Class D amp long term reliability

[dlaloum]

Looking into this I came across this article about high quality computer PSU failures. https://www.tomshardware.com/news/why-power-supplies-fail-psus,36712.html While I am sure it is possible to create a long life switching PSU to me it appears orders of magnitude more difficult than an old fashioned linear supply with only a handful of heavy duty parts and a transformer which is both inherently durable as well as providing a "shield" to protect more sensitive semi-conductors from being directly connected to the mains grid. I am still skeptical that for the quality of switching PSU we will find in audio equipment that these won't be the weak link for longevity.

My concern - based on similar experiences with SMPS's - is that Class D amp modules, are basically quite similar to SMPS's (switch more power supplies)...

And given long term failure rate of PC SMPS's I look on Class D longevity with some suspicion.... but I believe, based on first principles (so to speak), that they could be made reliable in the longer term with heavier duty heatsinking - reducing the operating temperature of all the components.

 

[antcollinet]

My concern - based on similar experiences with SMPS's - is that Class D amp modules, are basically quite similar to SMPS's (switch more power supplies)...

And given long term failure rate of PC SMPS's I look on Class D longevity with some suspicion.... but I believe, based on first principles (so to speak), that they could be made reliable in the longer term with heavier duty heatsinking - reducing the operating temperature of all the components.

From a reliability point of view they are similar to SMPS's in exactly the same way class AB amps are. In other words - not at all - or totally.

Reliability has absolutely nothing to do with the archtecture, and everything to do with designing and building for reliability.

For example - Here is a typical PC PSU:

https://www.amazon.co.uk/Legion-Silver-Silent-Power-Supply/dp/B07QK8S5MJ/ref=sr_1_58?crid=B0QVIQVUEMS3&keywords=pc+psu&qid=1686400583&sprefix=pc+psu%2Caps%2C87&sr=8-58

700W total output retailing for £116 - That is being designed down to a price.

Compare with one of the better Class D amp designs - lets say around £1000 for a similar total output power. Lets split that £1000 50:50 between amplifier part and PSU part, (since both are included) and you are still at around 4 to 5 times the £ per watt. This at least provides scope for design and manufacture for quality. (Quality and reliability don't come for free)

 

[Roland68]

My concern - based on similar experiences with SMPS's - is that Class D amp modules, are basically quite similar to SMPS's (switch more power supplies)...

And given long term failure rate of PC SMPS's I look on Class D longevity with some suspicion.... but I believe, based on first principles (so to speak), that they could be made reliable in the longer term with heavier duty heatsinking - reducing the operating temperature of all the components.

As already mentioned in my other postings, your fears are based on PC experience with consumer hardware, which only has to last a maximum of 2 years (in some countries only 1 year).

Anyone who saves on Class D amplifiers and power supplies just as much as they do on consumer PCs will also have the same experiences.
At Aliexpress you can get a TPA3255 board for $25, power supplies with about 400W for $25-30 and it works (for now).
Take a look at the XRK TPA3255 board from the neighboring forum, for example. But it also costs about $350. With a housing, if you install it yourself, from $500.
Add to that a cheap, high-quality power pack with 600-1000 watts like a MeanWell HRP or PSP for $200-450 and you have a combo that I can trust for 10 years without any problems. But as a finished product, you have to add at least the cost of distribution and packaging, cost of warranty service, retailer's profit, taxes and duties.
On the other hand, there is an Aiyima A07 for about $100 or other cheap amplifiers up to $300, which come with power supplies for $5-15.
Do you notice parallels to PC consumer hardware and professional IT hardware in the pricing?
You get what you pay for.

In industry, including PCs, a 5-10 year runtime in 24/7 operation is normal, often longer.
Here, too, I refer to professional PC power supplies, eg from Bicker Elektronik.

Take a look at HP Business (the high-end series) PCs and workstations, or Apple iMacs and workstations. Many of these devices are still running without failures after 10 years. This also puts the prices of these devices into perspective.

 

[fpitas]

I don't think we gain anything worthwhile considering bargain-basement PC power supplies.

 

[pseudoid]

Such is life in a capitalist society.

I'll come to @dlaloum defense:
I think he means that "Such is life - of tightwads - in a capitalist society."... they-get-what-they-paid-for!
ECON101: Caveat Emptor! SOC101: No whiners!

 

[fpitas]

I'll come to @dlaloum defense:
I think he means that "Such is life - of tightwads - in a capitalist society."... they-get-what-they-paid-for!
ECON101: Caveat Emptor! SOC101: No whiners!

There's an art to making a working product for the lowest cost. It's real engineering. But usually, reliability suffers.

 

[pseudoid]

There's an art to making a working product for the lowest cost. It's real engineering. But usually, reliability suffers.

... and we are NOT talking about non-capitalist societies... right?
You pay peanuts; you get monkeys. You hire "monkeys"; guess who will sweep the peanuts shells!

 

[fpitas]

... and we are NOT talking about non-capitalist societies... right?
You pay peanuts; you get monkeys. You hire "monkeys"; guess who will sweep the peanuts shells!

