In terms of durability a well designed class AB amplifier can go on for many decades, I will be surprised if class D amplifiers match their longevity.
Why wouldn't they?
Modules of that type tend to have a definite life expectancy, dependent on how well they're implemented. The key will be long term availability of replacement modules or equivalent drop in equivalent modules.
Why wouldn't they?
The "modules" used in typical implementations are plug and play and you are bound by the supplier maintaining stock of modules at whatever price they decide is reasonable (or unreasonable). It's all a recipe for mass landfill in 10-15 years.
repaired,The SMPS power supplies used frequently to power these modules are the weak point. Running capacitors at high frequencies and high currents produces heat, lots of it, and that heat is inside the cpacitors, in the electrolyte. They essentially cook themselves from the inside out.
The "modules" used in typical implementations are plug and play and you are bound by the supplier maintaining stock of modules at whatever price they decide is reasonable (or unreasonable). It's all a recipe for mass landfill in 10-15 years.
That's why SMPSs use capacitors specially designed for that purpose, with low losses at high frequencies.
Still not sure exactly what you mean by "modules" here. Yes, they are PCBs, just like in linear electronics. They can be repaired, just like linear electronics.
You'd better tell that to the many thousands of failed SMPS capacitors I've replaced over the years for premature failure from overheated electrolyte/venting.
But they are not repaired, that's the point. They are SMD, often with bespoke, unmarked components. Not only that, obtaining schematics for modules (n-core/ice etc) is challenging to say the least. Parts are not sold to do board level repair and you may only be able (for a short time) to obtain complete modules at exorbitant prices. Are you going to send an amplifier module across the world to be repaired or buy a new one? Nope, you'll shell out for a new board.
Sounds like a similar amount to the electrolytics I had to replace in linear/analog circuits.
The same goes for linear/analog electronics these days.
The point he is making is the linear power supplies don't fail as quickly.
Very soon.Ilkka, when you are going to publish more detailed user manuals with power response curve, group delay data etc.? (a la S360 manual)
I am not aware of mid range distortion issue on the 8351A model. The measured graph looks as it was designed. Having said that, the new 3rd generation coaxial driver in 8351B and 8361A is greatly improved in all performance aspects, including all sorts of nonlinearities.Have you managed to solve the mid range distortion issue (8351A), which was reported on several forums and measured by Knif Audio?
8351A Distortion & step response:
RRP of 8331A and 8341A will stay the same.Thx for the info @Ilkka Rissanen. Will the prices of 8341 and 8331 change? Will they also get this new coaxial driver?
What about power supplies for PCs? I haven't gone out of my way to do so, but I haven't read about any systemic issues about PC power supply failures for years now.The SMPS power supplies used frequently to power these modules are the weak point. Running capacitors at high frequencies and high currents produces heat, lots of it, and that heat is inside the cpacitors, in the electrolyte. They essentially cook themselves from the inside out.
Linear (transformer) supplies have massive amounts of copper and steel to dissipate what little heat they produce. The filter caps always run cool, as do the rectifiers. Their operational life is measured in multiple decades.
Failures in traditional linear supplies are very rare, easy to troubleshoot, and cost effective to repair. Failures in SMPS supplies are the opposite, expensive, require bespoke parts, often catastrophic in nature (multiple cascaded failure) and, due to unmarked or unidentified components, repairs are often simply not possible.
The "modules" used in typical implementations are plug and play and you are bound by the supplier maintaining stock of modules at whatever price they decide is reasonable (or unreasonable). It's all a recipe for mass landfill in 10-15 years.
I think it's OK if the gear isn't real expensive. Just treat it as disposable when it fails. I think one could go safely as high as $750 per powered monitor provided the item sold in very large quantities making parts availability more likely. Cheap stuff like JBL LSR 3's are no risk. Genelec's are pricey and that is a risk.
Perhaps that's why really expensive audiophile speakers will stay passive, and really expensive amplifiers will have transformers which weigh as much as concrete blocks.
@JJB70 Sorry about those headphones falling apart on you. Which Yamaha amps do you have in mind. The WXA-50 is clean, but only 55w @8 ohms. Links to measurements appreciated if you have them. I'm wary of the LS50 wireless and high end Genelec's due to this issue. I might take a chance on the Adam a7x, which is where I got the $750 from. In the US Guitar Center stocks them in over 100 locations which puts them on a par with LSR 3's, Yamaha and Rockit for availability. Mind you, I don't think Genelec's are junk. Also, nobody in Houston has them.