• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

Component ageing, firmware updates, ... - Do initial review measurement results still match after 2, 5, ... years of audio equipment use?

pogo

Major Contributor
Joined
Sep 4, 2020
Messages
2,416
Likes
934
The industry is familiar with the fact that field returns are measured with a larger tolerance. The reason for this is the consideration of component ageing.
The question now is, are the initial review measurement results of audio equipment tested still valid after 2, 5, ... years?
For example, what would be the measurable difference after years with a Fosi V3 Mono, which can get very warm? Does a NAD M23 have advantages due to its better thermal management?
Perhaps some earlier UUTs could be re-measured years later to get a better picture?

Edit:
I have added 'firmware updates, ...' to the thread title.
 
Last edited:
Perhaps some earlier UUTs could be re-measured years later to get a better picture?

Do you use the product daily or test it, pack it back in its box and pull it out 30 years later and retest? Big difference.

I've pulled gear out of long term storage that is NOS. I bought, tested, repacked and never touched for 10, 20 and even over 30 years. Everything (apart from mechanical, cassette decks/ belts etc) has been perfect out of the box.

Problem is, 30 years ago, I couldn't test the limits of a CD player or amplifiers with extremely low THD, so I have no reference as to how they performed originally. That said, I've not come across one of my NOS players that needed anything other than a loading belt or a bit of lubrication/clean. Recently compared two identical Sony ES players- one I've used for 30 years, the other was a stored backup. Both were extremely close to one another, better than original specs and equal to technical reviews at the time. Two identical upper range Pioneer CD players with factory adjusted trims for MSBs on the ladder DACs had not drifted or gone out of spec in their boxes since purchased in 1990. So 34 years with no change.

I recently unpacked an Akai microprocessor controlled graphic equalizer and spectrum analyzer HiFi component I bought closeout special on 2nd April 1990 (receipt was in the box). It had never been used. It operates perfectly and is absolutely mint of course. Like a time capsule.

But none of those get hot, or at least not significantly hot.

With the Fosi V3 mono, it's only new in the scheme of things. The SMPS plugpack will deteriorate, as they all do. The primary filter and secondary filter caps will slow-cook in the sealed case, as will the primary side switcher IC. The environment inside a sealed plastic high power SMPS brick is truly a horrible place for electronic components. But failure can take a number of years. Depending on the change in characteristics, the measurable impact may be quite significant or minimal. The amplifier may well outlast the power supply, which should be a relatively cheap and easy fix.
 
Last edited:
The industry is familiar with the fact that field returns are measured with a larger tolerance. The reason for this is the consideration of component ageing.
The question now is, are the initial review measurement results of audio equipment tested still valid after 2, 5, ... years?
For example, what would be the measurable difference after years with a Fosi V3 Mono, which can get very warm? Does a NAD M23 have advantages due to its better thermal management?
Perhaps some earlier UUTs could be re-measured years later to get a better picture?
As @restorer-john stated lots of gear lasts a long time and stays in spec. As per gear that has a limited lifespan and poor spec retentivity I serviced 1980s Technics micro stereos. The ones where they are very small components that stack together and are very attractive looking. The audio amp section with integrated power supply was well known to require a near complete recapping as the high temps within with the bursting caps and hot semiconductor that was crammed in a small case with minimal air flow for cooling. The audio would cut in and out, the power supply cut in and out and all sorts of issues occurred. Did it deviate from spec within ~8-15 years or more ? Yes, it did to a great degree. After a recapping the amp with power supply would fire up and work well until it again needed a recapping. Was it good gear? Yes. Was it designed to last 30 years of daily use. No. So it's a matter of analyzing many factors in the design of the gear and from that one can maybe get a general idea of a estimated reliability and life expectancy range with a very general idea of if it will remain in spec. As @restorer-john enjoys and likes some 1970s, 1980s and some 1990s gear because the Japanese manufacturers where putting out some great gear with high quality components and it stayed within spec. So if you want great gear that stays in spec and is reliable you need to pick and choose wisely and take your time to ensure you make a good decision. Be prepared to spend a bit more for this reliability and long term spec retentivity.
 
The challenge today is to either purchase sub $500 gear and expect 5 years maximum. A throw away world just like mobile phones.

Or spend $3000 and get 10 years +. Yet become locked into state of the art at time of purchase.

The ROI is self evident I think, with the cheaper gears allowing regular upgrades to newer specs and feature sets. Shiny and bright.

Caveat emptor indeed.
 
I use the Purifi reference design amp routinely for calibration of my system or when I doubt everything is working in my measurement system. The results remain identical to when I first tested them which I think was 3 years ago. Don't have history longer than that.
 
I think technically equipment with a lot of feedback will retain specs until they does not anymore . My layman intuition tells me that it is the circuit design with the quite large amount of nfb that creates the spec , i dont know what component should deteriorate until it flips gets unstable or suddenly have more thd or noise ?
 
I think technically equipment with a lot of feedback will retain specs until they does not anymore . My layman intuition tells me that it is the circuit design with the quite large amount of nfb that creates the spec , i dont know what component should deteriorate until it flips gets unstable or suddenly have more thd or noise ?
In my repairing experience the voltage amp sections of audio amps can be too lightly built and then the transistors are overloaded, they commence making fuzz, crackle and other distortion sounds due to being too hot and nowhere for that heat to go. That causes overloading of the resister networks setting the operational parameters of those transistors and that increases the amount of distortion until failure occurs due to the resisters being fusible and or the transistor(s) going into thermal runaway.
 
I would also consider mechanical contacts such as relays, internal and external plug contacts, potentiometers, ... as a source of error.
 
