I stumbled on Roger Sanders' ESL site, and read something on power supplies that puzzled me. Roger says:
Sanders has solved these problems by developing a voltage regulator that is essentially 100% efficient. There is no heat dissipated by the regulator system. There is no high-power/high-voltage switching that causes heat generation or RF problems.
The regulator in the Magtech amplifier maintains a stable voltage regardless of load or reasonable changes in the line voltage feeding the amplifier. It runs stone cold, produces zero RF energy, and is simple and reliable.
Linear Regulators without dissipation? What about the Vout-Vin difference times the current draw? It is not a switching regulator either, apparently. Does anyone know more about this design? Or maybe it is simply a cap multiplier which of course isn't really a regulator.
Jan
PS 1st post on the new forum. I love it!
Sanders has solved these problems by developing a voltage regulator that is essentially 100% efficient. There is no heat dissipated by the regulator system. There is no high-power/high-voltage switching that causes heat generation or RF problems.
The regulator in the Magtech amplifier maintains a stable voltage regardless of load or reasonable changes in the line voltage feeding the amplifier. It runs stone cold, produces zero RF energy, and is simple and reliable.
Linear Regulators without dissipation? What about the Vout-Vin difference times the current draw? It is not a switching regulator either, apparently. Does anyone know more about this design? Or maybe it is simply a cap multiplier which of course isn't really a regulator.
Jan
PS 1st post on the new forum. I love it!
When it comes to power supplies there is no 100% efficiency!
Linear or switching, they all suffer some power loss because there is no transistor with zero voltage drop ( IE zero saturation voltage ).
Even a simple cap multiplier has some losses due to the same Vsat.
In fact cap multipliers are just voltage followers, making them even more inefficient.
Many things are said on the internet, that in it's self does not make them true.
Linear or switching, they all suffer some power loss because there is no transistor with zero voltage drop ( IE zero saturation voltage ).
Even a simple cap multiplier has some losses due to the same Vsat.
In fact cap multipliers are just voltage followers, making them even more inefficient.
Many things are said on the internet, that in it's self does not make them true.
That's also my take on it. But I know the guy who wrote this and he is a very technical person, I don't think he would write something just for advertisement.
Maybe I should ask him directly.
Jan
Maybe I should ask him directly.
Jan
I'm not clear what problem 100% efficiency solves that 99% efficiency fails to solve. Its just marketing blah - he also says:
Our regulator is unique in that it runs cold and wastes no heat. So we offer virtually the only regulated power supply amplifier on today's market.
The 'So' is a non-sequitur here and obviously there are plenty of other amps using regulated supplies. Naim for example uses linear regulators and Soulution uses off-the-shelf SMPSU bricks. Chord uses their own in-house designed switching supply in their power amps, as do Benchmark.
Our regulator is unique in that it runs cold and wastes no heat. So we offer virtually the only regulated power supply amplifier on today's market.
The 'So' is a non-sequitur here and obviously there are plenty of other amps using regulated supplies. Naim for example uses linear regulators and Soulution uses off-the-shelf SMPSU bricks. Chord uses their own in-house designed switching supply in their power amps, as do Benchmark.
There are some unusual regulator circuits and devices.
Some examples:
https://ustpower.com/avr-guide-magnetic-induction-voltage-regulator-operation/https://ustpower.com/comparing-auto...oresonant-transformer-type-voltage-regulator/
Usually such regulators have drawbacks that make them not the best choice for most applications.
Some examples:
https://ustpower.com/avr-guide-magnetic-induction-voltage-regulator-operation/https://ustpower.com/comparing-auto...oresonant-transformer-type-voltage-regulator/
Usually such regulators have drawbacks that make them not the best choice for most applications.
@abraxalito: I don't think there are many power amps with linear regulators for the output stage. Maybe a few, but most shy away from it.
And besides the dissipation issue, you can make the point that without a regulator your amp in general can output higher peak powers (but not continuous powers), which is all that matters in music.
I fired off an email, let's see what the response is, if any.
Jan
And besides the dissipation issue, you can make the point that without a regulator your amp in general can output higher peak powers (but not continuous powers), which is all that matters in music.
I fired off an email, let's see what the response is, if any.
Jan
You could use a switch running at twice the line frequency, so 100 Hz or 120 Hz, similar to a rising- or trailing-edge dimmer. Since it's called Magtech, it could also control AC current using a saturable inductor with a DC control winding that controls the amount of core saturation. Those things are usually called mag amps, magnetic amplifiers or saturable core reactors. In any case, the efficiency will never be 100 %.
Let's put it more bluntly then:
-A linear 100% efficiency regulator is a physical impossibility!
You must have some voltage drop on the regulator in order to get a regulated output, thus you must have some power loss in the regulator in order to get a regulated output from a linear one.
That's how it works, that's it's physics, no matter what he or anyone else says.
PS: On a more realistic note, that text sounds just like marketing crap, nothing more!
The arrangement appears to be a combination of two power supplies, with the second turning on and off according to power requirement (voltage droop). From the following passage from the lengthy webpage (here) which many would choose to skip, it doesn't appear to be a linear regulator.
