Super Regulator

Hi viltsone

What is the input voltage? If you have 10V at the input and 5V at the output with a current of 2A you have 10W of dissipation which requires a more serious heat sink .

For improvements you can; replace R4 with CSS, also R1, R3, Q1 and D1 with CCS. A better op amp is the ADA4897. For 1V ref you can use ADR512, LM4041, ZXRE4041, AD1580 and similar. If you are going to jfet op amp try ADA4625, it is the best choice.
 
For improvements you can; replace R4 with CSS, also R1, R3, Q1 and D1 with CCS. A better op amp is the ADA4897. For 1V ref you can use ADR512, LM4041, ZXRE4041, AD1580 and similar. If you are going to jfet op amp try ADA4625, it is the best choice.
I am quite newbie in this. Could you please tell me what parts and values to use? I think it would be easier to make new schematic for that. I was planning replace D4 with GLED431 2,5V and D2 zener with RED Led and to add 33pf cap between opamp output and + input but I am not sure if it works like this way? I got some hints from your posted schematic in this post.

@grunf Any news about Walt Jung new regulator that you where working together?
 
AX tech editor
Joined 2002
Paid Member
I modified this schemaic to 5V version. D2 I am using 2,4V zener 0,5W because from local shop they have only this close to 2,5V. For D4 I am using LM4040 2,5V. Q1 and Q2 re BD140 and 2SC5200 for Qizl. D1 is GREEN Led. Input 13 V ac 3 Amps. OPA1611. R6 value is parallel combination with resistor and trim pot.
Problem is that Qizl 2SC5200 gets very hot in 3 minutes so I can not hold more than 3 seconds my finger on the SK-129-50 heatsink. What kind of heatsink is good to use for 2 amps output and what changes to make in this schematic for better reliability and noise specs?
Any suggestions for 1 V reference to use with OPA1641? Using Leds?

View attachment 1212330
You don't really need a 4V drop out. Lower the input to 22V for 20V out and you halve the dissipation.
Or even lower, it's work fine with 1V headroom but then you should be sure that Vin won't drop under load.

Jan
 
  • Like
Reactions: 1 user
@jan.didden Thank you. I did not thought about that. I am using transformer secondary out 9,35 V ac. After active rectifier I will get about 13,25 V dc and it drops to 12 V after regulator turned in. I am using it with 5 V out. So it went quite hot and I added a bigger heatsink but I see that I need better appropriate transformer for this. I added 4 Ohm resistor in series before input voltage and now regulator is much much cooler but now resistors are hot. :D

What would be a good transformer output for this application using 5 V out? What would be a good minimum voltage headroom for this 5 V out regulator?

I am using this regulator with my Volt 476 soundcard and I must say that it is a hole new sound for me. Using original wall adapter with Volt 476 it started to irritate me and I could not listen music. I did not enjoyed it. I got stressed. Now with this set up I am starting to enjoy it. Vocals and sound has a hole new spectrum and I feel live vibrations with it. It is much brighter and I can hear and enjoy music instruments in a new level. At the beginning I had some strange feelings because it sounded different with this regulator. There more I use it the more I like it and I think this regulator is also still opening up for better performance.
 
Another newbie question about the boards in the diyaudiostore. In this thread I remember seeing a post mentioning that the negative output half of the board can be configured to have positive output. Can't find it anymore, but wonder if this is possible. If yes, is there a resource (maybe a post in this thread) that explains how to do this?
 
AX tech editor
Joined 2002
Paid Member
There's no change to the board. The idea is to consider the board as a battery.
Depending on which side you ground, you have a pos or a neg voltage with respect to ground.
The only restriction is that the input (secondary, rect, cap, etc) should be completely floating and not connected to ground or the input stuff of another board in the system.

Jan
 
Jan, could you please clarify what you wrote? Does it mean that the setup below will blow things up if the negative board is populated and wired as a positive version?

