I'm new to these chip amps. I have some experience with vacuum tubes. Maybe this is common and someone can point me to a schematic or two. I need about 100 watts of power. So I think why not use two chip amps working 180 out of phase, one amp drives the positive speaker lead wire the other drives the negative speaker wire. So, I'd run the input (if it's not already a "balanced" source from say, an XLR cable) to a phase inverter (or an audio coupling transformer) and then to two 50W amps.
What I'm looking to build is a bass guitar amp. I would build four of the above (eight chips total) and each push/pull pair would drive one 10 inch speaker in a four speaker cabinet.
What I'm looking to build is a bass guitar amp. I would build four of the above (eight chips total) and each push/pull pair would drive one 10 inch speaker in a four speaker cabinet.
What you mean here is a bridged amp, this already exist so its nothing new, you can get four times the power doing this but remember that the amps will see only half the speakers impedance.
If two amplifiers in bridge mode share a 4ohm load, each amp will see 2 ohms, or 4 ohms if its a 8 ohm speaker.
This is important to remember or you´ll blow up the amps, or atleast run them very hot which will eventually make them fail, and when they do they till also kill your speaker.
I suggest you check the bpa300 thread, its about a 300 watt bridge amplifier using i think six LM3886.
If two amplifiers in bridge mode share a 4ohm load, each amp will see 2 ohms, or 4 ohms if its a 8 ohm speaker.
This is important to remember or you´ll blow up the amps, or atleast run them very hot which will eventually make them fail, and when they do they till also kill your speaker.
I suggest you check the bpa300 thread, its about a 300 watt bridge amplifier using i think six LM3886.
That BTL amp is a "Bridged" amp. I'm looking for "push/pull"
In a bridged amp you simply sum together several amps by wiring them in parallel.
In a push pull design one amp drives each speaker lead.
The difference is kind of like wiring the amps in series or in parallel.
Here is the simplest example of a push/pull design I can find.
http://www.drtube.com/schematics/hiwatt/hwpwr50w.gif
Notice the two triodes are 180 degrees out of phase. In a solid state design you don't need the transformer and can drive the speaker's voice coil directly. What I'm thinking of is roughly replacing each of the triodes with a cheap chip amp part.
I'm asking if anyone has seen this done
In a bridged amp you simply sum together several amps by wiring them in parallel.
In a push pull design one amp drives each speaker lead.
The difference is kind of like wiring the amps in series or in parallel.
Here is the simplest example of a push/pull design I can find.
http://www.drtube.com/schematics/hiwatt/hwpwr50w.gif
Notice the two triodes are 180 degrees out of phase. In a solid state design you don't need the transformer and can drive the speaker's voice coil directly. What I'm thinking of is roughly replacing each of the triodes with a cheap chip amp part.
I'm asking if anyone has seen this done
Hi,
look up the National datasheets and application notes for their range of chipamps.
They give examples of single, bridged (BA100), paralleled (PA100) and bridge/paralleled (BPA200).
They also give the design procedures and component values.
The idea of 4amplifiers, each dedicated to a single driver is good.
Each driver in it's own volume would be better. This would allow for a failed/shut down amplifier or blown speaker. The remaining three would keep operating.
One failed driver in a 4way cabinet will probably not sound nice.
Make sure all your drivers you intend to power from bridged chipamps are really 8ohm. Design each channel of the amplifier to drive 4ohms and 2r0. The 2r0 loading is not as severe as driving a real 4ohm speaker. I would consider a 1r4 load as a good test for a 4ohm amplifier.
look up the National datasheets and application notes for their range of chipamps.
They give examples of single, bridged (BA100), paralleled (PA100) and bridge/paralleled (BPA200).
They also give the design procedures and component values.
The idea of 4amplifiers, each dedicated to a single driver is good.
Each driver in it's own volume would be better. This would allow for a failed/shut down amplifier or blown speaker. The remaining three would keep operating.
One failed driver in a 4way cabinet will probably not sound nice.
Make sure all your drivers you intend to power from bridged chipamps are really 8ohm. Design each channel of the amplifier to drive 4ohms and 2r0. The 2r0 loading is not as severe as driving a real 4ohm speaker. I would consider a 1r4 load as a good test for a 4ohm amplifier.
At that power level, digital would be lighter and cheaper. Since you're dealing with an analog input, a hybrid digital is the way to go, unless you're also using DSP in which case a pure digital can also be used.
There are premade modules that do exactly what you want.
http://www.coldamp.com/opencms/open...s/class_d_amplifier_modules/BP4078/index.html
Here's an article on the theory of Delta Sigma conversion which is what's used in many pure and hybrid digital amplifiers:
http://www.beis.de/Elektronik/DeltaSigma/DeltaSigma.html
Note that in practice, it works better to have two Delta Sigma stages running in parallel, with the same clock signal but with one signal inverted by 180 degrees, so the carrier more or less gets cancelled out leading to lower losses at low signal levels. (The design of a Delta Sigma digital amplifier, especially the pure digitals, is very involved and it is not recommended for a beginner to design one. There are chips that integrate the Delta Sigma modulators and for some, even the power stages in one easy to use package.)
