41 volts no load at maximum line is probably ok - what does it drop to when playing music LOUD?
At max voltage I don’t think I’d run 4 ohm speakers hard. Not a good idea with STK modules either. Reduce the volts when running 4 ohms. 3886 is probably just as reliable at STKs were even at max voltage - STKs didn’t last forever even when they were common. But if you didn’t beat on them too hard they were reasonable. The monolithic chip amps are only going to fail from repeated thermal overstress so minimizing that goes a long way. STKs failed from that too, but also had another failure mode of die contamination due to the poor moisture seal on the air cavity package. That was totally random and goes away entirely with a 3386 or discrete solution.
If you DIY your own STK replacement, you can make the input stage anything you like. Balanced differential with XLR inputs if you like. Those little SMD input stage components are easy to figure out where to put, and aren’t usually a source of failure. It’s figuring out how to fit in 4 regular power transistors and the bias regulator, on something you can mount in the same space that’s the challenge. Anything custom up front adds very little overhead.
At max voltage I don’t think I’d run 4 ohm speakers hard. Not a good idea with STK modules either. Reduce the volts when running 4 ohms. 3886 is probably just as reliable at STKs were even at max voltage - STKs didn’t last forever even when they were common. But if you didn’t beat on them too hard they were reasonable. The monolithic chip amps are only going to fail from repeated thermal overstress so minimizing that goes a long way. STKs failed from that too, but also had another failure mode of die contamination due to the poor moisture seal on the air cavity package. That was totally random and goes away entirely with a 3386 or discrete solution.
If you DIY your own STK replacement, you can make the input stage anything you like. Balanced differential with XLR inputs if you like. Those little SMD input stage components are easy to figure out where to put, and aren’t usually a source of failure. It’s figuring out how to fit in 4 regular power transistors and the bias regulator, on something you can mount in the same space that’s the challenge. Anything custom up front adds very little overhead.
Datasheet limits are NOT to be used as DESIGN values, at all.
Just the above mentioned +/-10% variation in mains voltage will murder those running too close to the precipice edge.
Besides "maximum" values mentioned, Datasheet suggests +/-35V as design value into 8 ohm loads and +/-28V into 4 ohm.
Maybe they know a thing or two about their own product? 😉 😉
Personal experience: I make Guitar amps "since forever" and am quite known in Argentina.
Some 5 years ago, a new brand appeared, very good looking amps (they imitated Mesa Boogie appearance) an d, more worrying, good SOUND.
All were SS but sound was quite "tubey", and they began to sell a lot, chipping away some of my customers, which was worrying.
And then, Musicians started complaining in Forums "My xxxx amp blew on stage"
or even worse: "during a rehearsal, and I wasn´t even playing loud".
Eventually 20% or more blew 😱 , Factory was swamped with warranty repairs .... which blew again a couple Months later .... until they closed doors.
By then some 50% had failed, a catastrophic deal breaking event.
A few months later, a guy from a band who also used some of my amps called: "please Juan, I know this one is not yours, but Factory disappeared and nobody wants even to touch it, would you please take a look at it?"
I accepted, mainly because I was curious about it, so a look under the hood would be "interesting"
Power amp was a boosted LM3876 (earlier version of 3886) , straight driving a pair TIP35C/TIP36C, so it actually had a light duty job, since main current into speaker was provided by power transistors, it only had to supply base current.
And why did they explode, leaving bits of plastic shrapnel all over the chassis?
Because they were fed crazy +/-47V rails.
Yup, "factory approved" 94V total. 😱
My solution?
I had to use available "real" LM3886 (I had a couple TI factory samples), had to cut/paste tracks because pinout is not the same, plus original PCB had already been murdered by previous Techs, but main solution was to add a bucking 24V transformer (we have 220V mains), so rails dropped to way more sensible +/-42V (so 84V end to end).
It is still working for over 5 years now.
Just the above mentioned +/-10% variation in mains voltage will murder those running too close to the precipice edge.
Besides "maximum" values mentioned, Datasheet suggests +/-35V as design value into 8 ohm loads and +/-28V into 4 ohm.
Maybe they know a thing or two about their own product? 😉 😉
Personal experience: I make Guitar amps "since forever" and am quite known in Argentina.
Some 5 years ago, a new brand appeared, very good looking amps (they imitated Mesa Boogie appearance) an d, more worrying, good SOUND.
All were SS but sound was quite "tubey", and they began to sell a lot, chipping away some of my customers, which was worrying.
And then, Musicians started complaining in Forums "My xxxx amp blew on stage"
or even worse: "during a rehearsal, and I wasn´t even playing loud".
Eventually 20% or more blew 😱 , Factory was swamped with warranty repairs .... which blew again a couple Months later .... until they closed doors.
By then some 50% had failed, a catastrophic deal breaking event.
A few months later, a guy from a band who also used some of my amps called: "please Juan, I know this one is not yours, but Factory disappeared and nobody wants even to touch it, would you please take a look at it?"
I accepted, mainly because I was curious about it, so a look under the hood would be "interesting"
Power amp was a boosted LM3876 (earlier version of 3886) , straight driving a pair TIP35C/TIP36C, so it actually had a light duty job, since main current into speaker was provided by power transistors, it only had to supply base current.
And why did they explode, leaving bits of plastic shrapnel all over the chassis?
Because they were fed crazy +/-47V rails.
Yup, "factory approved" 94V total. 😱
My solution?
I had to use available "real" LM3886 (I had a couple TI factory samples), had to cut/paste tracks because pinout is not the same, plus original PCB had already been murdered by previous Techs, but main solution was to add a bucking 24V transformer (we have 220V mains), so rails dropped to way more sensible +/-42V (so 84V end to end).
It is still working for over 5 years now.