Thanks everyone, very helpful! I now see the obvious that the output voltage equals the total power supply voltage for the opamp. So a 24V output is not a problem for example for AD825 specified up to +/-15V meaning 30V total. This however makes me confused on how AD825 works in my +5V regulator. This opamp is specified from +/-5V, which is the total of 10V as the minimum. How can it possibly work at a twise lower power supply voltage than its specified minimum? I would appreciate any insight. Thank you!
The sample of AD825 you happened to receive, is accidentally better than what the datasheet promises. Many other units of AD825 are "only" exactly as good as the datasheet promises and no better. Manufacturing variations, Gaussian distribution, Bell Curve, etc.
BTW if your AD825 suddenly quits being better, and suddenly works only when its total supply exceeds 9.99V, Analog Devices won't give you a refund. "But, But, But, it USED TO work when I was violating the datasheet requirements!" doesn't cut any ice with them.
BTW if your AD825 suddenly quits being better, and suddenly works only when its total supply exceeds 9.99V, Analog Devices won't give you a refund. "But, But, But, it USED TO work when I was violating the datasheet requirements!" doesn't cut any ice with them.
Thanks Mark, but my question is not about my case 
The Super Regulator is specified to work from 3.3V. This means AD825 still works 3 times below its minimum voltage. Is this ability unique or common among opamps? Also, some opamps have a twice lower minimum voltage. Would it be worth trying them instead or does AD825 do everything it needs to in this circuit despite the lower voltage?
The Super Regulator is specified to work from 3.3V. This means AD825 still works 3 times below its minimum voltage. Is this ability unique or common among opamps? Also, some opamps have a twice lower minimum voltage. Would it be worth trying them instead or does AD825 do everything it needs to in this circuit despite the lower voltage?
Last edited:
Thanks Jan! AD825 works like a charm in my +5V regulator, but I wonder if it can be improved. What do you think of the following idea?
Say, I build a +5V/-20V regulator with the negative output on the positive ground and the positive pin of the positive opamp on the negative ground? This way I get a +5V output while the only purpose of the negative board is to feed the positive opamp with +20V (or even more?) instead of its own +5V? The unit would be powered by batteries with proper connections.
In case this would cause the opamp to be unbalanced, then how about feeding it from the output of a separate +/-15V board and have all 3 grounds connected? Or something like that?
Please let me know if this is a silly idea or not. My DAC is hyper sensitive to the quality of the USB power, so any improvement is critical.
Alex
Last edited:
Jan, is there any disadvantage in feeding the opamp by a battery? It typically stays within 13 -13.5V. Should I shunt the battery by a 100u cap? Thanks again!
Because the clean output is only +5V while a battery is 13V (or even +/-13V with two). I see your point on low voltage opamps, but they don't seem to have the same specs. Much lower bandwidth, slew rate, etc. while the USB 2.0 frequency is 480MHz.
The main advantage of the regulator in my case is a super low impedance at high frequencies. The main disadvantage of a battery is a high impedance. For this reason a battery doesn't work as a power supply for USB. However, there is no tangible change in the current consumed by the opamp, is there? So the relatively high battery impedance may be moot there unlike with the digital USB signal that constantly jumps from zero to the maximum.
Sorry, I am not an expert in this by any stretch, is my intuition way off?
The main advantage of the regulator in my case is a super low impedance at high frequencies. The main disadvantage of a battery is a high impedance. For this reason a battery doesn't work as a power supply for USB. However, there is no tangible change in the current consumed by the opamp, is there? So the relatively high battery impedance may be moot there unlike with the digital USB signal that constantly jumps from zero to the maximum.
Sorry, I am not an expert in this by any stretch, is my intuition way off?
I think it is. The regulator output is much cleaner, lower noise and lower impedance than any battery can hope to be.
This gives the opamp the best possible supply.
There is the issue of opamp bandwidth, that is correct, but there are surely opamps with similar bandwidth and slewrate as the AD825 but working on 5V (or even 3.3V).
I've had good results with the AD8031.
I don't understand the reference to the USB.
Jan
This gives the opamp the best possible supply.
