Sorry in advance if this opens a can of worms but a side topic in another thread peaked my curiosity on this question.
I read in multiple threads (as far back as the search function goes) that some people suggest using a regulated supply on the "gainclone" is just waisted effort and others both here and on other sites claiming good supply regulation can actually improve sound, even going so far as swearing by it sounding superior.
I'm curious of peoples opinion who have tried both and actually compared by ear both at low and high volume levels.
My guess is that far many more factors play into this than just regulated vs. unregulated voltage, but I am sure there must be some unbiased opinions here that have actually heard amplifiers employing both types that were based on the LM3886/3875/1875, etc..
Look forward to reading the replies 😎
I read in multiple threads (as far back as the search function goes) that some people suggest using a regulated supply on the "gainclone" is just waisted effort and others both here and on other sites claiming good supply regulation can actually improve sound, even going so far as swearing by it sounding superior.
I'm curious of peoples opinion who have tried both and actually compared by ear both at low and high volume levels.
My guess is that far many more factors play into this than just regulated vs. unregulated voltage, but I am sure there must be some unbiased opinions here that have actually heard amplifiers employing both types that were based on the LM3886/3875/1875, etc..
Look forward to reading the replies 😎
If one uses a split rail supply, there is no need for regulation unless running at absolute maximum power almost clipping. Single rail supplies should be smooth, clean, noise free and regulated.
Yes, that is a valid assumption.
But the point above makes a simplistic comparison difficult. Especially with so many people trying silly low capacitance supplies with gainclones.
A regulator adds almost as much heatsinking, size and cost as the amplifier itself. My stance is that these resources are better spent on building a more powerful discrete amp which will have better performance.
In my experience, insufficient PSRR is the biggest weakness of chipamps. Regulation helps get the impedance 'seen' by the chipamp down to lower levels but doesn't totally overcome the subjective weakness of having poor PSRR. I resort to other measures - like using an output transformer to present a lighter load to the chip - to get closer to the transparent sound I'm seeking.
I find that a "simple" linear PSU consisting of transformer + bridge rectifier + a pair of smoothing capacitors, give such good performance that there is absolutely no need to invent the need for complication to improve something that doesn't need improving.
Just build the chip amp and it's PSU properly.
Just build the chip amp and it's PSU properly.
I tend to agree with most of the posts above that appear "against" the idea on grounds of practicality and diminishing returns. Most chipamps set up properly have much more than adequate PSRR however giving it a regulated supply as well is not going to cause any harm...... With smaller gainclone chip such as LM1875 with its limited 5A output current, you could try it and see. Its something I am going to build when I get round to it "just for the hell of it" for my 1875 based Lab amplifier mono-blocks. Make sure there's plenty capacitance, after the regulator circuit to handle peak currents and that it can handle the maximum current output by the amplifier chip being used..... There is a slight advantage that you can run them close to maximum voltage for the chosen speaker load without having to take into account mains voltage variations. The major disadvantage is you now have another set of transistors/regulators to find a heatsink (and space) for....
What is it about split rail supplies that they don't benefit from regulation?
Hi,
No. If anything a single rail supply removes the PSSR issue from one rail,
usually the negative rail which has the worse PSSR of the two rails usually.
You could argue to equal the negative rail you need regulation, and
its half as complicated so more cost effective for single rail, fair point.
However a chip amp that fine with unregulated dual rail
will be just fine with single rail unregulated generally.
Note if an amplifier has poor PSSR in general, single rail regulated
may be be a better option than dual rail regulated. Typically not
a real issue for most chip amplifiers with good PSSR numbers.
rgds, sreten.
What about voltage protection for power spikes when running these amps near their limits on nasty AC mains?
I know it isn't unusual for some party or club situations where power can fluctuate greatly, or when storms move in and cause power line spikes or brownouts.
I have seen other high end amplifiers online using regulated supplies claiming to have tighter response, especially with bass frequencies. Guessing this is not much different than having a large reservoir of PSU capacitors for loud sound peaks.
If that is the case then wouldn't a simple but effective PSU regulator remove the need for a huge capacitance bank?
There must be some good reason some people are using regulation in their design from a professional standpoint besides the ones already mention by others here so far?
I know it isn't unusual for some party or club situations where power can fluctuate greatly, or when storms move in and cause power line spikes or brownouts.
I have seen other high end amplifiers online using regulated supplies claiming to have tighter response, especially with bass frequencies. Guessing this is not much different than having a large reservoir of PSU capacitors for loud sound peaks.
If that is the case then wouldn't a simple but effective PSU regulator remove the need for a huge capacitance bank?
There must be some good reason some people are using regulation in their design from a professional standpoint besides the ones already mention by others here so far?
