DIY Purifi Amp builds

[March Audio]

Some electronic components, capacitors in particular, have a finite life span, albeit a fairly long one. I prefer to switch my stuff off.

This. Just turn it off.

It would be interesting to test people who think they hear significant differences due to warm up under controlled conditions.

 

[JimM]

Warm up on class D does not seem to be as long as a class AB or Class A but I still notice a difference after about 15 to 30 minuets, but as I said in a previous post, I suspect it is the opamps or preamp rather than the class D amps. But on class AB or class A dependent on how the much bias is on the output transistors I have always been able to tell a difference in warm up time.
JimM

 

[barrows]

Some electronic components, capacitors in particular, have a finite life span, albeit a fairly long one. I prefer to switch my stuff off.

Generally, electrolytic capacitors get stressed more by inrush currents than by being on all the time. As long as heat levels in the chassis are not high, the electrolytic capacitors will last longer if they are left on, and not subject to many cycles of being turned on. I have had an NC-400 amp with two SMPS 600s on 24/7 for many years now and it is still going strong. In general, most solid state gear not subject to high heat levels will last longer when kept powered up.

 

[renaudrenaud]

Generally, electrolytic capacitors get stressed more by inrush currents than by being on all the time. As long as heat levels in the chassis are not high, the electrolytic capacitors will last longer if they are left on, and not subject to many cycles of being turned on. I have had an NC-400 amp with two SMPS 600s on 24/7 for many years now and it is still going strong. In general, most solid state gear not subject to high heat levels will last longer when kept powered up.

Ok but not very climate friendly.

 

[barrows]

Ok but not very climate friendly.

It is all relative I guess. Compared to my Pass Labs amplifier, which idled at well over 200 watts of power consumption, and basically also needed to be on all the time to sound its best, the few watts a Purif amp will consume is virtually nothing. With my audio system on 24/7, my monthly electric bill is about $45 for my entire home. I think I am doing OK in this regard-could always be better though! I know plenty of people who keep a few 100 watt lights on outside all night long for "security" purposes. There are a lot of ways to save some electricity.

SMPS 1200 idles at 9 W
a single Purifi module at 1.7 W
lets account for bit another 4 W for the eval module (a guess)

So: 9+1.7+1.7+4=12.4 watts total at idle. I think most of use could easily save 10x this amount of electricity daily by remembering to turn off the lights when leaving a room.

 

[JimB]

Generally, electrolytic capacitors get stressed more by inrush currents than by being on all the time. ...

Please give us some reference to this for further understanding. I understand high, continuous currents, but not "one-time" initial turn-on events. I might have some questions after I get to learn more. (And, it might belong in a different topic.)

 

[barrows]

Jim, I do not have any additional information to cite. This was explained to me by audio engineers at a company where I previously worked. The context was that I was the Customer Service Manager, and technical support person, responsible for advising dealers and customers on how to best use the products (audio components). In my capacity at this position I also handled the administration of nearly all repairs, over the course of my years there I learned quite a bit about audio component failure modes as well.
IME, electrolytic capacitors, of good quality, when sensibly implemented (\within reasonable voltage tolerances, and not subject to super high temperatures) will easily last over ten years when constantly powered, without any degradation in performance. And the engineers told me that power cycling is the most stressful thing for those parts, as the inrush currents can be very high (talking power supply caps here).

I saw plenty of my share of failed electrolytic as well-the cause was almost always way too high heat levels, and/or capacitors which were running way too close to their rated voltages (sometimes even above). Well engineered products will have a fair amount of voltage margin in the specification of electrolytic capacitors, and this really can help with lifespan. Using 105 C. rated caps is a good idea as well, if there is any heat inside the chassis.

 

[JimM]

Yes inrush current can be harmful to a lot of the components, but the use of thermistors as fairly cheap solution can mitigate most of those problems, I haven't noticed the Hypex SMPS power supply exhibiting a turn on problem, like some of my class AB amps even with a properly designed power supply. So I assume they have power ramp up by some means to help this problem or it's just the nature of switch mode power supplies. I cringe when turning on a AMP and the lights dim momentarily. If I wasn't so old and lazy I would learn more about SMPS.
JimM

 

[barrows]

Yeah, lights dimming, haha. Been there done that! Of course, lights dimming means a momentary voltage drop, so the fact that the lights dim when powering up big amps is probably a good thing, as the lower voltage during in rush probably spares the parts of some stress.