There is the issue of opamp bandwidth, that is correct, but there are surely opamps with similar bandwidth and slewrate as the AD825 but working on 5V (or even 3.3V).
I've had good results with the AD8031.
I don't understand the reference to the USB.
Jan
Attachments
Last edited:
Thanks Jan, I appreciate it! The reference to USB simply is that I use the regulator for the USB power supply of my DAC. The USB signal essentially Is a step wave between zero and the maximum. So the current consumption by the USB input circuit from the regulator also is a step wave at the USB frequency. And so the regulator should deal with a sharp voltage drop in hundreds of MHz. I don’t know how AD825 does it, but it sounds fantastic. A 5V battery sounds bad. Super capacitors sound better than a battery, biput nowhere near the Super Regulator.
I don't think there's a sharp voltage drop - there better should not be!
There is a fast changing current draw, but not the whole USB load is varying like that.
There is a steady DC current draw and then some varying one on top of that due to the data flow.
The regulator together with the decoupling caps is designed to supply that varying current without the voltage changing.
Jan
Exactly, I've seen this on the o-scope. There is a 250mA DC draw modulated on top by a square wave of the signal at the USB frequency. Thanks for your opinion on using a battery to feed the opamp. Your expert insight points me in the right direction.
Hi Jan,
I want to try the idea of feeding the opamp in the +5V regulator from the output of the +/-15V regulator - build a power supply for a power supply I am building the +/- regulator for my phono preamp anyway, so nothing to lose by trying. I think connecting both grounds should take care of most things, so I have only one question. The opamp output in the +5V regulator should be at around +2.5V and at the same time around the middle point between the power supply lines. So there are two ways of doing this:
1. Feed the opamp with +/-15V. The opamp output is at +2.5V, but not at the midpoint between the power supply lines, although not too far from it.
2. Feed the opamp with +16.25V / -13.75V. If my thinking is correct, the opamp output still would be at +2.5V and also at the middle point between the power supply lines.
If these options make sense, which one would you recommend from the standpoint of the best performance (disregarding any added complexity)?
Thanks so much,
Alex
I want to try the idea of feeding the opamp in the +5V regulator from the output of the +/-15V regulator - build a power supply for a power supply
I am building the +/- regulator for my phono preamp anyway, so nothing to lose by trying. I think connecting both grounds should take care of most things, so I have only one question. The opamp output in the +5V regulator should be at around +2.5V and at the same time around the middle point between the power supply lines. So there are two ways of doing this:1. Feed the opamp with +/-15V. The opamp output is at +2.5V, but not at the midpoint between the power supply lines, although not too far from it.
2. Feed the opamp with +16.25V / -13.75V. If my thinking is correct, the opamp output still would be at +2.5V and also at the middle point between the power supply lines.
If these options make sense, which one would you recommend from the standpoint of the best performance (disregarding any added complexity)?
Thanks so much,
Alex
The opamp needing to be at the mid point refers to the midpoint of the supply it regulates, in this case the +5V so its output preferably at 2.5V.
The reason for this is that it is in the middle of the regulation range for that supply, between 0V and 5V.
This is completely independent of it's own supply voltage. If you feed it from +15V or +/-15V, it's output should not change (if it did, the +5V would go haywire!).
I don't think feeding the opamp separately makes a difference for the +5V output, but it does give you more choice in opamps of course.
Jan
The reason for this is that it is in the middle of the regulation range for that supply, between 0V and 5V.
This is completely independent of it's own supply voltage. If you feed it from +15V or +/-15V, it's output should not change (if it did, the +5V would go haywire!).
I don't think feeding the opamp separately makes a difference for the +5V output, but it does give you more choice in opamps of course.
Jan
Jan, another quick question, if I may. I am looking for a voltage references as good as LM329, but for 2.5V and 10V for the +5V and +/-18V regulators. I see some 3-pin units designed to work without a resistor like LT1460. Is it OK to use such devices and remove R5 in the following schematics?
https://linearaudio.nl/sites/linearaudio.net/files/superreg article V2.3.pdf
https://linearaudio.nl/sites/linearaudio.net/files/superreg article V2.3.pdf