Class D amp long term reliability

[EJ3]

Fan subject.
I always wonder if the benefits or using a fan and having the components a bit cooler is not countered by bringing more dust to the inside and risking an electrical failure in the long term.

I am talking about the case when a fan is not absolutely necessary, obviously.

Anyone with an informed opinion?

Note that opening the device to clean the dust from time to time also brings some additional risk.

There are many ways to filter the air that is being pulled in (It's not like you need sub-micron filtering.
Sometimes all you need is a dryer sheet cut to fit & taped into place.
A layer or 2 of gauze.
If you want to get serious about it: a couple of layers of gauze with 2 more layers that have been sprayed lightly with the (normally used on cars) K&N air filter (a very tacky substance that catches most any minute solid that tries to pass through it at the low flow rates we are using here {compared to a car at wide open throttle).
Why the initial 2 layers of gauze not having this oil on them: they will absorb some of the tacky oil (which is not something that drips), but sort of migrates a bit over time.
So I would suggest another layer of gauze over the top.
Now you have a 5 layer gauze filter sandwich that will keep dust, pollen, etc from the inside of whatever you are pulling air though.
There are many ways to defeat the dust.
For me, I use are filters on my home ventilation system made by the company that supplied the ones for the NASA Space Shuttle program.
Officially they are MERV 8 in airflow but they filter as if they are MERV 13 while flowing like MERV 8 filters (The Name of the company is :Colorfil.
I also use 2 Honeywell HPA 3200's with HEPA, SMOKE & VOC filtration in them. Each one covers up to 520 square feet of a 8 foot ceiling building.
(My use case, at the moment is 1100 SQ. Feet [but that includes a small screened in patio], so we are filtering under 1000 SQ. feet).
These units are available at or through Home Depot, though you will likely have to order the additional VOC part.
Without that they are still a HEPA filtration level of filtering.
It's not hard or expensive to keep ones place so that you don't have to dust & vacuum every week.
Another thing that we do is have slippers inside of all doorways to the outside, so that, the moment that you come in, you change from your outdoor shoes.
Yep, you guessed it, my wife is Asian. But it's a good policy IMHO.
EJ3

 

[Yorkshire Mouth]

What would be likely to lengthen/shorten the life of a Class D amp?

If you sat one in a box for 5 years, I presume it wouldn’t wear out.

If you use it for four hours a week, is it likely to last longer than 40 hours a week? And would it last 10 times longer?

Would running it at full welly wear it out faster than 50%?

If so, if you’re using it as a power amp, and turn the volume to full, and control from a pre-amp, does that count as using it at full tilt, or dies that depend on the output/how hard you’re driving your speakers?

Many thanks.

 

[antcollinet]

First as I've said here many times - there is nothing about the topology of class D that makes them any more or less reliable than any other type of amp.

It one hundred percent comes down to the quality of the design - how close to (particularly thermal) limits components are run at - and that is the same whatever class of amp you are talking about.

Having said that, and to answer your question - heat is the enemy. Heat x time in particular. Running at high output levels generates more heat. But again - the same for any class of amp. It is quite possible for a well designed class D to have much better reliability than a poorly designed class A or class AB.

Oh - and in the "heat is the enemy" - class D putting out less additional heat at full tilt gives the designers the *possibility* of designing a more reliable amp. If they choose to take it.

 

[Yorkshire Mouth]

First as I've said here many times - there is nothing about the topology of class D that makes them any more or less reliable than any other type of amp.

It one hundred percent comes down to the quality of the design - how close to (particularly thermal) limits components are run at - and that is the same whatever class of amp you are talking about.

Having said that, and to answer your question - heat is the enemy. Heat x time in particular. Running at high output levels generates more heat. But again - the same for any class of amp. It is quite possible for a well designed class D to have much better reliability than a poorly designed class A or class AB.

Oh - and in the "heat is the enemy" - class D putting out less additional heat at full tilt gives the designers the *possibility* of designing a more reliable amp. If they choose to take it.

Thank you.

So if you don’t run it at full tilt, and your listening sessions are not so long, and less frequent, then your Class D (or Class Anything) amp is less prone to giving up the ghost, all else being equal?

 

[antcollinet]

Thank you.

So if you don’t run it at full tilt, and your listening sessions are not so long, and less frequent, then your Class D (or Class Anything) amp is less prone to giving up the ghost, all else being equal?

yes - but not to the extent I'd limit my use of it.

 

[tmtomh]

For home use, I personally hate fans on hi-fi equipment, and IMHO there's no need for a fan unless you have to have an amp inside a closed/hidden cabinet with insufficient natural air movement and ventilation. Of course if you already have an amp that you like and it happens to run hot and you are concerned about that, then by all means stick a fan or cooler under (or on top of) it - I'm not trying to be judgy here.

Also, my understanding is that "crappy" caps are rated for "only" 85 degrees, and that's Celsius. So an amp running at 105F shouldn't be a problem, even if the internal temperature is bit higher than that in certain areas.

As for reliability, mass adoption of high-performing, high-power Class D amps for home use is relatively new, but so far the reliability record seems to look pretty good, with most failures seemingly being due to the use of off-brand caps and/or truly moronic thermal design (or lack of design) with casing, heatsinking, and so on.

My own personal sanity-check question for Class D would be, Is the topology used in respectable-brand professional amps? And the answer is Yes. That tells me pretty much all I need to know RE the reliability of the topology as a whole.

