logo elektroda
logo elektroda
X
logo elektroda
Advanced Search
logo elektroda

12V DC to 0-30V DC 5A Schematic for Buck-Boost Supply

Ask a question on the Forum

User question

12v dc input, 0-30v dc output 5A schematic

Artificial Intelligence Response

Direct answer to the question

A 12 V DC to 0–30 V DC, 5 A supply is not a simple linear-regulator schematic.
The correct solution is a high-power switching converter, preferably:

  • Best single-stage topology: 4-switch non-inverting buck-boost
  • Easiest to debug: boost to ~34–36 V, then buck down to 0–30 V

Key points:

  • 30 V × 5 A = 150 W output
  • With realistic efficiency of 88–92%, your 12 V input source must supply about 14–16 A continuous
  • If your 12 V source is only 5 A, then 30 V at 5 A is impossible
  • A simple LM723 / 2N3055 style schematic will not work from a 12 V input for this requirement
  • A boost-only converter also does not fully solve it, because you want 0–30 V, which includes voltages below 12 V

A practical engineering answer is therefore:

  1. Use a buck-boost controller
  2. Use synchronous MOSFETs
  3. Design for 20 A-class input path
  4. Add CC/CV control
  5. If you require a true 0.00 V output, add a small negative auxiliary rail for the control amplifier, or use an output disconnect arrangement

Recommended practical schematic

Below is a practical text schematic for a single-stage non-inverting synchronous buck-boost supply.

1) Power stage schematic

 +12V INPUT
|
F1 20A
|
Reverse polarity
protection MOSFET
|
TVS diode
|
+-------+-----------------------------+
| |
CIN1..CINn |
(low-ESR electrolytic + ceramic) |
| |
GND |
|
Q1 Q3
+---MOSFET-----MOSFET---+
| | | |
| SW1 SW2 |
| +----L1----+ |
| |
| |
+---MOSFET-----MOSFET-+
Q2 Q4
| |
GND GND
U1 = 4-switch buck-boost controller
drives Q1..Q4
|
+---------------------> +VOUT
|
COUT1..COUTn
(50V low-ESR caps)
|
GND

2) Output sensing and control

+VOUT ---- R1 ----+---- R2 ---- GND
|
FB ---> U2A error amplifier ---> U1 feedback pin
Output current shunt:
LOAD- ---- RSHUNT ---- GND
Voltage across RSHUNT ---> U2B current limit amplifier ---> U1 current/COMP clamp
Voltage set potentiometer ---> U2A reference
Current set potentiometer ---> U2B reference

3) True 0 V support

+5V auxiliary ---- charge pump inverter ---- -2V rail
-2V rail powers U2A/U2B so the control loop can regulate all the way to 0V.

Without that small negative rail, many controllers can regulate down only to about 0.5–1.0 V, not true zero.

Example component class selection

These are engineering-grade example values, not final production values.

Block Suggested value / class
Input fuse 20 A
Input capacitors 2,000–3,000 µF total low-ESR + 10 µF + 100 nF ceramics
MOSFETs Q1–Q4 80 V N-channel, low \(R_{DS(on)}\), low gate charge
Inductor L1 10–15 µH, saturation current > 25 A, low DCR
Output capacitors 1,000–1,500 µF total, 50 V low-ESR + ceramic bypass
Output shunt 10 mΩ, Kelvin-connected, 3–5 W
Switching frequency 150–300 kHz typical
Auxiliary rail +5 V and about -2 V for control op-amp
Controller family LM5175 / LM5176 / LTC3780 / LT8390 class

Detailed problem analysis

This requirement has three separate engineering difficulties.

1) Voltage conversion difficulty

You want:

  • Input: 12 V
  • Output: 0–30 V

That means the converter must do both:

  • Step-down when output is below 12 V
  • Step-up when output is above 12 V

So:

  • Buck converter only: not enough
  • Boost converter only: not enough
  • Buck-boost: correct answer

This is why the first offline answer, which proposed a simple boost stage, is incomplete for a true 0–30 V output.

2) Power difficulty

Maximum output power is:

\[ P{out} = V{out} \times I_{out} = 30 \times 5 = 150\text{ W} \]

If efficiency is 90%:

\[ P_{in} = \frac{150}{0.9} \approx 167\text{ W} \]

At 12 V input:

\[ I_{in} = \frac{167}{12} \approx 13.9\text{ A} \]

In practice, design for:

  • 15 A minimum continuous
  • 18–20 A preferred margin

If the source is a battery and input falls to 11 V:

\[ I_{in} = \frac{167}{11} \approx 15.2\text{ A} \]

So your:

  • connector
  • fuse
  • wiring
  • PCB copper
  • inductor
  • MOSFETs
  • current sensing

must all be sized for this current.

