 5W (half of that into 4 ohms). So two 5W amps when bridged into 8 ohms should now drive 4x the power, i.e., 20W (as above). Rohitbd (talk) 13:37, 2 January 2008 (UTC)

I think you're assuming that current rating has some direct relationship with power rating. Amps are limited by their supply voltage and by the rate at which they can dissipate heat without being destroyed. An amp could be capable of sourcing a bazillion amps without being destroyed, but if the supply voltage is limited to a 9 V battery, the amp will have a very low power rating. Connecting two of them in bridged configuration into the same load will allow 4x the power, no? — Omegatron 02:42, 3 January 2008 (UTC)

The issue is that the comparison can seem unfair. the amplifiers listed above would need to be capable of 200W @ 4ohm each. Two of such amplifiers would be capable of a total of 400W. A bridged combination of these amplifiers would need an 8ohm load to achieve 400W output. This same 8ohm speaker, attached to either of the amplifiers alone, would only draw 100W. Thus the bridged configuration allows for 4x power. It is necessary that the amplifier is capable of the output current and any additional heat associated with this extra power. —Preceding unsigned comment added by 71.32.46.111 (talk) 09:25, 28 September 2008 (UTC)

This is pointless and idiotic. It's is true that if you took a commercial stereo Amp and connected it in Bridge configuration it wouldn't be able to put out four times the (continuous) power. But it probably would be able to manage four times peak power, which is what usually matters (nobody listens to continuous sine waves). But the real issue here is to the designer of a new amplifier. The Bridge configuration is almost universally used in Car Stereo's and small domestic equipment (MP3 players, etc) which must run off low voltages. 119.18.11.19 (talk) 03:32, 31 March 2014 (UTC)

This is not a myth! If a 2 channel amplifier is bridgeable the output can be 4 x the single output (or close to it) so long as the impedance is the same and the power supply is capable of supplying the current. Another limiting factor is the efficiency of the amplifier. A lot of power can be lost in heat. Using low power amplifiers will provide up to 4 x the single output when bridged. I have done this many times and I am in the industry. I measure THD and am speaking of values of around 1 to 3%. Only when high power amplifiers are used does the loss become greater due to the limitations of the supply and design/cost. — Preceding unsigned comment added by User: (talk • contribs)

This is a completely unreferenced section that's apparently propagating someone's WP:TRUTH and confuses a bunch of things, such as power and maximum power. It should probably be deleted or substantially rewritten. Fact of the matter is that many power opamp data sheets advertise approx. 3x-4x increase in power correctly (as long as SOA == safe operating area) of the opamp is not exceeded. To pick an example, TEA2025 (http://www.st.com/web/en/resource/technical/document/datasheet/CD00000172.pdf) claims 0.02W in 32ohms at 3V single-ended and 0.06W in 32ohms at 3V bridged (3x increase in power). It also claims 1W at 6V in 4ohms single-ended and 2.8W at 6V in 4ohms in bridge config (that's about 2.8x the power). The same op amp also claims 2.3W in 4ohms at 9V (single ended [i.e per stereo channel]) but only 4.7W in 4ohm bridged at 9V; there's only a 2x increase in power at this higher voltage because the thermal envelope (SOA) of the chip is maxed out; 4.7W or so the max power the chip can safely dissipate. 86.121.137.150 (talk) 21:01, 9 December 2014 (UTC)


 * The myth section is misleading. Bridging amplifiers to get well over 2x power is quite normal in situations where a limited supply rail means that otherwise unfeasibly low loudspeaker impedances would be required. Stub Mandrel (talk) 16:12, 26 January 2018 (UTC)


 * Stub, only if the power supplies are underwhelming for the rest of the amp and your building circuit(s) they're on can handle it does doubling the number of supplies suddenly have much advantage. The way most pro bridgeable stereo amps are rated is for their bridge mode up to their 1% THD point and then measured for their respective lower THD in stereo mode at arbitrary divisions of that rating per impedence for each channel and the fact their limiters are then set by design based on that means you usually get two-thirds more available voltage bridging than the simple summation of the individual stereo channels at that given impedence, but at the expense of increased distortion, stress on the components, heat, and risk of shut down. This seeming increase in available gain is partly an artifact of rating, measurement, and hardware limiting and not really real, and a little part of it is a real power increase, albeit poor-quality gain increase by running the unit with floating ground and in a forced lower impedence state.  If the amp is also not completely up to spec, all that distortion, stress, heat, and risk of shut down gets even worse and the lifespan will be much lower for the amp when bridged.  Two-thirds more voltage is only an increase in 4.4dB and less than double the perceived volume even at 20hz and is WAY less than double the volume at higher frequencies where you need about four times the gain (+12dB) to actually double the volume.  So it usually makes no sense to run a pair (or multiples of a pair) of identical subs from a single bridgeable stereo amp in bridge mode for such a moddest increase of available low-quality headroom instead of just running the amp in cleaner, safer stereo mode that will also keep the sound in stereo if you do, what little stereo imaging there is in the sub frequencies.  -DJ Reticuli 2605:A000:1301:8B09:9CBB:419B:92A6:76AC (talk) 22:24, 8 March 2019 (UTC)

DC offset & blocking capacitors
"One requirement of this configuration is that the output DC offset voltage of the amplifiers must always be equal (in magnitude and sign), preferably as close to zero as possible at no signal. Unequal offset will cause some direct current to flow through the load and the amplifiers, wasting power in all three. Practically, a small offset may be acceptable depending on final requirements or specifications."

