Different Ground Reference

[Olegkuskhov]

HI, im new in electronic world. I plan to make a Linear Power supply which the specification is 0-30 V  0-5 A  CV CC which i need for some circuit testing, because buying new one is rather expensive for me :'(. I have bunch of schematics lying arround and I decided to try this schematic from "The Post Apocalyptic Inventor". I attached the schematic below.

I have tons of question about this circuit, but not to be greedy, i will ask it one by one  .

I've marked the main ground which in the schematic is represent by 'Strip line'  and the op amp ground which written as V03 ( a bit strange when He use V03 +5V and V03). I've heard about ground loop and notorius things about it. But my question is why do we need to separate the ground reference?. If so, why the voltage adjustment part where there's some voltage divider on it use the main GND and V03 at the same time?.

Last, i'm sorry for my english because i'm not from a country that speak english. Any advice or correction for me are extremely welcomed.

[ArdWar]

Vo₃ is actually "ground" referenced to the positive regulator output voltage (let's call it V_out from now on). It's done that way probably so that the whole control part of the circuit not needing a high voltage supply. It only need V_out+5, V_out, V_out-5 supply. Far more reasonable than 43V, 0V, -5V if it's common ground referenced to the regulated part of the circuit

edit: Made a quick sketch of the circuit's general idea. Pin 3 of RV1 should be to a voltage slightly higher than Vout, set by the 1K pot in your schematic.

[Olegkuskhov]

Ahh i see what u mean there. But what happen when we mixed those ground reference. What i mean is, we still use the schematic as it is, but instead of separating the ground reference we mixed the Vo3 like u said and the main ground. Will the circuit still works?. If not, why is that happen?

[EPAIII]

If you look at the upper left corner you see that the line/mains transformer has two secondary windings. The second secondary winding is center tapped. That center tap is the origin of the To3 "ground".

To3+5 and To3-5 come from the two three pin regulators that are supplied by that winding. This is a way of using two rectifier diodes, one for the plus supply and one for the minus supply, instead of four diodes in a bridge. This is called half wave rectification because each diode only conducts for one half of the AC cycle. So larger filter capacitors are needed as opposed to using a full wave bridge.

The apparent purpose for this separate "ground" is to allow it to float, probably close to the output voltage of the main regulator. So the two op amps are being powered by 5 Volts above and below that output Voltage. This is why you can see the notation Vo5 at that main output on the right.

Frankly I don't like that circuit and I think you could find something a lot simpler for a first power supply.

[David Hess]

I've marked the main ground which in the schematic is represent by 'Strip line'  and the op amp ground which written as V03 ( a bit strange when He use V03 +5V and V03). I've heard about ground loop and notorius things about it. But my question is why do we need to separate the ground reference?. If so, why the voltage adjustment part where there's some voltage divider on it use the main GND and V03 at the same time?.

It really should be called common instead of ground.

There are two "commons", the output negative and the output positive.  All of the control circuitry is referenced to the output positive marked Vo3.  There is a floating +/-5 volt supply also referenced to Vo3.  The advantage of this configuration is that the control circuitry can control a higher voltage than its supply voltage.  The disadvantage is the requirement for a floating supply.  In this case it allows the 30 volt LM324 operational amplifier to control up to 30 volts, which would otherwise be difficult.  Alternative designs could have used a common 40 volt operational amplifier.

The voltage error amplifier is comparing Vo3 at its inverting input to a divided down output, plus an offset generated by the Vo3+5V voltage divider, which becomes the reference.

Hmm, that capacitor across the 100k potentiometer is a mistake because it would add phase lag but it shows no value so was probably left out.

HP/Harrison made a lot of power supplies with this bootstrapped configuration.

[Olegkuskhov]

Hi thank you for all the answer. It seems that using different ground or we can call it "common" instead is a trivial approach when designing analog circuit. I've made some electronics projects but based on MCU, now i want to get my hands wet in analog world. I appreciate all the answer. I know maybe it's a bit too much to make this PSU design as we need to know exactly how to use the op amp properly. I came from EE major so i have maybe i could say 25% knowladge of beginner EE . I understand almost the Schematic Part that i posted here except the op amp things.

I got this same hardness feeling when first tinkering electronics in digital world. I can't barely understand anything, even  i don't know how to calculate resistor value for any given LED specs. But after making some projects i guess i have the guts to try analog world and hoping for the same happy ending. Cheers every one, appreciate your time a lot for answering my questions .

[ArdWar]

Hmm, that capacitor across the 100k potentiometer is a mistake because it would add phase lag but it shows no value so was probably left out.

That 1nF across pot actually act as sort of feed forward

[David Hess]

Hmm, that capacitor across the 100k potentiometer is a mistake because it would add phase lag but it shows no value so was probably left out.

That 1nF across pot actually act as sort of feed forward

It goes to ground so adds up to 90 degrees of phase lag.  It is the opposite of feed forward or a lead network.

To add lead it would need to be across the 8.2k resistor.

[Dave]

It goes to ground so adds up to 90 degrees of phase lag.  It is the opposite of feed forward or a lead network.

To add lead it would need to be across the 8.2k resistor.

No, he's right. Vo3 is the fixed, common reference of this circuit, GND is the node whose potential actually "moves" with respect to other supply rails in this circuit.

Since this is a floating regulator circuit, it would make much more sense to select what is currently Vo3 to be the GND. Kind of like it is done in E3631A schematic. Doing that would avoid this sort of confusion.

[David Hess]

I can see that now.  I wonder why the lead capacitor does not show a value.

[Olegkuskhov]

Hi after sometimes not open this forum, i have more understanding about the circuit i ask before. I see that this is a common practice when designing this kind of circuit. But i have one more question to ask. Does the ground later in pcb designing being connected together?. If so, i'm glad that my knowladge is good enough hahaha. If not, what is the reason for that. I can say this because how in earth the op amp is able to control the majn circuit if it's ground is not connected together. Thanks in advance.

[Terry Bites]

 It is common practice to have separate ground routes, one  for small signal grounds and another for high current grounds.

You don't want currents from the high current ground flowing in the small signal control circuit grounds. ie The control loop and reference circuits.

If it does, that will lead to regulation errors, noise and instability. This situation is known as a "ground loop".

The two grounds, power ands small signal, are bonded together at a single point near the mains inlet.

This principle should be applied to all circuits where small signal circuits are adjacent to heavy current circuits. eg Audio power amplifiers, motor controlers etc.
You might look up "force and sense" as appled to the psu output teminals. This is an appliciation of the same principal.