"The key to efficient operation lies in the way that the coupling transistors are operated. First, they are either fully switched on or fully turned off. This means that they have either infinite resistance or essentially none. This prevents them from putting any resistance in the circuit that would cause them to dissipate heat.
.......
In some ways, the Magtech's power supply is like a switcher in that its transistors are either on or off. But there is no specific oscillation involved as in a switcher. Also, it is relatively simple and operates very little and at low frequencies, so its reliability is outstanding (no failures have ever occurred). And because it never switches under power, there is no noise or radio frequency problems with it. "
"The key to efficient operation lies in the way that the coupling transistors are operated. First, they are either fully switched on or fully turned off. This means that they have either infinite resistance or essentially none. This prevents them from putting any resistance in the circuit that would cause them to dissipate heat.
.......
In some ways, the Magtech's power supply is like a switcher in that its transistors are either on or off. But there is no specific oscillation involved as in a switcher. Also, it is relatively simple and operates very little and at low frequencies, so its reliability is outstanding (no failures have ever occurred). And because it never switches under power, there is no noise or radio frequency problems with it. "
Here is the Sanders white paper. Zero voltage switching, an old idea and certainly not 100% efficient.
https://www.ti.com/lit/ml/slup089/slup089.pdf
But he does use the word "virtually" so perhaps he has not actually measured the efficiency.
http://sanderssoundsystems.com/technical-white-papers/161-the-magtech-regulated-power-supply-wp
https://www.ti.com/lit/ml/slup089/slup089.pdf
But he does use the word "virtually" so perhaps he has not actually measured the efficiency.
http://sanderssoundsystems.com/technical-white-papers/161-the-magtech-regulated-power-supply-wp
Last edited:
So would it be that at some point, the 'switch' is closed and it supplies the amp & charges the buffer cap, then when the cap is full, it switches off and the cap supplies the amp?
Or it switches off if the input falls below the set value and the cap supplies the amp?
That would conceptually not different from a switcher, complete with switching noise (albeit at lower repetition rate).
Jan
As far as I know, the first solid-state amplifier from The Acoustical Manufacturing Co. (QUAD) got a regulated supply with current limiter after some complaints from the BBC about amplifier failures.
Was that the Quad 50E in 1966? That's the same year that Dynaco produced their Stereo 120,
which has a regulated power supply..
which has a regulated power supply..
Last edited:
OK, Sanders pointed me to that Whitepaper: http://sanderssoundsystems.com/technical-white-papers/161-the-magtech-regulated-power-supply-wp
I see how it works, neat! Although I will have to point out that it is a bit misleading to talk about shunt regs in power amps and be silent on series regulators. Nobody in their right tech mind would design a shunt reg. for power amps, to suggest that as an option is misleading.
The other things is that when you use a higher unregulated idle voltage and accept some sagging under high load, you get higher peak power and lower continuous power, but peak power is where audio lives.
But his concept certainly is interesting.
Jan
I see how it works, neat! Although I will have to point out that it is a bit misleading to talk about shunt regs in power amps and be silent on series regulators. Nobody in their right tech mind would design a shunt reg. for power amps, to suggest that as an option is misleading.
The other things is that when you use a higher unregulated idle voltage and accept some sagging under high load, you get higher peak power and lower continuous power, but peak power is where audio lives.
But his concept certainly is interesting.
Jan
According to the white paper, they use a zero crossing switch. So the second rectifier is connected to or disconnected from the buffer capacitors at zero crossings of the current, so switching losses and interference will be quite small.
The zero voltage switching losses are low by nature, and also because of the 120Hz maximum
switching frequency. The conduction losses are also low because of using a bipolar transistor,
which has a low saturation voltage.
switching frequency. The conduction losses are also low because of using a bipolar transistor,
which has a low saturation voltage.
Last edited:
I have read most of it and... well a veeeery long introduction with half truths and down right misinformation on how unregulated power supply is bad for a power amplifier. Any decent power amplifier will have a decent amount of supply ripple rejection, enough to make it irrelevant as long as there is sufficient filtering capacity.
On the other hand there is no linear voltage regulator to speak of, in no sense of the meaning, it's just a form of a switcher ( sort of like a PFC ).
But even so, zero voltage switching only removes switching losses, which at that low frequency are rather small anyway.
The main losses are in conduction and they are the same with or without ZVS, thus making the voltage regulator anything BUT a 100% efficient!
As i said, marketing crap...
As opposed to what? A mosfet?
Apparently they are worried about supply voltage sagging and its effect on the output power. The PSRR has nothing to do with that.
I don't remember the type number. It was before they learned how to make SOAR protection circuits.
That is only relevant if you run the amplifier close to the clipping point, meaning that it would clip sooner under a heavy load.
But you should never do that anyway!
Outside of that, the PSRR has everything to do with preventing unwanted modulation, because any type of signal modulating the power supply would just look like a voltage ripple, which is supposed to be taken care of by the PSRR.
But you should never do that anyway!
Outside of that, the PSRR has everything to do with preventing unwanted modulation, because any type of signal modulating the power supply would just look like a voltage ripple, which is supposed to be taken care of by the PSRR.