I imagine that many folks that operate a power amp in bridged mode do so for the purpose of bi-amping speakers, however I am interested in running both channels of a stereo amp in parallel in order to double the output power, thus using each stereo amp as a mono-block. I am having some trouble getting my head around this and need some help.
It's my thinking that connecting the two input RCA positives and then strapping the 8 ohm tap ought to get the job done. I'm thinking that the O/Ts act as a voltage source so that putting two in parrallel will double the available current but leave the voltage unchanged so that the impedance needs would not change, ie the two strapped 8 ohm taps would be connected to an 8 ohm speaker.
It has been suggested that the two channels should be run out of phase and/or the O/Ts operated in series. Both ideas I am inclined to dismiss as incorrect, however they were suggested in earnest by serious people.
What is the correct way to operate a stereo amp as a mono-block and what is the correct way to think about the physics of the problem?
Thanks
It's my thinking that connecting the two input RCA positives and then strapping the 8 ohm tap ought to get the job done. I'm thinking that the O/Ts act as a voltage source so that putting two in parrallel will double the available current but leave the voltage unchanged so that the impedance needs would not change, ie the two strapped 8 ohm taps would be connected to an 8 ohm speaker.
It has been suggested that the two channels should be run out of phase and/or the O/Ts operated in series. Both ideas I am inclined to dismiss as incorrect, however they were suggested in earnest by serious people.
What is the correct way to operate a stereo amp as a mono-block and what is the correct way to think about the physics of the problem?
Thanks
Never parallel solid state power amplifiers. If your amp can't deliver the load currents replace it or re-design it.
Bridge mode runs 2 amplifiers out of phase and then connects the load between the 2 'hots'. This doubles the voltage and current quadrupling the power output but you don't get something for nothing. Each amplifier sees 1/2 the load impedance so don't try this with 4 ohm speakers unless the amp is rated for 2 ohm loads. In theory any amp can be bridged but in practice it's often not a good idea.
Creating the out of phase signal can be done many ways - transformers (yuck), opamp inverters, tube inverters or tap into the feedback network of a solid state amp and use the amp itself as the inverter.
G²
As stated operating two amps out of phase is called bridging, and this is the usual method for solid state amps. Paralleling is just what the name implies. This is often the best course of action with a tube amp. I am generally not afraid to blow something up, and I am always looking to get more power, so I have tried both methods on several tube amps. Many tube amps will react poorly to bridging since no two amps will have exactly the same output impedance they will not share the load equally. The two NFB loops (if used) will fight each other possibly creating an unstable situation. Low frequency oscillation is a possibility. I have paralleled many tube amps without blowing anything up, but some don't sound so good. Many do. I am currently working on an amplifier using two Simple P-P boards tweaked up for extreme power and paralleled. I get 60 watts out of each board running the two channels in parallel.
Your thinking is correct except for one flaw. Paralleling the channels does leave the voltage unchanged, and it does double the AVAILABLE current, but attaching the same old 8 ohm speaker across the same voltage source, will not DRAW any more current. In order to use the available double current you must halve the load impedance to draw double the current. This means connecting the 8 ohm speaker to the 16 ohm tap.
Since I built my own amps, I found that you can do this with one OPT if it is big enough to handle the increased power. Wire the plates of the two channels in parallel, and run them through an OPT of half the normal impedance, or the same OPT with the 8 ohm speaker on the 16 ohm tap.
As Stratus said you run the channels out of phase and connect you speaker between the positve terminals. The effective load per channel is halved as said and in a tube amp with multiple secondaries you need to connect for 4 ohm for a 8 ohm speaker to maintain correct loading of the output tubes. This will give you the double power instead of the theoretical 4 times with an ideal voltage source.
Theoretically you could achieve the doubling of power without inverting phase by connecting the channels in parallell and shift the speaker connection to 16 ohm (if available) for your 8 ohm speaker. Haven't tried this and threr might be practical aspects I haven't thought of.....
//Olof
Oups, already covered by George........too slow
Theoretically you could achieve the doubling of power without inverting phase by connecting the channels in parallell and shift the speaker connection to 16 ohm (if available) for your 8 ohm speaker. Haven't tried this and threr might be practical aspects I haven't thought of.....
//Olof
Oups, already covered by George........too slow
Last edited:
Paralleling may work, but can be problematic. Bridging is guaranteed to work for stereoblocks with a common ground.
Stratus46 provided the key fact. When you bridge, each amp section "sees" only 1/2 the load. That means 16 Ω speakers get connected to 8 Ω taps. and 8 Ω speakers get connected to 4 Ω taps. Unless you employ custom "iron" that has 2 Ω taps, 4 Ω speakers are out, when bridging "hollow state" stereoblocks.
I'll disagree about the suitability of a trafo for producing the requisite 180o out of phase drive signals. Driven from a low impedance source, Sowter's model 8920 will do quite nicely.
Stratus46 provided the key fact. When you bridge, each amp section "sees" only 1/2 the load. That means 16 Ω speakers get connected to 8 Ω taps. and 8 Ω speakers get connected to 4 Ω taps. Unless you employ custom "iron" that has 2 Ω taps, 4 Ω speakers are out, when bridging "hollow state" stereoblocks.
I'll disagree about the suitability of a trafo for producing the requisite 180o out of phase drive signals. Driven from a low impedance source, Sowter's model 8920 will do quite nicely.
