Hello,
Can the more experienced among you explain how two chips are wired in parallel or bridged?....Iv'e built a LM3886 clone thats been up and running and sounding good for almost two years.
I havn't visited the forum in quite some time and was interested in the new LM4780?(the one thats two 3886 on the same die)and how you actually,physically connect the pins for bridge and or parallel mode.
I have looked at the NSC App. notes and schematics but it's a little beyond my knowledge level.
Here's the questions I'm asking and they may sound stupid so bear with me:
1) If the LM4780,a two channel stereo chip, is composed of two LM3886 chips how do you physically combine them to make a more powerfull one channel amp.
2) I'm running 4ohm speakers and understand that parallel chips share the load for better performance yes?
Thanks for your help
Can the more experienced among you explain how two chips are wired in parallel or bridged?....Iv'e built a LM3886 clone thats been up and running and sounding good for almost two years.
I havn't visited the forum in quite some time and was interested in the new LM4780?(the one thats two 3886 on the same die)and how you actually,physically connect the pins for bridge and or parallel mode.
I have looked at the NSC App. notes and schematics but it's a little beyond my knowledge level.
Here's the questions I'm asking and they may sound stupid so bear with me:
1) If the LM4780,a two channel stereo chip, is composed of two LM3886 chips how do you physically combine them to make a more powerfull one channel amp.
2) I'm running 4ohm speakers and understand that parallel chips share the load for better performance yes?
Thanks for your help
I am not sure if your mean simultaneous series/ parallel bridging or just plain... I will explain the former:
You will need 2 LM4780's to create one hi-powered channel.
In Chip 1, you use the non-inverting inputs fed by a common source. On the outputs you use 0.22ohm to connect the two outputs together to form one channel. (Like paralleling discrete output devices). You have just parallel bridged this chip. This chip 1 will become the top half of your overall Paralell/Series bridge channel.
In chip 2 you you do the same as chip 1 except you use the "inverting" inputs fed by the same common source feeding Chip1. You parallel bridge the 2 outputs via 0.22 ohm on this Chip as well. This chip forms the bottom half of your bridge.
Lastly you connect the 2 chip's outputs to a load to get output (i.e. Series bridge).
Chip 1's outputs goes to the "+" of your speaker and Chip 2's outputs goes to the "-" of your speakers.
As mentioned earlier, your dissipation is 3x to 4x depending how stiff your power supply is, so you need to dissipate all that heat.
See http://cache.national.com/ds/LM/LM4780.pdf it shows both paralell and series bridging, you will need to do both.
I do not work with Chip amps so I do not know how reliable the LM will be in this application, I just mentioned how it's done since you asked.
You will need 2 LM4780's to create one hi-powered channel.
In Chip 1, you use the non-inverting inputs fed by a common source. On the outputs you use 0.22ohm to connect the two outputs together to form one channel. (Like paralleling discrete output devices). You have just parallel bridged this chip. This chip 1 will become the top half of your overall Paralell/Series bridge channel.
In chip 2 you you do the same as chip 1 except you use the "inverting" inputs fed by the same common source feeding Chip1. You parallel bridge the 2 outputs via 0.22 ohm on this Chip as well. This chip forms the bottom half of your bridge.
Lastly you connect the 2 chip's outputs to a load to get output (i.e. Series bridge).
Chip 1's outputs goes to the "+" of your speaker and Chip 2's outputs goes to the "-" of your speakers.
As mentioned earlier, your dissipation is 3x to 4x depending how stiff your power supply is, so you need to dissipate all that heat.
See http://cache.national.com/ds/LM/LM4780.pdf it shows both paralell and series bridging, you will need to do both.
I do not work with Chip amps so I do not know how reliable the LM will be in this application, I just mentioned how it's done since you asked.
on the side note:
yes. about a year ago i built a bridged T clone network on a lm4780 for the purpose of a 120W @ 8ohm application. I called it the hi3gc (high input impedance inverting gain clone), and used a 100k + 1uF capacitor for the DC blocking filter. the results were good overall, but a 70mV offset existed (50mV on one amp, -20mV on the other, 70mV combined). otherwise everything was fine. Mine was run hot -- the thing was barely specced to run according to national. however since i ran it for at most 10 minutes at high power (this was a dorm room) i figured i was fine.
the phase inversion for bridging was done by a preamp inverting amplifer.
in a good parallel configuration, each amp will see twice the load impedance -- supplying half the current. however parallel can be bad because you could have 10V on one amp, and 10.1V on the other amp. the 0.1V difference seems trivial until you concider the 0.1ohm resistance between the two amps, in which case one amp will seem like a 1A current sink! the use of output resistors and other matching networks help to limit the cross-current to low values.
yes. about a year ago i built a bridged T clone network on a lm4780 for the purpose of a 120W @ 8ohm application. I called it the hi3gc (high input impedance inverting gain clone), and used a 100k + 1uF capacitor for the DC blocking filter. the results were good overall, but a 70mV offset existed (50mV on one amp, -20mV on the other, 70mV combined). otherwise everything was fine. Mine was run hot -- the thing was barely specced to run according to national. however since i ran it for at most 10 minutes at high power (this was a dorm room) i figured i was fine.
the phase inversion for bridging was done by a preamp inverting amplifer.
in a good parallel configuration, each amp will see twice the load impedance -- supplying half the current. however parallel can be bad because you could have 10V on one amp, and 10.1V on the other amp. the 0.1V difference seems trivial until you concider the 0.1ohm resistance between the two amps, in which case one amp will seem like a 1A current sink! the use of output resistors and other matching networks help to limit the cross-current to low values.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.