> should I just start experimenting with low level switching power supplies?
No.
If "Class A is like driving with the parking brake on", then switch-mode supplies are like doing watch-repair on a vibrating bed.
Switch-mode supplies are now common, almost mandatory, in VERY high power commercial sound. When you want a dozen 2 KW amps in a 48" road-rack, the weight savings pay for the much more complicated design. But gold-ear tests usually favor the meaty sound of conventional amps. Absent a compelling weight issue, and a clear path to a low-price low-ripple design, I would not go there.
Ordinary SMPS design won't work for Class-B audio. A common SMPS has a minimum output current: it can't drive its effective impedance high enough to control very small current. They can be pretty good: many PC power supplies can idle with just the load of their internal DC fan. But it is a design problem, and one not covered in cookbook design tools.
Watch surplus sales. Monster caps sometimes turn up at attractive prices.
If power noise-ripple is really a problem, go to a 2-stage filter. Once you get past the bare-minimum capacitance (say 1000uFd), even a very lame 2-stage gives better ripple rejection and much better rejection of ripple harmonics, with little effect on speech/music power.
Happy 2005!
No.
If "Class A is like driving with the parking brake on", then switch-mode supplies are like doing watch-repair on a vibrating bed.
Switch-mode supplies are now common, almost mandatory, in VERY high power commercial sound. When you want a dozen 2 KW amps in a 48" road-rack, the weight savings pay for the much more complicated design. But gold-ear tests usually favor the meaty sound of conventional amps. Absent a compelling weight issue, and a clear path to a low-price low-ripple design, I would not go there.
Ordinary SMPS design won't work for Class-B audio. A common SMPS has a minimum output current: it can't drive its effective impedance high enough to control very small current. They can be pretty good: many PC power supplies can idle with just the load of their internal DC fan. But it is a design problem, and one not covered in cookbook design tools.
Watch surplus sales. Monster caps sometimes turn up at attractive prices.
If power noise-ripple is really a problem, go to a 2-stage filter. Once you get past the bare-minimum capacitance (say 1000uFd), even a very lame 2-stage gives better ripple rejection and much better rejection of ripple harmonics, with little effect on speech/music power.
Happy 2005!
With all due respect... 
While the information that I've garnered about SMPS does indicate that it's not something to be undertaken lightly by the average DIYer, for high power applications, it seems to be the only way to fly.
Weight and size issues aside, the general efficiency of a SMPS is hard to beat. As to minimum current issues, with most of the amplifiers that I've seen there is signifigant bias current, which will definitely require enough current to keep a SMPS stable.
For the "golden eared" -- I don't believe everything that they say. As far as I'm concerned, the power supply shouldn't provide anything to the amplifier's sound. If it does, it's defective and needs to be improved, or replaced with something that doesn't.
While the information that I've garnered about SMPS does indicate that it's not something to be undertaken lightly by the average DIYer, for high power applications, it seems to be the only way to fly.
Weight and size issues aside, the general efficiency of a SMPS is hard to beat. As to minimum current issues, with most of the amplifiers that I've seen there is signifigant bias current, which will definitely require enough current to keep a SMPS stable.
For the "golden eared" -- I don't believe everything that they say. As far as I'm concerned, the power supply shouldn't provide anything to the amplifier's sound. If it does, it's defective and needs to be improved, or replaced with something that doesn't.
PRR said:
A switched mode PS, if the output inductor is large enough, can tolerate a fairly low load -- it's all a matter of working the equations and then dealing with the cost and size of the inductor. To this end the DIYr should design the supply so that it burns some watts constantly -- the QEX article on a modded ATX supply would run 100ma min, 20 amps max -- and the output inductors were designed for this.
The other big problem I see for DIY SMPS design are the EMI or RFI problems --
Again, I suggest that DIYrs search on the Nat Semi website and search under Sanjaya Maniktala's name for articles on optimizing the magnetics in SMPS.
An interesting series of posts and opinions.
The difference between a good power amp and a great power amp is in the details. Good power amps are availlable in the comercial market place for less money than the parts will cost you to build your own, so unless you just want experience or knowledge, why build just another good power amp. Just buy one.
Great power amps from commercial sources cost way more because they are built in low volumes and use more expensive parts (such as oversize filter capacitors even going so far as non-electrolytic capcitors for main storage) They may use a separate chassis for power they often use more regulation or choke input filters, they might be fully balanced end to end effectively doubling the number of parts, the list just goes on and on. They use a host of ideas to make little improvements that are all well down the diminishing returns curve and well up the increasing price and complexity curve.
Yes the basic schematic of an average power amp is not that big a deal but I believe you should be wary of people who tell you this is all simple and why bother with that complicated stuff. The complicated stuff is what makes the design special instead of ordinary. Like I said you can buy ordinary for not much money it will probably look pretty and be quite reliable and work fine. It just wont be a great power amp.
The difference between a good power amp and a great power amp is in the details. Good power amps are availlable in the comercial market place for less money than the parts will cost you to build your own, so unless you just want experience or knowledge, why build just another good power amp. Just buy one.
Great power amps from commercial sources cost way more because they are built in low volumes and use more expensive parts (such as oversize filter capacitors even going so far as non-electrolytic capcitors for main storage) They may use a separate chassis for power they often use more regulation or choke input filters, they might be fully balanced end to end effectively doubling the number of parts, the list just goes on and on. They use a host of ideas to make little improvements that are all well down the diminishing returns curve and well up the increasing price and complexity curve.
Yes the basic schematic of an average power amp is not that big a deal but I believe you should be wary of people who tell you this is all simple and why bother with that complicated stuff. The complicated stuff is what makes the design special instead of ordinary. Like I said you can buy ordinary for not much money it will probably look pretty and be quite reliable and work fine. It just wont be a great power amp.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.