This is the most important correction to the simplistic “12 V in, 30 V 5 A out” idea:
it is feasible only if the 12 V source can provide very high current.

3) True 0 V output difficulty

Many DC-DC controllers regulate to an internal reference such as:

  • 0.8 V
  • 1.0 V
  • 1.2 V

That means the lowest direct output is usually not 0 V.

If you need a real laboratory-style 0.00 V to 30.00 V, you typically need one of these methods:

Method A: Small negative rail for control loop

Use a charge pump or auxiliary winding to generate about -1.5 V to -3 V for the error amplifier.

Advantages:

  • True 0 V regulation possible
  • Best for real bench-supply behavior

Disadvantages:

  • More complexity

Method B: Output disconnect MOSFET

The converter regulates from about 0.8 V upward, and a MOSFET disconnects output for “0 V”.

Advantages:

  • Simpler

Disadvantages:

  • Not true analog 0 V regulation under load

Method C: Linear post-regulator

Switching pre-regulator plus linear pass stage.

Advantages:

  • Very clean output

Disadvantages:

  • Very high heat at low output voltage and high current
  • Not attractive at 150 W class

For your requirement, Method A is the best engineering solution.

Why the common LM723 / 2N3055 schematic is not appropriate

A lot of “0–30 V 5 A” schematics use:

  • transformer
  • rectifier
  • large filter capacitor
  • linear regulator
  • pass transistor(s)

That works only when the raw DC bus is already higher than 30 V, usually derived from mains AC.

It does not solve:

  • starting from 12 V DC
  • stepping above 12 V
  • doing so efficiently at 150 W

So if you start from 12 V DC, you need switch-mode conversion.

Recommended topology comparison

Topology Can do 0–30 V from 12 V? Complexity Efficiency Recommendation
Linear regulator only No Low Very poor Not suitable
Boost only No, not below 12 V Medium Good Incomplete
Buck only No, not above 12 V Medium Good Incomplete
Boost + buck Yes High Good Very practical
4-switch buck-boost Yes High Very good Best single-stage

Engineering recommendation

If you are designing from scratch:

  • For easiest implementation: use 2-stage

    • Stage 1: 12 V to 34–36 V boost
    • Stage 2: 36 V to 0–30 V buck with CC/CV

  • For best final design: use single-stage 4-switch synchronous buck-boost

Current information and trends

In current power electronics practice, the preferred approach for this power range is:

  • Synchronous conversion
  • Current-mode control
  • Kelvin-sensed shunts
  • Low-DCR shielded inductors
  • 4-switch buck-boost controllers
  • Digital or analog CC/CV loops

Current design trends emphasize:

  • higher switching frequency for smaller magnetics
  • lower \(R_{DS(on)}\) MOSFETs
  • better EMI control through layout rather than only filtering
  • accurate current limiting and soft-start for battery-powered systems

For a 150 W design, discrete legacy controllers such as TL494/SG3525 can still be used, but they are usually more difficult to stabilize and optimize than a modern dedicated buck-boost controller.

Supporting explanations and details

How the 4-switch buck-boost works

It uses two half-bridges and one inductor:

  • Left half-bridge connected to input
  • Right half-bridge connected to output
  • Inductor between the two switching nodes

Operating modes:

Buck mode

When \(V{out} < V{in}\)

  • input side switches actively
  • output side behaves as synchronous rectifier

Boost mode

When \(V{out} > V{in}\)

  • output side switches actively
  • input side handles energy pumping

Buck-boost transition mode

When \(V{out} \approx V{in}\)

  • both bridges participate
  • controller blends operation smoothly

This gives continuous regulation across the whole range.

Why MOSFET voltage rating matters

Do not use marginal MOSFETs such as 55 V parts in a 30 V output boost/buck-boost design unless you have carefully verified transients.
A safer engineering choice is typically:

  • 80 V MOSFETs

because switching spikes and ringing can easily exceed the nominal bus voltage.

Inductor sizing reality

At 150 W output, the inductor does not just carry 5 A.

In boost or buck-boost operation, inductor average current is much closer to the input-side power current, so it may be:

  • 14–16 A average
  • 18–22 A peak, depending on ripple and transient load

Therefore choose:

  • 10–15 µH
  • Isat > 25 A
  • low DCR
  • good thermal performance

Output current sensing

For a 5 A output limit, a practical low-side shunt is:

\[ R_{shunt} = 0.01\ \Omega \]

At 5 A:

\[ V_{shunt} = 5 \times 0.01 = 50\text{ mV} \]

Power in shunt:

\[ P = I^2R = 25 \times 0.01 = 0.25\text{ W} \]

Still choose 3–5 W physical rating for robustness, lower temperature rise, and pulse tolerance.

Use Kelvin connections to the shunt.
Do not sense current through heavy copper traces.