In my repairing experience the voltage amp sections of audio amps can be too lightly built and then the transistors are overloaded, they commence making fuzz, crackle and other distortion sounds due to being too hot and nowhere for that heat to go. That causes overloading of the resister networks setting the operational parameters of those transistors and that increases the amount of distortion until failure occurs due to the resisters being fusible and or the transistor(s) going into thermal runaway.
I think we're seeing something similar here right now: Link
The failures seem to be piling up.
 
I think we're seeing something similar here right now: Link
The failures seem to be piling up.
I don't think there is a correlation because I was detailing a class A/AB amplifier voltage amp section and your example is a class D amplifier. But if a BJT transistor or FET is overheating distortion will rise substantially and failure may occur if not addressing the heat issue. It applies to class A/AB/D/C and H etc.
 
Probably quality of components,apart from everything else and up to a price.
Example,ancient 20 years old interface specs:

specs.PNG
(link)


..after 20 years of abuse,traveled round the world,etc. :


20 years.PNG

(don't be fooled about a low powered device,it runs hot on occasions,mic class A pres and the chip inside to their part to it)

((that's why I love AKM :) ))

Do I expect the same from today's gear at this price? (around 100 euro back then)
Nope,simply because no one would use components like this and charge them as much.See today's Topping interfaces who are comparable in price for example doing the same SINAD with this poor old thing.
 
The industry is familiar with the fact that field returns are measured with a larger tolerance. The reason for this is the consideration of component ageing.
The question now is, are the initial review measurement results of audio equipment tested still valid after 2, 5, ... years?
For example, what would be the measurable difference after years with a Fosi V3 Mono, which can get very warm? Does a NAD M23 have advantages due to its better thermal management?
Perhaps some earlier UUTs could be re-measured years later to get a better picture?
I believe there are too many variables in design and component selection to make any broad statements, aside from the fact that excessive heat should always be avoided.

There are vintage amplifiers that Amir has reviewed which performed well in measurements. Generally speaking, large old amplifiers with ample "empty" space inside can indicate their ability to withstand heat-induced failures. However, compact designs with hot enclosures do not necessarily mean that the unit will have a short lifespan, especially if the temperature issues have been addressed by using higher-rated components.
 
I would also consider mechanical contacts such as relays, internal and external plug contacts, potentiometers, ... as a source of error.
My experience as well. Add electrolytic capacitors, displays, ... and not at last some digital components. Slightly OT and without getting into the details but quite few synths from the 80s don't work anymore due to failures of some digital components (e.g. from 'bit rot') and lack of replacement parts.
OTOH my two sets of Shure E4 IEMs (single balanced armatures) from 2005/2006 sound just as good as when I bought them almost 20 years ago; I just had to replace the cables on both sets, the PVC didn't make it that far.
 
A good example is probably my Sony MDS-JA3ES. The laser unit destroyed itself over time. In the beginning, the error correction was able to intervene, but at some point the repetitions of the previously read out value as replacement values became too much and led to more distortions and even drop outs.
 
Last edited:
The SMPS plugpack will deteriorate, as they all do. The primary filter and secondary filter caps will slow-cook in the sealed case, as will the primary side switcher IC. The environment inside a sealed plastic high power SMPS brick is truly a horrible place for electronic components.

Amen to that. I've endured two "power brick" failures over the past couple of months. On the other hand, the suitably ventilated industrial SMPS that runs my two A07s has held up perfectly, even with its cooling fan disabled, for four years essentially 24/7/365.
 
if the temperature issues have been addressed by using higher-rated components.
... or division into several components, better thermal connection, better component separation...
To stay with the NAD M23 example, this is probably the reason why the PSU has such huge dimensions compared to other stereo 1ET400A implementations.
 
The results remain identical to when I first tested them which I think was 3 years ago.
...Actually, it's more than five years ago. ;)
 
The industry is familiar with the fact that field returns are measured with a larger tolerance. The reason for this is the consideration of component ageing.
The question now is, are the initial review measurement results of audio equipment tested still valid after 2, 5, ... years?
For example, what would be the measurable difference after years with a Fosi V3 Mono, which can get very warm? Does a NAD M23 have advantages due to its better thermal management?
Perhaps some earlier UUTs could be re-measured years later to get a better picture?
Working on studios meant I saw a wide range of gear, most of which was heavily used, often every day, all day. What this taught me about drifting specs and failures are:
1) Bathtub curve is correct. a) Some things fail on day 1 or 2, but an identical device from the same batch works for years. b) You can sometimes get a sudden batch all of which fail at roughly the same time after many years of good service.
2) Cooler running stuff tends to stay on spec for longer and often is more reliable. BUT, that's not a universal rule, assuming components are specified for high enough temperatures, hot, stable reliable gear is possible.
3) Belts, bearings, relays, switches and cogs etc. have limited life. Some devices such as tape machines and turntables can get out of spec as mechanical components age.
4) A brand tells you nothing about reliability or staying on spec. A vendor may have a great, reliable product. They bring out a version 2 and it's nothing but trouble. Then version 3 comes out and it's fine. Every designer has off-days.
 
4) A brand tells you nothing about reliability or staying on spec. A vendor may have a great, reliable product. They bring out a version 2 and it's nothing but trouble. Then version 3 comes out and it's fine. Every designer has off-days.

I feel like the more conservative companies like McIntosh or even Marantz spend a lot of time on validation and Q&A for their hifi line. There are definitely issues with their AV software, but I think “company policy” for things like Q&A, supplier validation, etc. help.

Yamaha is ISO 9001, which will add a layer of company policy which goes beyond company culture.

This is the pro audio division, not sure what consumer division follows.
 
There was a long period of time when many electrolytic capacitors were mis-manufactured with a bad electrolyte.
They would loose capacitance, which would impact both the DC supplies and low frequency audio output.
 
Back
Top Bottom