 

[ChrisPa]

Ok but not very climate friendly.

Depends

In the winter it's removing (using Barrows calculations as a finger in the air) a 12W requirement from the heating system. (My radiators each have independent smart thermostats, so they will compensate for whatever heat the amps are providing)

In the summer (and in the North of England summer is a concept rather than a thermal occasion) I don't fly anywhere on holiday, so the amps are still being climate friendly

 

[Count Arthur]

It's not only capacitors: http://www.csl.mete.metu.edu/Electromigration/emig.htm

 

[March Audio]

Generally, electrolytic capacitors get stressed more by inrush currents than by being on all the time. As long as heat levels in the chassis are not high, the electrolytic capacitors will last longer if they are left on, and not subject to many cycles of being turned on. I have had an NC-400 amp with two SMPS 600s on 24/7 for many years now and it is still going strong. In general, most solid state gear not subject to high heat levels will last longer when kept powered up.

Do you have some supporting evidence for this? At best it is a generalisation and in my experience / opinion incorrect.

I can certainly tell you that I have had to replace many caps in products that have been left on all the time but never in products that are switched off.

The point you are missing here is that whilst your nc400 and smps may still be functioning, the caps performance and value will have been continuously degrading. This will ultimately start to affect the amps performance. It doesnt mean the caps have exploded, leaked or stopped the amp from functioning completely.

It's just a waste of electricity and on going cost to keep it switched on permanently for no benefit.

 

[March Audio]

Yes inrush current can be harmful to a lot of the components, but the use of thermistors as fairly cheap solution can mitigate most of those problems, I haven't noticed the Hypex SMPS power supply exhibiting a turn on problem, like some of my class AB amps even with a properly designed power supply. So I assume they have power ramp up by some means to help this problem or it's just the nature of switch mode power supplies. I cringe when turning on a AMP and the lights dim momentarily. If I wasn't so old and lazy I would learn more about SMPS.
JimM

The Hypex supplies have a 10 ohm NTC resistor to reduce inrush current.

 

[March Audio]

Just some general info on cap lifing. Note it doesn't have to be "hot" in the sense of high temperature to degrade life.

https://www.illinoiscapacitor.com › ...PDF
Life expectancy of Aluminum electrolytic capacitors - Illinois Capacitor

The rate of increase in operating life is for the life to double for every 10°C decrease in temperature (Arrhenius’s law).

So if your amp gets to a modest 35 deg internally, you are halfing the cap life over being switched off at 25 deg. That's ignoring the effect of number of charge discharge cycles, ripple current etc.

So when electrolytic cap life typically starts at just 1000 hours (at rated temps and voltage) , you probably want to do whatever you can to lengthen this life. Being switched on only when in use is obviously the most significant thing you can do to help and keep the amp performing at spec as long as possible.

 

[March Audio]

Jim, I do not have any additional information to cite. This was explained to me by audio engineers at a company where I previously worked. The context was that I was the Customer Service Manager, and technical support person, responsible for advising dealers and customers on how to best use the products (audio components). In my capacity at this position I also handled the administration of nearly all repairs, over the course of my years there I learned quite a bit about audio component failure modes as well.
IME, electrolytic capacitors, of good quality, when sensibly implemented (\within reasonable voltage tolerances, and not subject to super high temperatures) will easily last over ten years when constantly powered, without any degradation in performance. And the engineers told me that power cycling is the most stressful thing for those parts, as the inrush currents can be very high (talking power supply caps here).

I saw plenty of my share of failed electrolytic as well-the cause was almost always way too high heat levels, and/or capacitors which were running way too close to their rated voltages (sometimes even above). Well engineered products will have a fair amount of voltage margin in the specification of electrolytic capacitors, and this really can help with lifespan. Using 105 C. rated caps is a good idea as well, if there is any heat inside the chassis.

There are too many generalisations here.

"well engineered" is an interesting comment. Well engineered also means not pointlessly over engineering. Putting excessively over rated components increases cost. I would suggest most domestic hifi products these days are not designed in terms of life span for 24/7 operation.