Finally, there are tons of great old Class AB amps out there that need their caps replaced.... after 30 or more years. And many of those caps are not super-premium and have sat inside hot-running cases on hot boards for all that time. That suggests to me that a well-designed, well-built amp of any topology should - and more often than not will - last for a good long time.

 

[Yorkshire Mouth]

I wonder how much of the criticisms are based on people using these camps in California.

I suspect the dangers aren’t so great in northern England.

 

[martin900]

First as I've said here many times - there is nothing about the topology of class D that makes them any more or less reliable than any other type of amp.

It one hundred percent comes down to the quality of the design - how close to (particularly thermal) limits components are run at - and that is the same whatever class of amp you are talking about.

Having said that, and to answer your question - heat is the enemy. Heat x time in particular. Running at high output levels generates more heat. But again - the same for any class of amp. It is quite possible for a well designed class D to have much better reliability than a poorly designed class A or class AB.

Oh - and in the "heat is the enemy" - class D putting out less additional heat at full tilt gives the designers the *possibility* of designing a more reliable amp. If they choose to take it.

High frequency in hundreds of KHz (SMPS and such) + ripple does wear electrolytics, esp cheaper ones. Anything in high freq will be the first to go bad.

 

[mhardy6647]

What would be likely to lengthen/shorten the life of a Class D amp?

Heat.
Input power anomalies (e.g., a power surge).

 

[antcollinet]

High frequency in hundreds of KHz (SMPS and such) + ripple does wear electrolytics, esp cheaper ones. Anything in high freq will be the first to go bad.

Losses in capacitors cause heat that causes wear. Losses are higher at high frequencies it is true. That just requires the designer to take that into account, and rate the capacitors appropriately.

Just as he must take into account the relatively high ripple current seen in bulk capacitance in the typical linear PSU.

Design choices cause reliability issues, not topology.

 

[Sokel]

I would be more afraid of the times than the class.

To explain myself,there were tremendous shortages in the supply chain,specially of quality components.
I myself waited patiently for a year and a half for some Caddock MPxxx 30w resistors (don't ask) as I only trust Mouser and the shorts for genuine stuff.

Manufacturers must have a hard time with this and we can see that as capacitors for example differ from batch to batch.
Some may have to do some compromises probably.
Signs of the times

 

[dlaloum]

I wonder how much of the criticisms are based on people using these camps in California.

I suspect the dangers aren’t so great in northern England.

Cultural aspects such as how warm people keep their homes indoors, and whether they have air conditioning would also impact things...

Places with lots of AV, often tend to keep their temperatures lower indoors... and Europeans tend to run their homes very warm in winter (from my experiences...)

So there is this counterintuitive aspect that gear in California might be kept cooler trough its life, than gear in Northern England

 

[dlaloum]

Heat.
Input power anomalies (e.g., a power surge).

Power up and Powering down, can also count as "power anomalies"!

 

[DMill]

I have a question for engineers out there. Would how hard you drive an amp also have an effect? Obviously routinely clipping the amp would but my question still stands.

 

[Blumlein 88]

Getting back to the original question, class D amps from good brands are reliable. I have one in daily use over 14 years old. Used for a decade to power some extremely difficult speaker loads. I know people with similar ages. I think the question has been answered within reason. Heat kills electronics. What else is there to know?

 

[dlaloum]

I have a question for engineers out there. Would how hard you drive an amp also have an effect? Obviously routinely clipping the amp would but my question still stands.

Yes, in the sense that being driven hard, will generate more heat, and heat is the electronics killer.

But, as has been pointed out over and over, most people, most setups, most of the time, only use single figure watts.... Hence most amps are never "driven hard" mostly power reserves are for the occasional peak, for dynamics, to ensure the amp can easily handle difficult loads...

So I have an amp that is rated at 440W/ch @ 8ohm (1200W/ch @ 2ohm) ... and my typical PEAK usage is under 4W... and I have never lit up the led which indicates 16W peak.

But I have speakers that go down to 1.6ohm - and this amp sounds good with them, where my AVR (rated at 100W/ch @8ohm) - sounds merely OK - messes up a bit with dialogue/midrange... and clearly doesn't cope well with those speakers (I have another pair of speakers nominally 8ohm, which the AVR handles just fine...!)

My class D 440W amps, idling along at under 4W do the job superbly, and stay very cool indeed, they barely get slightly warm.
They have been in my system for almost 15 years, and were purchased used.

There is something to be said for having enough power to meet the old Aston Martin & Bentley specifications from the UK (before they had to declare power by law) - the specifications sheet stated "more than sufficient power under any circumstance"...
It allows for the amps to idle along staying cool, and live a long trouble free life!

 

[antcollinet]

Obviously routinely clipping the amp would

Counter-intuitively clipping is not a problem for an amp. While the amp is in clipping, the output transistors are hard on, almost no voltage across them, so almost no power dissipated in the transistor. (Not operating in the linear region).

 

[dlaloum]

Counter-intuitively clipping is not a problem for an amp. While the amp is in clipping, the output transistors are hard on, almost no voltage across them, so almost no power dissipated in the transistor. (Not operating in the linear region).

For the speakers on the other hand...

 

[egellings]

Counter-intuitively clipping is not a problem for an amp. While the amp is in clipping, the output transistors are hard on, almost no voltage across them, so almost no power dissipated in the transistor. (Not operating in the linear region).

There is power dissipated in a saturated, all-the-way-on transistor, but it's tiny; buncha current times a few millivolts.

 

[antcollinet]

There is power dissipated in a saturated, all-the-way-on transistor, but it's tiny; buncha current times a few millivolts.

I think that is what I said