Designing SMD AMplifier Board, please suggest & Comment ?

[dashpuppy]

This isn't totally done yet but i'm looking for some suggestions and improvments remarks maybe some positive and bad feedback ?  This is the driver stage of the amplifier, the outputs are on another board..

[dashpuppy]

NO advice from all the good guys on here ?

[ejeffrey]

You haven't specified your goals/specs, provided a schematic, or talked about why you did it this way or what else you considered.  So there isn't much for anyone to comment on, and there is a fairly big barrier to trying to reverse engineer a schematic from your layout.

[dashpuppy]

You haven't specified your goals/specs, provided a schematic, or talked about why you did it this way or what else you considered.  So there isn't much for anyone to comment on, and there is a fairly big barrier to trying to reverse engineer a schematic from your layout.

Well that makes sense  

The goal is to keep it compact. It's an Nelson Pass Aleph 2 SMD PCB layout, the board shown is the driver stage. I have the IRF9610's in the center with holes in the middle of the 3 so i can push a pair of small bolts through and mount to a small thin heatsink and clamp against the fets. They don't get tooo too hot, but they get warm, if i pull the heat away they will last longer

The board uses all SMD devices and using 0204 MELF resistors @ 1% precesion.  The power resistors are at the top because those are 3watt SMD dale units. They don't get hot..

Schematic Attached


Output Board 

[dashpuppy]

TTT

[SiliconWizard]

I don't really like the trace between R52 and R7. It's a mess and gets too close to a plated hole for my taste.
I would probably route it on the bottom layer.

[dashpuppy]

I don't really like the trace between R52 and R7. It's a mess and gets too close to a plated hole for my taste.
I would probably route it on the bottom layer.

YUP ! that part is a nasty mess LOL !  I have better plans for that area, like you suggested under side it wil go !

[schmitt trigger]

Fill all the unused real state with a low impedance ground plane.

[SiliconWizard]

Yep.
Just one tip: provide sufficient clearance around wide traces if you fill the void with a ground plane (which you should do). Depending on the PCB class and the manufacturer, you may get subtle short-circuits between wide traces and the ground plane if you don't give enough clearance. Got bitten with this once

[Audioguru]

I notice that your schematic is completely wrong:

[dashpuppy]

Fill all the unused real state with a low impedance ground plane.

I will be using a Ground plane Here is a screen shot.

[dashpuppy]

I notice that your schematic is completely wrong:

[Audioguru]

The Nelson Pass Aleph 2 amplifier is a room heater (class-A dissipation is 300W). It produces an output of 100W into 8 ohms. Its distortion is fairly high compared to modern amplifiers.

[dashpuppy]

The Nelson Pass Aleph 2 amplifier is a room heater (class-A dissipation is 300W). It produces an output of 100W into 8 ohms. Its distortion is fairly high compared to modern amplifiers.

Actually you are incorrect.

I know what it is, I have built TONS of nelson's designs.

[schmitt trigger]

Yep.
Just one tip: provide sufficient clearance around wide traces if you fill the void with a ground plane (which you should do). Depending on the PCB class and the manufacturer, you may get subtle short-circuits between wide traces and the ground plane if you don't give enough clearance. Got bitten with this once

Very Correct!
Most PCB design packages allow to set the default clearance between planes and other traces.

[dashpuppy]

Yep.
Just one tip: provide sufficient clearance around wide traces if you fill the void with a ground plane (which you should do). Depending on the PCB class and the manufacturer, you may get subtle short-circuits between wide traces and the ground plane if you don't give enough clearance. Got bitten with this once

Very Correct!
Most PCB design packages allow to set the default clearance between planes and other traces.

Yup i have a default gap between all the pcb traces. And have Eagle to remind me that things get to close to each other..

[dashpuppy]

I notice that your schematic is completely wrong:

Sorry you are correct, i fixed this!  Thought you were thread crapping Sent pm

[dashpuppy]

Totally re-vamped the board.

[schmitt trigger]

There are a couple of taces, close by the N$2 label in yellow, which could benefit from a little more clearance.

[dashpuppy]

There are a couple of taces, close by the N$2 label in yellow, which could benefit from a little more clearance.

Thanks

Im thinking about trashing this and trying again .. Wish i could pass the file around and see what others can do..

[Zero999]

The Nelson Pass Aleph 2 amplifier is a room heater (class-A dissipation is 300W). It produces an output of 100W into 8 ohms. Its distortion is fairly high compared to modern amplifiers.

Actually you are incorrect.

I know what it is, I have built TONS of nelson's designs.

I can't comment on the distortion because I haven't seen the whole schematic or done any tests/simulations, but he's almost right about the power dissipation. Note I said almost right, because an efficiency of 30% is optimistic. The theoretical maximum efficiency for a class A design, without a transformer, is 25%. In reality, it will be much lower than, probably around 2% under typical operating conditions.

No doubt there are plenty of cheaper, modern class AB and even class D amplifier designs, capable of the same output power level, using a fraction of the input power of this design, with similar or better distortion and noise figures.