Ethical and legal aspects

For electronics engineering, the main issues here are safety, compliance, and reliability rather than ethics in the abstract.

Safety aspects

  • A 12 V source may appear “safe,” but a battery or high-current supply can deliver very dangerous fault current
  • Add:
    • fuse
    • reverse-polarity protection
    • thermal shutdown
    • over-current protection
    • output short-circuit protection

EMC / regulatory aspects

If this will be sold or used in commercial equipment, consider:

  • conducted emissions
  • radiated emissions
  • immunity
  • proper enclosure grounding
  • cable filtering

Fire risk

At 150 W, poor layout or undersized parts can overheat:

  • MOSFETs
  • inductor
  • connectors
  • shunt
  • PCB vias

Practical guidelines

Best implementation method

If you want the highest chance of success:

Option 1: Two-stage design

12V IN --> synchronous boost to 36V --> synchronous buck CC/CV to 0-30V / 5A

Why this is easier:

  • easier loop compensation
  • easier testing
  • each stage is familiar
  • easier fault isolation

Option 2: Single-stage 4-switch buck-boost

Use if:

  • you are comfortable with SMPS compensation
  • you can do proper PCB layout
  • you want better integrated performance

PCB layout rules

These are critical.

Keep high-current loops short

Especially:

  • input cap → MOSFETs → shunt → ground
  • MOSFET → inductor → MOSFET loop
  • output cap return loop

Use wide copper

For 15–20 A class current:

  • heavy copper planes
  • multiple vias between layers
  • no thin trace bottlenecks

Separate grounds

Use:

  • power ground
  • analog ground

Join them at a single controlled point near the controller sense reference.

Kelvin sense

For:

  • shunt resistor
  • output voltage sense
  • compensation network reference

Snubbers

Provide RC or RCD snubbers if ringing appears on switching nodes.

Thermal guidelines

  • Heatsink MOSFETs if needed
  • Expect forced airflow at full power
  • Place inductor away from sensitive feedback circuitry
  • Verify temperatures at:
    • full load
    • low output voltage
    • maximum output voltage
    • short-circuit / current-limit mode

Test plan

Bring it up in this order:

  1. Test controller auxiliary rails only
  2. Test gate drive with no power stage load
  3. Power from current-limited bench supply
  4. Verify switching nodes on oscilloscope
  5. Test with dummy load at low power
  6. Increase to 25%, 50%, 75%, 100% load
  7. Check:
    • efficiency
    • ripple
    • thermal rise
    • transient response
    • current limit behavior

Possible disclaimers or additional notes

  • If your source is 12 V / 5 A, your requested output is impossible.
  • If your source is 12 V battery or 12 V / 20 A SMPS, the design is feasible.
  • A full reliable 150 W SMPS is not a beginner project.
  • The exact compensation network, MOSFET choice, and inductor value depend on:
    • controller IC
    • switching frequency
    • PCB layout
    • desired ripple
    • transient response

So a “complete final schematic” requires choosing the exact controller first.

Suggestions for further research

If you continue this design, the most valuable next topics are:

  • non-inverting buck-boost compensation
  • current-mode control
  • inductor ripple current design
  • synchronous rectification timing
  • shunt Kelvin sensing
  • EMI reduction in SMPS layout
  • true 0 V regulation using auxiliary negative rails

A good next engineering step is to decide between:

  1. single-stage buck-boost
  2. two-stage boost + buck

Then the schematic can be finalized around one controller family.

Brief summary

Your specification is feasible, but only under the right conditions:

  • 12 V to 0–30 V at 5 A requires a switch-mode design
  • 30 V at 5 A = 150 W, so the 12 V input must supply roughly 15 A
  • A boost-only circuit is not enough for a true 0–30 V range
  • A linear LM723 / 2N3055 style PSU is not suitable from a 12 V input
  • The correct practical solution is:
    • 4-switch synchronous buck-boost, or
    • 36 V boost + 0–30 V buck

If you want, I can next give you one of these in a full pin-by-pin schematic:

  1. Single-stage LM5175/LM5176-style 12 V to 0–30 V / 5 A design
  2. Two-stage 12 V→36 V boost + 36 V→0–30 V buck design
  3. KiCad-style netlist/BOM with approximate component values

User question

from an existing 230v transformer to produce a 0-30v dc, 5A power supply with CC and CV schematic pdf . This is an upgrade to an old psu.

Artificial Intelligence Response

Direct answer to the question

Yes, you can upgrade an old transformer PSU to 0–30 V DC, 5 A, with constant-voltage and constant-current control, but it is only practical if your existing transformer secondary is roughly 28–30 VAC at 6–8 A. If it is only 24 VAC, you will usually not get a reliable 30 V at 5 A with a linear regulator.

