100W LM3886 Parallel Stereo Power Amplifier
This amplifier is based on the PA100 parallel amplifier detailed
in National Semiconductor's application note - AN1192.
Since my DIY speaker is 4-ohm and somewhat difficult to drive, I want to
have a more powerful amplifier to match with it. Therefore I designed
this amplifier which uses two LM3886 per channel, in parallel circuit.
This amp can deliver about 50W into a 8-ohm speaker and 100W into a
4-ohm speaker. This is a stereo amplifier and therefore 4 LM3886s are
used. The LM3886 circuit is in a non-inverted configuration, so the
input impedance is determined by the input resistor R1, i.e. 47k. The
680 ohm and 470pF resistor capacitor filter network is used to filter
out the high frequency noise at the RCA input. The 220pF C4 and C8
capacitors are used to shot out the high frequency noise at the LM3886
input pins.
I used high quality audio grade capacitors at several locations: 1uF
Auricap at the input for DC blocking, 100uF Blackgate for C2 and C6, and
1000uF Blackgate at the supply filter.
The PCB is designed in a way that the power ground is separated
from the signal ground, as you can see from the below layout. The signal
ground is located in the middle and surrounded by the power ground.
There is a thin trace near C5 connecting them. The PCB layout is done by
using PADS PowerPCB 5.0. I think it is a powerful layout software.
After finished the layout, I send it to a PCB manufacturer in HK to make
the PCB. After the PCB is made, I found some of the drill holes are not
large enough.... I have to make it bigger manually.
Finished PCB.
The 20k and 1k resistors are hand matched to 0.1%. For the output resistors, I used six 0.5W 1% 1ohm resistors in parallel per 3886 output instead of one 3W resistor because 3W 1% resistors are hard to find.
I used the insulated version - LM3886 TF, so that I can mount it directly to the case and heat sink with thermal compound.
Coupling capacitor is Auricap 1uF 450V. I used this high quality capacitor because it is in the main signal path.
The HF filter capacitors are Silver Mica 47pF and 220pF.
The power supply filter used is Blackgate 1000uF 50V.
C2 and C6 are Blackgate 100uF 50V. For better result, the bi-polar version should be used. However I am not using it because the bi-polar blackgate is too big to fit into my PCB.
The 680 ohm + 470pF filter network is installed at the RCA. This help filtering the high frequency noise before it gets onto the board.
The WIMA 0.1uF supply decoupling capacitors are soldered directly on the pins of LM3886 at the back of the PCB. This helped to remove some high frequency noise.
The 3886 are mounted on a 3/8 inch aluminum plate then to the case. Outside of the case I used 3 PC CPU heatsinks. I used Arctic Silver thermal compound between the aluminum layers to improve heat conductivity.
With all these big heatsinks, it only get slightly warm when listen at normal volume.
The power supply used is a regulated power supply. I used 10000uF per rail before the LT1083 regulator. After the regulator, I have 100uF on the regulator board. The advantage of using regulator is that the power supply ripple voltage is removed. If power regulation is not used, I can hear very little 50/100Hz hum from the speaker.
Power Supply Schematic :
The high current MUR860 diode is used to ensure high current flow.
The voltage regulator used is LT1083, it can provide about 8A of current.
Transformer used here is a 500VA 2x 25V. The power supply is then regulated by 2 LT1083, after the regulation, the voltage is 30V.
Will consider to use this supply circuit instead, the TIP2955 is capable of providing 15A of current:
I did some DC measurement and the result is quite good, I got 7 mV of DC offset at the speaker terminal. The voltage difference between the output of the 2 chips is less then 1 mV.
The sound of this amplifier is similar to my LM3875 amplifier, which is very clean and detail. It has no hum, no hiss and no noise. Compared to the LM3875 Gainclone, this amp can deliver twice the power to my 4-ohm speaker, and it improves the dynamics and bass punch a lot.
Finished PCB.
The 20k and 1k resistors are hand matched to 0.1%. For the output resistors, I used six 0.5W 1% 1ohm resistors in parallel per 3886 output instead of one 3W resistor because 3W 1% resistors are hard to find.
I used the insulated version - LM3886 TF, so that I can mount it directly to the case and heat sink with thermal compound.
Coupling capacitor is Auricap 1uF 450V. I used this high quality capacitor because it is in the main signal path.
The HF filter capacitors are Silver Mica 47pF and 220pF.
The power supply filter used is Blackgate 1000uF 50V.
C2 and C6 are Blackgate 100uF 50V. For better result, the bi-polar version should be used. However I am not using it because the bi-polar blackgate is too big to fit into my PCB.
The 680 ohm + 470pF filter network is installed at the RCA. This help filtering the high frequency noise before it gets onto the board.
The WIMA 0.1uF supply decoupling capacitors are soldered directly on the pins of LM3886 at the back of the PCB. This helped to remove some high frequency noise.
The 3886 are mounted on a 3/8 inch aluminum plate then to the case. Outside of the case I used 3 PC CPU heatsinks. I used Arctic Silver thermal compound between the aluminum layers to improve heat conductivity.
With all these big heatsinks, it only get slightly warm when listen at normal volume.
The power supply used is a regulated power supply. I used 10000uF per rail before the LT1083 regulator. After the regulator, I have 100uF on the regulator board. The advantage of using regulator is that the power supply ripple voltage is removed. If power regulation is not used, I can hear very little 50/100Hz hum from the speaker.
Power Supply Schematic :
The high current MUR860 diode is used to ensure high current flow.
The voltage regulator used is LT1083, it can provide about 8A of current.
Transformer used here is a 500VA 2x 25V. The power supply is then regulated by 2 LT1083, after the regulation, the voltage is 30V.
Will consider to use this supply circuit instead, the TIP2955 is capable of providing 15A of current:
I did some DC measurement and the result is quite good, I got 7 mV of DC offset at the speaker terminal. The voltage difference between the output of the 2 chips is less then 1 mV.
The sound of this amplifier is similar to my LM3875 amplifier, which is very clean and detail. It has no hum, no hiss and no noise. Compared to the LM3875 Gainclone, this amp can deliver twice the power to my 4-ohm speaker, and it improves the dynamics and bass punch a lot.
2 komentar — Skip to Comment
Can you please share PCB parts layout with me
onesimus.tozeyana@gmail.com
Strange experienced.. Everyone kept telling me not to go over 84 volt or else the chip would die. I have many of these little LM3886T, so I took the risk in testing one of them. I have a power supply that can produce +-36ct or +-48ct at 4amp. I am using the same size heatsink as in the picture without the fan. I first tested with +-36 parallel of 6 speakers, probably some loss through the wires but this brought it down to 2ohm. It sounds great playing "AC/DC Who made Who". Pretty loud with lots of punch. The heatsink was warm but nothing hot. I then switched it to the +-48, this is 96v total exceeded 12v, thinking it will die or not produce any sound because of the SPIKE. Well, it played the same "Who made Who" louder with stronger punches. I left it playing for days, thinking it could only max out for a short amount of time. It didn't give up. The heatsink was hot, but wasn't melting crayon hot. I measured the voltage at the chip's pins, taking in 96volt. when I turned it up full power, the voltage dropped a bit, but it was still 94volt. Has anyone tried higher volts?
Posting Komentar — or Back to Content