In reality, it's a tradeoff. If you end up with a pile of non-working product that used up thousands of dollars of parts, you're in trouble.

 

[marX]

There's an art to making a working product for the lowest cost. It's real engineering. But usually, reliability suffers.

The company I worked for previously have a company director (who is also the CTO) telling us " factor of safety is for fools". He would ask us to not use it in the engineering calculations. He's very similar to that Stockton Rush dude in a sense.

 

[DonH56]

The company I worked for previously have a company director (who is also the CTO) telling us " factor of safety is for fools". He would ask us to not use it in the engineering calculations. He's very similar to that Stockton Rush dude in a sense.

Muntzing lives.

Run away from that company.

 

[Sokel]

There are two sides to that coin - Class A /B/ AB amps have always been built with decent heatsinks, they generate heat and need the heatsinks to ensure long term reliability.

Because the Class D amps generate so much less heat, they are often sold with minimal heatsinking, or just bolted to the case - for the case to act as heatsink....

Without a means of disposing of the excess heat, the fact that they run cooler and more efficient won't eliminate the risk of overheating... and therefore abbreviated lifetime.

All things being equal - ie: using similar levels of heat management / heatsinks, a class D amp should be more reliable and have a longer lifetime due to much lower temperatures.... But manufacturers leverage the reduced heat emission, to remove expensive heatsinks or heatsink cases.... and the end result is..... variable

Plenty of thermal images of class D amps showing that some of their components are getting seriously hot.... with good heatsinking given class D's efficiency, this should not happen.

Such is life in a capitalist society.

That's not a fault by the companies that make the modules,they even state it themselves:

Now,what is adequate is open to interpetation probably in some OEMs.

 

[marX]

Muntzing lives.

Run away from that company.

Haha I wish!

I'm going back to the same company after I have completed my work here in the UK. But it's not that bad though. I would nod when he says but use the same calculation anyway. But to save cost, I would try to be more creative but won't compromise on safety. Ease of manufacturing would take a hit though.

On topic:

Why not add a small fan in the case for more air circulation?

 

[pseudoid]

Haha I wish!
On topic:
Why not add a small fan in the case for more air circulation?

That is a 4-letter word, and if you continue using it; I will be forced to press the "Unlike" button...

 

[marX]

Offtopic:

I made a Yolo V7-based Python code and ran it on a Pi 4 as my Uni project. The CPU was getting really hot (like 73C) when it ran without any forced convection but adding a tiny fan to the case, the temperature reduced to around 53C.

 

[antcollinet]

Offtopic:

I made a Yolo V7-based Python code and ran it on a Pi 4 as my Uni project. The CPU was getting really hot (like 73C) when it ran without any forced convection but adding a tiny fan to the case, the temperature reduced to around 53C.

A little bit of air flow makes a huge difference.

Problem is there is no such thing as a quiet fan. For audio you need to get the thermal design right without one.

 

[pseudoid]

A little bit of air flow makes a huge difference.
Problem is there is no such thing as a quiet fan. For audio you need to get the thermal design right without one.

It would be neat if someone could figure a miniaturized heat-pipe, radiator and/or liquid cooling implementation that becomes the integral part of an equally small Raspberry Pi case. << None of those 4-letter words. (fans)

 

[antcollinet]

It would be neat if someone could figure a miniaturized heat-pipe, radiator and/or liquid cooling implementation that becomes the integral part of an equally small Raspberry Pi case. << None of those 4-letter words. (fans)

heat pipe/water cooling both unnecessary. Both are a solution for tranferring heat from CPU to heatsink. But you still need a heatsink.

With the PI - the whole thing is small enough to have that heatsink in direct contact with the CPU - like this:

Aluminium Armour - Heatsink Case for Raspberry Pi 4

Your Raspberry Pi 4, as small and powerful as it is, still requires passive cooling and case protection to function efficiently and maintain its computing power.

thepihut.com

But in any case - for audio applications if not doing DSP - additional cooling is not necessary.

 

[GXAlan]

The company I worked for previously have a company director (who is also the CTO) telling us " factor of safety is for fools". He would ask us to not use it in the engineering calculations. He's very similar to that Stockton Rush dude in a sense.

The only time that sentence makes sense is if it is followed by a few more sentences.

“Factor of safety is for fools. It gives you a false sense of security. You have to consider novel mechanisms of failure that aren’t accounted for by a simple multiplier.”

 

[pseudoid]

" factor of safety is for fools"

Are you referring to the 'derating of components' during circuit design, analysis and simulation (e.g. MonteCarlo, etc.)?

 

[marX]

Are you referring to the 'derating of components' during circuit design, analysis and simulation (e.g. MonteCarlo, etc.)?

Not circuit design in particular. I'm primarily a mechanical engineer. But our design team consists of electrical and electronics engineers as well.

The CTO would often comes up to us say this to challenge us not to factor in any safety buffer while designing new machines. His aim obviously to save money but he hides it behind the statement that we are incapable of designing stuff without adding a factor of safety.

But adding a factor of safety is an industry practice (in mechanical design at least).