No, this is a requirement for direct coupled capacitorless outputs, and applies equally to bridged and single ended drive. It has nothing to do with bridged outputs per se.

Bridged amps usually do use capacitorless output because the bridge topology makes this an inviting opportunity to save cost, but not all bridged amps do so, or can, and other amp types also often use capacitorless outputs.

Bridged amps can, and in a minority of cases do, run perfectly happily with unmatched and not very stable output offsets plus an output capacitor to block any dc offset.

"Amplifiers that use a single supply with an internal DC offset (at half the supply voltage for maximum undistorted voltage swing) must not drive that direct current into the load, so they typically use output coupling capacitors so that the speaker sees zero volts DC. In the case of a bridged amplifier, however, this is not necessary, since if both amplifiers have the same offset, the net voltage across the load is zero."

This isnt really true, though there is some truth within it. The fact that each output of a bridged amp quiesces at nominally the same V does not make it suitable for capacitorless working. For _any_ amplifier, small and stable V_offset is required for a capacitorless output, and such a condition is not automatic simply because a pair of amps are bridged. There have been numerous cases of hobbyists not appreciating this, and bridged capacitorless amps destroying themselves and/or speakers as a result. Tabby 17:36, 3 December 2007 (UTC)

Moved from main page (Ref: Expert needed)
I have moved the following from the main article. The context is indicated by comment annotations E1 and E2 in the main article.

E1
(edit) Because the above is so authoritively written, I am loathe to overwrite the "facts"in this section. Perhaps the originator will choose to investigate and clarify. It should read TWICE the power. see other edits below when the formula above is used correctly (end of edit) (edit2) No, it is 4 times. when you double voltage, current doubles too, so four times output. This really should be in the talk page...

E2
(Edit) The statements above fail to mention that when two amplifiers are connected in bridge mode the output impedance of the pair is DOUBLED. Thus, if "2R" is substituted for "R" in the formula given, (together with the implied "2Vsquared) it will be seen that the output power is also DOUBLED and not quadrupled as stated.

The first one appears correct to me; the second one I cannot comment on (My experience in analogue circuit design - and particularly amplifier design - is lacking) OwenS | T | C | 13:28, 3 September 2010 (UTC)

LAST WORD?? somehow I doubt it......

All that is needed to trash the X4 power theory is common sense. If an amplifier is rated at 100 watts MAXIMUM power into a specified impedance then THAT IS ALL IT CAN DO. PERIOD. So, two 100 watt 4 ohm amplifiers, HOWEVER combined can only output 200 watts. ( Someone has already said that above )

Lets say that that impedance is 4 ohms for arguments sake. If a 2 ohm speaker is substituted the amplifier CANNOT deliver twice the power to it. If you do, soon enough something will smoke and the amplifier will fail. This wpuld be the situation in the fallacious version of bridged mode. The two bridged amplifiers each see HALF the connected load impedance. In the case of two 4 ohm amplifiers connected wrongly to a 4 ohm speaker, each amp sees 2 ohms..At full(attempted) output the amplifier is seriously overloaded and it will fail.

Common Sense. Consider this. If an amplifiers output could magically be doubled by a different mode of connection amplifiers might be built in a box with that circuitry already implemented. If a 20 megawatt power station could be substituted for two 5 megawatt stations connected in bridged mode THAT IS HOW THEY WOULD ALL BE BUILT !! —Preceding unsigned comment added by 87.114.212.37 (talk) 11:10, 18 December 2010 (UTC)

The 4X idea is really not a "myth" exactly. Some Class-D products are in fact already produced to operate this way by default. If the current capability is there, you do get 4X the power, in a sense. Consider: a stereo power amp is capable of driving 8 ohm loads to 100W. In an ideal amp (i.e., ignoring various losses, "nominal" designation of loads, and other issues), if we switch the loads to two 4 ohm speakers, the current doubles and we get twice the power - 200W each side. Now connect both 4 ohm loads in series, and connect this 8 ohm total load across the two power amps, feeding them with opposite phase signals. Provided each amp has the current capability to run 4 ohm loads to 200W, this 8 ohm combined load (4+4 ohms) is driven to 400W. Hey, that's 4X right? Well, the "trick" of it can be viewed a couple ways - in the first case, we drove 8 ohms to 100W - TWICE (stereo, remember?), so 200W total power delivery occurring, really. In the bridged case, although the total load was 8 ohms, each amp sees a 4 ohm load (the center point, as long as we are talking about hypothetical perfectly matched amps and loads, remains at 0V potential). If you bridge a single 4 ohm load you are asking each power amp to drive 2 ohms. Many pro sound reinforcement (public address) amps and bass guitar amps can do 2 ohms these days. Bottom line, two amps bridged can theoretically deliver 4X the power into a given load as one of the amps alone could deliver into the same load. Have a nice day. -Jeffrey —Preceding unsigned comment added by 76.102.195.68 (talk) 02:06, 3 March 2011 (UTC)

As above, this is idiotic. For a designer, a Bridge configuration is quite standard, and yes they do put out four times the power. There are dozens of Audio chips which operate in this manner. But to argue that an existing off-the-shelf amp can't be tricked into putting out four times the power is pointless. It just shows that the writer has little knowledge of electronics design. 119.18.11.19 (talk) 03:43, 31 March 2014 (UTC)

Merger of Bridge-tied load and Bridged and paralleled amplifiers
Also the Bridge-tied load BTL article should be merged here, or at least the duplication eliminated somehow. 86.121.137.150 (talk) 21:16, 9 December 2014 (UTC)
 * Agreed and ✅ Klbrain (talk) 13:10, 6 August 2017 (UTC)