1694774364393.png


I happened to be reading about Mr. Jung's shunt regulator design when you posted your answer. The PDF on his site says something similar:

1694774453044.png


Unfortunately, "details [are] not described" there... Can you (or someone else, feel free ;)) please elaborate on how to wire a fully floating raw DC input? (I am aware of how much I do not know about electronics design and how easy it is to unwittingly build something that will self-destruct.)
 
I'm reading Mr. Jung's explanation and interpret it as: you need two of the first picture, or a transformer with dual secondary windings. A center-tap transformer won't work.
Thanks. But the first picture is by definition floating, isn't it? Also if you use a single secondary (no center tap).

That is what confused me in Jan's answer. In the full wave bridge rectifier it is normal to 'define' one of the outputs as 0V, which is then often called GND. But it's still floating if you don't connect it to the protective earth, right?
 
I ran into problem that LED is not turning on but it regulates the output voltage. The led turns sometimes on for a 1-2 seconds at the beginning and then it goes off. Could it be that I turned this regulator on without op amp OPA1641 and voltage reference? It was working at the beginning. I did some swaps to another board for different voltage reference. I replaced D2 zener with a new one. I got voltages on op amp pin 6 about 5.5 volts. Output voltage is 12 V with LM369 and 6,8V zener with a R6 about 685 ohm trim. Pin 2 and 3 have voltages about 6,85 volts. Could it be Q2 issue? Input voltage 13,25V. With a lower output voltage is the same that led is not turning on.

What else to check? Any help would be appreciated. :)
 
I increased the input voltage and now LED is on. :LOL:

What would be a really good choice of caps to use in this regulator in different positions? Right now I have Nichicon UKZ 100uF and C4 position is Panasonic 120uF FR series. Other regulator has Panasonic FR caps. Somehow I do not have a good feeling about Nichicon UKZ caps in this regulator. Any better suggestions? Someone mentioned Panasonic Poly cap in specific location in this regulator gave a good sound improvement but I cant remember what location and who said that.

Can I hear in listening music if this regulator is oscillating?
 
Without a scope, you will hardly see the oscillations.They cannot be heard because they are over 20 kHz if there are any.
As for the electrolyte, the best choice is UKZ and UFG, and only then Panasonic FC or Rubycon ZLJ. They must have as little leakage current as possible so that low ESR or even polymers (the worst choice) are out of the question.
 
  • Thank You
Reactions: 1 user
What about C3 or end cap cap capacity if I have modified this with grunf schematic and I want to use this for 1,5 amps than should I increase cap uF-s? For 2,5 amps out? Or is 100-120uF enough? How C3 capacity size or ripple current affects sound quality? Should I look also cap ripple current for this C3 position?
 
What about C3 or end cap cap capacity if I have modified this with grunf schematic and I want to use this for 1,5 amps than should I increase cap uF-s? For 2,5 amps out? Or is 100-120uF enough? How C3 capacity size or ripple current affects sound quality? Should I look also cap ripple current for this C3 position?
In Super regulator I mean C4 / C8 caps. In grunf schematic it is C3.
 
My own mostly SMT take on the super regulator, set up for ~18V 1A out. The idea of using an extra transistor to drive the base of the pass transistor was taken from grunf, as I really liked how that took the burden off of the opamp. In addition, I used a modified GLED431 and connected it to a CCS, with all LEDs changed to a pair of 4148s. However, the two SOD-123 diodes take up about the same amount of room as a 3mm LED so it's easy to modify the circuit to use those instead if I change my mind.
1695321648357.png

1695322202315.png

1695322290948.png
 
Last edited:
  • Like
Reactions: 1 users
Member
Joined 2011
Paid Member
Looks like your circuit requires the pass transistor (D44H11 / D45H11) to have a current gain of 153 when the load current is 1000 mA. Unfortunately the datasheet only guarantees a current gain of 60. So you may have to sort your transistors by measured current gain, and discard the ones with low beta.