I reconsidered the LM3886 version as I want to compare with the error corrected type. I noticed that I had done error input ground wiring. The sound is great, very interesting to listen to. I passed the most difficult music to reproduce , the Requiem of Verdi, I got full resolution. You find bellow the original in mp3 and what I recorded with my modest phone in m4a. Try the original on your Amp, if you hear distorted, curse the mp3 that helps you sleep.
There is startup and shutdown plop to adjust.
To mention that my 2 way speaker is 106db/1w and the recorder is 1.5m from the mouth of a giant 0.7m x 0.7m diameter horn which makes the classAB amp run in its worst operating range.
There is startup and shutdown plop to adjust.
To mention that my 2 way speaker is 106db/1w and the recorder is 1.5m from the mouth of a giant 0.7m x 0.7m diameter horn which makes the classAB amp run in its worst operating range.
Last edited:
To get best bass from LM3886 the magnetic field of the output, Vcc, Vss should cancel. For that the Vcc should be adjacent to output that is pin2 which in not connected. The +V must be supplied by pin2 and at the root of the chip is linked to pin1.
This why the lm3886 must be customized. At the root of pin1 and 2 a magnetic lead from electrolytic capacitor links them as shown in the picture bellow. The Vcc, Vss tracks come in parallel from the bottom side to pin2 , 4 and the ground from the upper side to join the output and bifurcate sideways in parallel. An ceramic 100nF 100v smd can be installed between the Vcc and Vss at the pins.
This why the lm3886 must be customized. At the root of pin1 and 2 a magnetic lead from electrolytic capacitor links them as shown in the picture bellow. The Vcc, Vss tracks come in parallel from the bottom side to pin2 , 4 and the ground from the upper side to join the output and bifurcate sideways in parallel. An ceramic 100nF 100v smd can be installed between the Vcc and Vss at the pins.
Attachments
Hello HAYK,
i was looking at your schematics and i noticed what you are actually doing (posting this just for information). it's a simple VAS with its own NFB (from the output of the complete amp) in front of a chip amp.
you could look at it like two seperate gain stages in composite configuration.
maybe try to make the first gain stage with long tail pair input and an EF VAS to improve the input distortion and higher open loop gain
i was looking at your schematics and i noticed what you are actually doing (posting this just for information). it's a simple VAS with its own NFB (from the output of the complete amp) in front of a chip amp.
you could look at it like two seperate gain stages in composite configuration.
maybe try to make the first gain stage with long tail pair input and an EF VAS to improve the input distortion and higher open loop gain
Last edited:
I love the design of this amp with the 3886 + fet input. My only apprehension is the floating PS ground being neg speaker output, which essentially makes it cap coupled.
There has to be a way to DC couple it and still retain the midrange magic you describe. It would be better to get the PS caps out of the speaker signal path, since the caps can be an avoidable source of HD at lower frequencies.
The slight bump in the LF area (warmer mids you describe) is probably caused by increased resistance and H2 generated from caps in series instead of solid ground. Have you tried bridged output with second chip or even 4x 3886 and only single supply rail? You could even do fully balanced input this way, probably better common mode rejection too.
There has to be a way to DC couple it and still retain the midrange magic you describe. It would be better to get the PS caps out of the speaker signal path, since the caps can be an avoidable source of HD at lower frequencies.
The slight bump in the LF area (warmer mids you describe) is probably caused by increased resistance and H2 generated from caps in series instead of solid ground. Have you tried bridged output with second chip or even 4x 3886 and only single supply rail? You could even do fully balanced input this way, probably better common mode rejection too.
There is a market for higher end low power chip based amps. Most of the time, with active speakers, the only basic choice you have is a compact, inexpensive chip amp, which is usually the weakest link.
A well designed chip amp could be very useful for a higher power active 3 way speaker. Most of the time, the noise alone can spoil it when running high sensitivity CD + WG. A higher quality, low THD+N modular amp option would be great.
A bridged, parallel output version may be enough for MF and LF. A single high current 35V rail could do 100W into 8 ohms this way.
A well designed chip amp could be very useful for a higher power active 3 way speaker. Most of the time, the noise alone can spoil it when running high sensitivity CD + WG. A higher quality, low THD+N modular amp option would be great.
A bridged, parallel output version may be enough for MF and LF. A single high current 35V rail could do 100W into 8 ohms this way.
@ImpatientIcecream An LTP with CCS would also allow for balanced input. Incorporating NFB into that can further improve THD and drive the chip better, converting high input impedance, lowering noise.
The purpose of the MOSFET is to simulate the transfer function of single ended triode. It is not a VAS as the load, the power opamp negative input is very low impedance. The MOSFET acts as V/I converter and the power opamp as I/V converter. The total NFB @20khz is 70db instead of 24db for the LM3886 alone with26db gain. In post 1 you can see the noise is about 46db lower than stand alone.
The single supply with floating ground, I keep for tweeter amplifier with Chinese TDA2050 that uses 48v 0.5A CCTV supply. With TDA7293-94 and lm3886, dual 36v supply is used.
The PCB for TDA7293 is going to be modular double parallel chips where the user decides to use a single or double. With LM3886 I will keep it single. I could make a bridged all keeping the even harmonics untouched on simulator, I posted it more than a year ago on EEVBLOG website where I started this subject. I want to say that I need the circuit to be low cost using if possible the Chinese versions, the TDA7293-4 work good and cost about $1. For LM3886 I think the original need to be used and is not cheap, so single enough.