I have seen tube amps go unstable when bridged. If the amp is P-P and uses a good deal of GNFB with somewhat marginal stability (many Williamson designs) the amp may motorboat wildly at a low frequency. I remember vividly an old amp (a Pilot maybe) that was in good condition, and worked well in normal Stereo, but oscillated at about 1 Hz when bridged. It was so bad that the red spots on the EL34 plates appeared to pulsate. It worked OK when paralleled.
I am sure that there are cases where paralleling doesn't work well either. With tube amps of varying design there can't be one method that is "guaranteed" to work. I have tested both methods on most of the amps that I build and on a non feedback amp they both generally work. A paralleled SE amp is still SE. A bridged SE amp becomes push pull, but usually remains class A, unless you drastically reduce the load impedance. Paralleling is the simplest since no external phase splitter is needed.
I have also found ways to bridge the SSE boards through a single OPT and get 50 watts out of one board using 6L6GC's. It is generally called push pull AB1. Yes, an external phase splitter is needed.
Cheapie Edcors seem to work OK too when driven by a low impedance source. The Triode Electronics universal driver transformer is limited in frequency response, but worked OK for a guitar amp. It also has 4 secondary windings. You can wire 4 amps in parallel - bridged mode for extreme power but blown parts are possible. Made a wicked loud screamin guitar amp out of 2 SSE's.
Last edited:
There are some amplifiers that will operate in parallel mode. Some Crown amplifiers are designed to work in a dual parallel mode as well as bridged or stereo mode. In the parallel mode they will drive a lower impedance load than what they will normally drive in stereo mode. In the world of commercial audio in which I live I can tell you that I have never used a stereo amplifier in parallel mode. When I needed more power I just installed a larger amplifier.
As others have already gone thru the concept of bridging and proper hook up I won't repeat that again other than to mention not to connect any bridged amplifier to a load lower than 8 ohms unless the amplifier manufacturer specifies that the amplifier in question is designed to run a load lower than 8 ohms in bridged mode. Some of the newer QSC amplifiers will run a load lower than 8 ohms in bridged mode.
If you are thinking of biamping I would suggest purchasing an amplifier capable of operating the low frequency section of the speakers and probably using the amplifier that you have now to power the high frequency section of the speakers. In most cases the speaker designer/engineer has some guideline as to what power to use as a minimum and maximum. Generally you can use what ever you have as long as it is above the minimum guidelines that are set forward by the manufacturer.
As others have already gone thru the concept of bridging and proper hook up I won't repeat that again other than to mention not to connect any bridged amplifier to a load lower than 8 ohms unless the amplifier manufacturer specifies that the amplifier in question is designed to run a load lower than 8 ohms in bridged mode. Some of the newer QSC amplifiers will run a load lower than 8 ohms in bridged mode.
If you are thinking of biamping I would suggest purchasing an amplifier capable of operating the low frequency section of the speakers and probably using the amplifier that you have now to power the high frequency section of the speakers. In most cases the speaker designer/engineer has some guideline as to what power to use as a minimum and maximum. Generally you can use what ever you have as long as it is above the minimum guidelines that are set forward by the manufacturer.
Thanks guys for the thoughtful responses. I didn't say so in my post, because I was seeking a general discussion, but I'm fooling around with a push pull application. I was confusing bridging and paralleling.
It's bridging I want to follow-up on first. First, I think Eli has it right with 8 ohm loads into 4 ohm taps rather than 16. But, I know better than to bet against George; so who's got it right here?
But the signal inversion is more interesting. I understand the transformer idea and I even found a manufacturer that offers a "module" with a phase inverting transformer as a plug-in for their amp. But given that I'm fooling around with a pp amp, why not just lift and swap the wires at the plate and cathode of the phase inverter on one channel, or swap leads at the coupling caps, or swap O/T primary leads on one channel at the plates? Wouldn't that get the same result with a lower parts count?
I'll post a few questions on paralleling later, after I mull it over a bit.
Thanks for guiding me to the light.
It's bridging I want to follow-up on first. First, I think Eli has it right with 8 ohm loads into 4 ohm taps rather than 16. But, I know better than to bet against George; so who's got it right here?
But the signal inversion is more interesting. I understand the transformer idea and I even found a manufacturer that offers a "module" with a phase inverting transformer as a plug-in for their amp. But given that I'm fooling around with a pp amp, why not just lift and swap the wires at the plate and cathode of the phase inverter on one channel, or swap leads at the coupling caps, or swap O/T primary leads on one channel at the plates? Wouldn't that get the same result with a lower parts count?
I'll post a few questions on paralleling later, after I mull it over a bit.
Thanks for guiding me to the light.
Ok, my bad. I had it right when I declined to bet against TubeLab but no so much when I misinterpreted his post. His comments:
refer to paralleling, not bridging. I get it now.
refer to paralleling, not bridging. I get it now.
If your amp has gNFB it will suddenly become positive. Non feedback amps will work just fine. Even with local feedback it may work but you need to find how and where it is applied.
//Olof
Good point, Olof. I hadn't considered that yet. In this situation, feedback is from the transformer secondary to the phase splitter cathodes. That would be positive if I swapped transformer primaries at the plates. It looks like my idea is too clever by half.
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.