Interesting document I found on this...
https://www.extron.com/download/files/whitepaper/powersupp_reg.pdf
"Power Supply Regulation in Audio Power Amplifiers"
Really good read so far. I am still skimming through right now.
Has some interesting points on the uses of regulated supplies and where they can benefit in some amplifier designs.
https://www.extron.com/download/files/whitepaper/powersupp_reg.pdf
"Power Supply Regulation in Audio Power Amplifiers"
Really good read so far. I am still skimming through right now.
Has some interesting points on the uses of regulated supplies and where they can benefit in some amplifier designs.
I'm not sure about the output transformer, but without a doubt an improvement in the PSRR of my 7297 was a big step up. It was more important that it had adequate power than the highest level of PSRR, however.
Other features have certainly aided in getting past the performance of some class A amps, but it wouldn't be there without the dual function of regulation and improved PSRR.
Perhaps not all chipamps are equal, either. However in general they seem incapable of performing at somewhat exquisite levels without essentially externally correcting what a big old class A like an F5 does inherently because of using discrete transistors. The benefit? Cheaper, easier to work with, way less power waste.
The biggest advantage of classA operation is that its possible to have constant power draw from the rails, leading to inconsequantial levels of noise on the supplies. Not all classA amps though aim to be power supply invariant - for example I see plenty of SE classA amps where the load is fed across the CCS. To me this looks to miss the point of classA - when the load's fed across the active element the PSU noise drops significantly.
I like regulated supplys,
i can get closer to the maximum supply voltage without havin to worry about the main's fluctuation. from a higher supply voltage small chips like tda1554q and alikes, and sta540 (my new favourite) preforms a lot better, the highs get "muddy" from unregulated powersupplys. the stuff i listen to do have some sudden highly powerdemanding notes, and there is a noticable difference.
it largely depends on how hard is the amp driven.
i can get closer to the maximum supply voltage without havin to worry about the main's fluctuation. from a higher supply voltage small chips like tda1554q and alikes, and sta540 (my new favourite) preforms a lot better, the highs get "muddy" from unregulated powersupplys. the stuff i listen to do have some sudden highly powerdemanding notes, and there is a noticable difference.
it largely depends on how hard is the amp driven.
If your highs get "muddy" doesn't that just mean you do not have enough storage capacitance or does that mean the transformer is undersized or have poor regulation?
my non regulated sta540 had 2x10,000 uF paralell, and a 5A continous rated xformer.
should had been enough, i guess.
should had been enough, i guess.
Buuuut.. when using a large overrated regulator (or regulator with large current boosting transistor) you will not dissipate that much heat.
Also this depends on the difference in voltage between the input and output. The less difference, less heat will be produced.
If designed properly I believe a regulator would not add much to the cost of a design and a much smaller heatsink could be used, or heck even the same heatsink that the amplifier chips are screwed to.
Anyways, my last chipamp I built using a TDA7384 that came out of a car stereo head unit was built around a regulated supply. Reason for this was because the chip could not be used above 14v and I just happened to have an old Radio Shack PSU that was beat up with a regulator in it that could provide the amperage the chip needed, so I ripped the guts out and used that.
I noticed when testing using that power supply compared to one I had threw together on the bench that was unregulated at the same voltage and amperage that some of these claims of better more defined bass and treble held true, at least in my case.
It very well could have been my ears that day, but this was before I even questioned if there was a difference between reg/unreg supplies.
a verry large powersupply running as an unregulated one, and a regulated but smaller powersupply will preform the same, actually it is sometimes cost effective to use a reglauted powersupply.
allso, for class D amplifiers i had not noticed any difference at all.
even running them from a pc smps made no difference.
for class A it allso makes no difference if its regulated or not.
Class AB it counts a lot.
i like regulated powersupplys, they protect the amplifier from main's peaks,
make sure i can go near the limits of the chip, that few volts excess leaves headroom for rail sagging so i don't have to purchase a large powersupply.
heat wise, yess it makes some heat, but regulators are cheap nowdays, and one has to be pretty unlucky if can't salvage an older cpu cooler. and that is by far enough for a chipamp regulator.
the additional costs are the 1 or 2 regulators needed, or if more current is desired then a pair of current boosting silicone. really not a big thing.
allso, for class D amplifiers i had not noticed any difference at all.
even running them from a pc smps made no difference.
for class A it allso makes no difference if its regulated or not.
Class AB it counts a lot.
i like regulated powersupplys, they protect the amplifier from main's peaks,
make sure i can go near the limits of the chip, that few volts excess leaves headroom for rail sagging so i don't have to purchase a large powersupply.
heat wise, yess it makes some heat, but regulators are cheap nowdays, and one has to be pretty unlucky if can't salvage an older cpu cooler. and that is by far enough for a chipamp regulator.
the additional costs are the 1 or 2 regulators needed, or if more current is desired then a pair of current boosting silicone. really not a big thing.