[dashpuppy]

The Nelson Pass Aleph 2 amplifier is a room heater (class-A dissipation is 300W). It produces an output of 100W into 8 ohms. Its distortion is fairly high compared to modern amplifiers.

Actually you are incorrect.

I know what it is, I have built TONS of nelson's designs.

I can't comment on the distortion because I haven't seen the whole schematic or done any tests/simulations, but he's almost right about the power dissipation. Note I said almost right, because an efficiency of 30% is optimistic. The theoretical maximum efficiency for a class A design, without a transformer, is 25%. In reality, it will be much lower than, probably around 2% under typical operating conditions.

No doubt there are plenty of cheaper, modern class AB and even class D amplifier designs, capable of the same output power level, using a fraction of the input power of this design, with similar or better distortion and noise figures.

Power dispation is correct. BUT people don't buy a Class A amplifier for it's effeciancy. They buy / BUILD them becuse they sound good

I've tried Class d and many high end brands IMO they all sound like CRap! IMO They should be used for one thing, and one thing only. Subwoofers thats it.

[Audioguru]

The Nelson Pass Aleph 2 amplifier costs $14,000 dollars (!!) a pair. It is a box covered with heatsinks.
A reviewer said it caused his electricity bill to increase 40%. The review was full of audiophool remarks.
Its spec's say its distortion is <1% which is very high but the distortion is mostly second and other even harmonics that sound musical like an old vacuum tubes amplifier.

[dashpuppy]

The Nelson Pass Aleph 2 amplifier costs $14,000 dollars (!!) a pair. It is a box covered with heatsinks.
A reviewer said it caused his electricity bill to increase 40%. The review was full of audiophool remarks.
Its spec's say its distortion is <1% which is very high but the distortion is mostly second and other even harmonics that sound musical like an old vacuum tubes amplifier.

Name a good class A amplifier that doesn't suck up power. They all do, Infact the Krell KSA 50 & 100 eats even more power.  I'm a Nelson Pass fan so thats what I use

[Zero999]

The Nelson Pass Aleph 2 amplifier is a room heater (class-A dissipation is 300W). It produces an output of 100W into 8 ohms. Its distortion is fairly high compared to modern amplifiers.

Actually you are incorrect.

I know what it is, I have built TONS of nelson's designs.

I can't comment on the distortion because I haven't seen the whole schematic or done any tests/simulations, but he's almost right about the power dissipation. Note I said almost right, because an efficiency of 30% is optimistic. The theoretical maximum efficiency for a class A design, without a transformer, is 25%. In reality, it will be much lower than, probably around 2% under typical operating conditions.

No doubt there are plenty of cheaper, modern class AB and even class D amplifier designs, capable of the same output power level, using a fraction of the input power of this design, with similar or better distortion and noise figures.

Power dispation is correct. BUT people don't buy a Class A amplifier for it's effeciancy. They buy / BUILD them becuse they sound good

I've tried Class d and many high end brands IMO they all sound like CRap! IMO They should be used for one thing, and one thing only. Subwoofers thats it.

Or people might buy/build them because they believe they sound better. It highly subjective and could even be the placebo effect, if the amplifiers are good enough and are operated well within their power ratings. See the study linked below, which tested whether audiophiles could tell the difference between three different amplifiers: one very old solid state design, a new (back then) solid state design and a valve/tube design. The volume was set the same, taking care not to overload the amplifiers. The same speakers were used and someone, in another room, switched between different amplifiers, randomly. The end result was no one could tell the difference, any better than chance, meaning all of the amplifiers sounded the same!
http://www.keith-snook.info/wireless-world-magazine/Wireless-World-1978/Valves%20versus%20Transistors%20DCD.pdf

The senses can easily be tricked. Aesthetics, brand and marketing are very influential. It's almost certain that, if I put the same amplifier circuit inside two enclosures, one which looks nice and high quality and another which looks cheap and nasty and asked people to rate how they sound, they'd pick the nicer looking amplifier, as the one which sounded better. Similar experiments have been conducted with the same food on different plates and in different packaging, with similar results: nicer packaging and crockery make it taste better!

Of course in real life, people do drive their amplifiers into clipping (overload) and that's where the differences start to appear: an old valve design might distort subtly, in a pleasing manner, whilst a modern transistor amplifier will harshly clip the signal, resulting in nasty distortion. This is why many musicians prefer old valve/tube amplifiers and your class A amplifier might sound similar.

It's true that class D amplifiers weren't very good when they were originally introduced, but some modern ICs have imperceptibly low levels of distortion, when operated within their ratings. Of course, lower distortion doesn't mean it will sound better to everyone. As I mentioned before, it's highly subjective. Some people like certain types of distortion.

If you want low distortion, your best bet is to over-rate the amplifier so it's never driven into clipping. The main source of distortion then becomes the speakers and room acoustics, which are far more important, than the amplifier design.