There are premade modules that do exactly what you want.
http://www.coldamp.com/opencms/open...s/class_d_amplifier_modules/BP4078/index.html
Here's an article on the theory of Delta Sigma conversion which is what's used in many pure and hybrid digital amplifiers:
http://www.beis.de/Elektronik/DeltaSigma/DeltaSigma.html
Note that in practice, it works better to have two Delta Sigma stages running in parallel, with the same clock signal but with one signal inverted by 180 degrees, so the carrier more or less gets cancelled out leading to lower losses at low signal levels. (The design of a Delta Sigma digital amplifier, especially the pure digitals, is very involved and it is not recommended for a beginner to design one. There are chips that integrate the Delta Sigma modulators and for some, even the power stages in one easy to use package.)
leadbelly said:
Thanks. You're right. Different terms are used by different groups. I just read National's data sheet. They say "bridged" to "bridged" it is. Thanks again
About the four compartment cabinet. You are thinking hi-fi. Many guitar cabinets are open back but not many bass cabinets. The reason is the instrument amps do NOT re-produce recorded music. They are the source and the amps "color" the sound many times dramatically altering it.
One reason for using one (bridged) amp per speakers is that I'd only have to design one PCB and then I'd have an entire range of amps for 100W up.
It will be a while until I need this. Got a few projects in the pipeline. I'm building an amp now with four 6L6 tubes and then after that a small single end amp.
Push Pull is a term that applies only to tube amps, or transformer coupled amps.
It means two devices that conduct on alternate half cycles of the input signal.
The internal circuitry of the LM3886 already has two output transistors connected in a totem pole configuration that conduct on alternate cycles, so internally the amp already IS push pull in that sense.
Connected two chips in a bridge (BTL) configuration drives each amp out of phase from the other and the output is taken between the two amps rather than from either amp to ground. The result is that the two amps drive the load out of phase or on alternate cycles of the input signal (similar to push pull) and the output voltages of the amps are in series with the load. By Kirkoff's law this results in double the voltage across the load, which results in 4 times the power into the load vs a single amp. The amps must be able to handle a load impedance of 1/2 the value normally used in a single amp configuration.
It would also be possible to drive the load through an output transformer that would match the load to the chips. In this case the center tap of the transformer might be connected to ground and each side of the primary to it's respective amp. The turns ratio of the transformer would match the load to the amps (the impedance ratio is the square of the turns ratio). You'd probably have to have the transformer custom wound, but it would allow combining two chip amps in a bridge configuration as if it were a push pull amp while not presenting too low an impedance to the chips. A pair of amps could then be run into a 4 ohm impedance from 25 volts ang deliver twice the power. But why go through the cost of a transformer when a parallel connection will do the same thing?
It means two devices that conduct on alternate half cycles of the input signal.
The internal circuitry of the LM3886 already has two output transistors connected in a totem pole configuration that conduct on alternate cycles, so internally the amp already IS push pull in that sense.
Connected two chips in a bridge (BTL) configuration drives each amp out of phase from the other and the output is taken between the two amps rather than from either amp to ground. The result is that the two amps drive the load out of phase or on alternate cycles of the input signal (similar to push pull) and the output voltages of the amps are in series with the load. By Kirkoff's law this results in double the voltage across the load, which results in 4 times the power into the load vs a single amp. The amps must be able to handle a load impedance of 1/2 the value normally used in a single amp configuration.
It would also be possible to drive the load through an output transformer that would match the load to the chips. In this case the center tap of the transformer might be connected to ground and each side of the primary to it's respective amp. The turns ratio of the transformer would match the load to the amps (the impedance ratio is the square of the turns ratio). You'd probably have to have the transformer custom wound, but it would allow combining two chip amps in a bridge configuration as if it were a push pull amp while not presenting too low an impedance to the chips. A pair of amps could then be run into a 4 ohm impedance from 25 volts ang deliver twice the power. But why go through the cost of a transformer when a parallel connection will do the same thing?
odd
I thought this was a solid state push pull amplifier....
An externally hosted image should be here but it was not working when we last tested it.
I thought this was a solid state push pull amplifier....
Sure, but that's not a push pull amplifier.bacon665 said:
A mixer circuit with a push pull input signal to the grids, a second signal input to the cathodes, and the plates tied together for the output will CANCEL OUT the input signal from the grids and ONLY output the sum or difference between the two input signals. This is a balanced mixer circuit. But this has nothing to do with an amplifier circuit.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.