I found this scheme on the net and tried to carry out the project with 4 LM3886 chips, the result obtained is very pleasant, the pcb groups the two amplifier stages and two psu. The sound is very different from that obtained from the single Lm3386, with this stage you can see better driving at low frequencies and excellent transparency on medium-high frequencies.
pcb also accept an antibum-protection.
if interest pcb is available in 3pz...with BOM
thanks
pcb also accept an antibum-protection.
if interest pcb is available in 3pz...with BOM
thanks
Attachments
The purpose of paralleling so many chips is to drive very low impedance speakers or to bridge 2 amps and really take advantage of the amp with any speaker.
First: Nice work!
Maybe some ideas? It is already enough for bridging, if you parallel 3x LM3886. Even the bridged amp will play down to 2 Ohm. 4 of them parallel without a bridge is a little overkill.
With such a bridge of 6x LM3886 you can run full voltage and have 4 times the output of a single chip. Even if you do not need such power, sound will improve with the enhanced headroom.
scafas is probably right about one thing, more chips in parallel sound better, in most cases. The speaker control gets better because of the lower output impedance.
I´m sure his amp will work fine, but it may be an advantage to make the PCB a little more compact. Then, have some additional film cap´s near each chips legs. Should be better than only electrolitic´s.
Maybe some ideas? It is already enough for bridging, if you parallel 3x LM3886. Even the bridged amp will play down to 2 Ohm. 4 of them parallel without a bridge is a little overkill.
With such a bridge of 6x LM3886 you can run full voltage and have 4 times the output of a single chip. Even if you do not need such power, sound will improve with the enhanced headroom.
scafas is probably right about one thing, more chips in parallel sound better, in most cases. The speaker control gets better because of the lower output impedance.
I´m sure his amp will work fine, but it may be an advantage to make the PCB a little more compact. Then, have some additional film cap´s near each chips legs. Should be better than only electrolitic´s.
Last edited:
each lm3886 already has its capacitors near the power supply pins, you can see it from the photo. I think putting this bridge amplifier is a purely experimental thing seen and considered that in real life at home we will never have 2 ohm speakers. There is only one pcb available for construction, whoever wants to use it can communicate it to me via private message. thank you all . P.S. the amp sounds really good!😉
comparing to a single chip amplifier :
bridge parallel will give more power in all speakers , from 2 ohms to 16 ohms,
when just paralleling will give you more power into low impedance speakers 1-2-4 ohms only.
Making amps able to drive very low loads has usually no trade of in sound quality with "regular" 4-8 Ohm loads.
If you use it full range, parallel output transistors add slam to the low frequency and make the amp sound more relaxed.
This is a very personal experience, as I build a lot of scale able amps.
IMO with A/B amps, if you have clean 200 watt / 4 Ohm available, even relative small 2-way speakers sound more sorted out. Weaker amps often have this tendency to sound fantastic with some special music, but loose the grip when you play compact, multi instrument music, like whole orchestras or live bands.
Maybe populate your second PCB, use one transformer and power supply for each board and get two of these:
Fully-Differential Audio Operational Amplifier OPA1632 Module ADC Driver Board | eBay
You may be surprised how the amp improves, bridged.
If you use it full range, parallel output transistors add slam to the low frequency and make the amp sound more relaxed.
This is a very personal experience, as I build a lot of scale able amps.
IMO with A/B amps, if you have clean 200 watt / 4 Ohm available, even relative small 2-way speakers sound more sorted out. Weaker amps often have this tendency to sound fantastic with some special music, but loose the grip when you play compact, multi instrument music, like whole orchestras or live bands.
Maybe populate your second PCB, use one transformer and power supply for each board and get two of these:
Fully-Differential Audio Operational Amplifier OPA1632 Module ADC Driver Board | eBay
You may be surprised how the amp improves, bridged.
one pcb available !!!!
last pcb available:
diy pcb amplificatore LM3886 JEFF ROWLAND CLONE power amplifier | eBay
last pcb available:
diy pcb amplificatore LM3886 JEFF ROWLAND CLONE power amplifier | eBay
There is an app-note from TI describing this in all possible aspects
https://www.ti.com/lit/an/snaa021b/snaa021b.pdf
new board available
5 pz. bare board available with black solderesist .
link:
diy pcb LM3886 jeff rowland clone amplifer circuito stampato | eBay
5 pz. bare board available with black solderesist .
link:
diy pcb LM3886 jeff rowland clone amplifer circuito stampato | eBay
I do not think this version can drive low impedance better than a single LM4886. The JR Concentra used three chips in parallel for bridged mode to better handle 3 - 4 ohm speakers. Two in parallel bridged gives four times the power of a single chip into four ohms.
Agree that bridged LM3886 can be better than single. Also parallel can be better. Never tried bridged parallel.
Currently using bridged LM3886 with efficient 8 ohm speakers. With a crossover and active sub. That addresses the doubling of output impedance. But would love to try the six LM3886 version of Jeff Rowland.
Agree that bridged LM3886 can be better than single. Also parallel can be better. Never tried bridged parallel.
Currently using bridged LM3886 with efficient 8 ohm speakers. With a crossover and active sub. That addresses the doubling of output impedance. But would love to try the six LM3886 version of Jeff Rowland.
There was a spreadsheet available from National that let you calculate various configurations. As I recall, two chips in parallel will do full power into 4 ohms at up to +/-30VDC, whereas three chips will go to +/-35VDC. I haven't got the sheet to hand, but adding a fourth chip seems unnecessary?