 

[March Audio]

It is all relative I guess. Compared to my Pass Labs amplifier, which idled at well over 200 watts of power consumption, and basically also needed to be on all the time to sound its best, the few watts a Purif amp will consume is virtually nothing. With my audio system on 24/7, my monthly electric bill is about $45 for my entire home. I think I am doing OK in this regard-could always be better though! I know plenty of people who keep a few 100 watt lights on outside all night long for "security" purposes. There are a lot of ways to save some electricity.

SMPS 1200 idles at 9 W
a single Purifi module at 1.7 W
lets account for bit another 4 W for the eval module (a guess)

So: 9+1.7+1.7+4=12.4 watts total at idle. I think most of use could easily save 10x this amount of electricity daily by remembering to turn off the lights when leaving a room.

Well we have moved on to led light bulbs.....

Over 10 years a Purifi amp will use about 1350kWh. Here in Oz I think the price is about $0.25 /kWh, so about $330. So maybe not a high cost, but I think that misses the point somewhat. Its just wasteful and I would rather spend the extra maybe $250 to $300 it costs to keep it switched permanently on something else.

 

[barrows]

@March Audio , different strokes I guess. I work from home and use my system pretty much all the time. I find class D amps need to be warmed up for at least a few hours to sound their best, for me, that means leaving them on all the time, except when I leave town, when I unplug everything.

As to the caps condition in my NC-400 amp I will find out soon, as I plan to re-cap the NC-400 amp and the two SMPS 600s once my Purifi build is up and running, so I'll be able to check out the old caps. I figure the NC-400 has been up and running 24/7 for about 8 years now, so not a bad time to refresh it.

My electric is built is low, and turning off one 60 watt light bulb saves more power than an idling class D amplifier. If we are going to be thta concerned about power consumption, might as well give up on listening to music at home entirely.

 

[March Audio]

@March Audio , different strokes I guess. I work from home and use my system pretty much all the time. I find class D amps need to be warmed up for at least a few hours to sound their best, for me, that means leaving them on all the time, except when I leave town, when I unplug everything.

As to the caps condition in my NC-400 amp I will find out soon, as I plan to re-cap the NC-400 amp and the two SMPS 600s once my Purifi build is up and running, so I'll be able to check out the old caps. I figure the NC-400 has been up and running 24/7 for about 8 years now, so not a bad time to refresh it.

My electric is built is low, and turning off one 60 watt light bulb saves more power than an idling class D amplifier. If we are going to be thta concerned about power consumption, might as well give up on listening to music at home entirely.

Your system simply wont be in use "all the time". Highly surprised if it is used more than 12 hours a day.

Now try that "warm up test" blind and under controlled conditions. Im quite confident you wouldnt be able to tell the difference. Also the interesting thing is that the Hypex amps measured noise and distortion performance tends to go down slightly when warm.

Your analogy to a 60 watt bulb is a total red herring. Firstly the world is moving on to be more energy efficient. LED bulbs will be more like 5 watts. Saying "oh its only a light bulbs worth" doesent make leaving any unused bit of kit on any more justifiable. More to the point, its not just your specific energy consumption thats the issue. Its when a "just a few watts" waste is multiplied by millions of households it becomes an issue.

Well if you had turned it off you wouldnt need to go to the bother and expense of recapping your NC400. It would carry on for another 8 years quite happily.

 

[Count Arthur]

Correct me if I'm wrong, but wasn't the whole "warm up" thing more of an issue with linear power supplies that used large transformers, where inrush current could be very high when the power was initially turned on.

Really large transformers need a "soft start", or the inrush current can trip the RCD/circuit breakers on the mains. I've also seen it suggested that a soft start reduces the "warm up" time and allows amps to "settle down" quicker after swicth on: https://www.lcaudio.com/index.php?page=315

I imagine that this inrush current surge is less of an issue with a well designed SMPS that doesn't rely on a large bank of capacitors for current reserve in the power supply.

My Hypex UCD based amps have a linear PSU, with large troidal transformer with a bank of caps and soft start modules. I can't say I've noticed the sound improving after being on for a sginificant amount of time.

 

[Dimitri]

waste is multiplied by millions of households it becomes an issue.

Some people call it an On-OFF switch. I call it an ON switch and use it once