For a classic low-noise bench supply, the most suitable retrofit is:

  • Linear power stage
  • LM723 for the CV loop
  • CA3140 for the adjustable CC loop
  • BD139 driver
  • 3 to 4 external pass transistors such as 2N3055 with emitter ballast resistors
  • Large heatsink + forced air
  • Preferably a small auxiliary 9–12 VAC winding or a small auxiliary transformer for the control rails

This is consistent with the intended use of the LM723: TI specifies that it is a series regulator, adjustable from 2 V to 37 V, supports external pass transistors for currents beyond 10 A, and includes current-limit/foldback application examples in the official datasheet. (ti.com)

Important correction: a plain LM723 circuit does not naturally give true 0 V output, because its normal output range is 2–37 V. For a real 0–30 V supply, add a small negative control rail (for example −5 V) or accept that the minimum output may sit a little above zero. (ti.com)

I cannot attach a binary PDF directly in chat, but I can give you a PDF-ready schematic below that you can copy into KiCad/EasyEDA or print to PDF.

Detailed problem analysis

1) Feasibility of your existing transformer

For a linear 0–30 V / 5 A supply, the raw DC after rectification must remain above the regulated output at full load:

[ V{RAW(min)} \approx V{OUT(max)} + V{dropout} + V{ripple} ]

For a practical linear stage:

  • (V_{OUT(max)} = 30\text{ V})
  • dropout reserve: about 3–4 V
  • ripple allowance: about 2–3 V

So you want roughly:

[ V_{RAW(min)} \approx 35\text{ to }37\text{ V} ]

That means the transformer should usually be:

  • 28 VAC to 30 VAC secondary
  • 6 A to 8 A
  • about 200 VA to 250 VA

Quick rule:

Secondary Result
18 VAC Not suitable
24 VAC Good for about 0–24/27 V linear, but not 30 V at 5 A
28–30 VAC Correct range for 0–30 V / 5 A linear
>30 VAC Possible, but thermal dissipation gets worse

2) Why the thermal design is the real challenge

At low output voltages, a linear PSU burns most of the power in the pass transistors:

[ P{pass}=(V{RAW}-V{OUT})\cdot I{OUT} ]

Example at 5 V / 5 A with about 38 V raw DC:

[ P_{pass}=(38-5)\cdot 5=165\text{ W} ]

That is a very large continuous heat load. This is why many professional bench supplies use:

  • multi-tap transformers
  • relay tap switching
  • or a switching preregulator

If you do not add tap switching, the supply can still work, but 0–10 V at 5 A becomes a severe thermal case.

3) Why LM723 + CA3140 is still a good retrofit choice

The LM723 remains a valid classical choice for a retrofit because it was designed for series regulators, supports an external pass transistor, and TI’s official datasheet includes both external NPN pass and current-limit/foldback example topologies. (ti.com)

The CA3140 is useful for the current loop because it can run from 4 V to 36 V supply and its input common-mode range can go 0.5 V below the negative rail, which is helpful when sensing low-value shunts near ground or slightly below it in a retrofit PSU. (renesas.com)

The 2N3055 is still usable in this type of build, but you must respect its 60 V VCEO, 15 A collector current, 115 W dissipation at case = 25°C, and especially its safe operating area/second-breakdown limits. That is why one transistor is not enough for a reliable 5 A bench supply; use at least 3, preferably 4, plus emitter ballast resistors. (onsemi.com)

4) Recommended topology for your upgrade

I recommend this architecture:

  1. Existing 230 V mains transformer
  2. Bridge rectifier + bulk capacitors
  3. LM723 voltage loop
  4. CA3140 current-limit loop
  5. BD139 driver transistor
  6. 3× or 4× 2N3055 pass bank
  7. 0.1 Ω current-sense resistor
  8. Optional tap switching
  9. Auxiliary 12 VAC or small auxiliary transformer for control power
  10. −5 V auxiliary rail for true 0 V output

This gives a proper lab-style CV/CC crossover:

  • below current limit: CV mode
  • above current limit: CC mode

Current information and trends

The traditional LM723 approach is still technically valid, and TI still documents it as usable in linear or switching regulator roles with external pass devices. (ti.com)

Current design practice, however, tends to prefer one of these for 0–30 V / 5 A:

  • switch-mode CC/CV
  • hybrid switch-mode preregulator + linear post-regulator
  • multi-tap linear with relay selection

For an old PSU upgrade, a linear retrofit still makes sense if you want:

  • low output noise
  • simple repairability
  • predictable transient behavior
  • reuse of transformer, chassis, and heatsink

But if you want continuous low-voltage/high-current use, a hybrid preregulator is the better modern direction.

Supporting explanations and details

Recommended PDF-ready schematic

A. Power input and rectifier

 230 VAC mains
│
F1 2A(T)
│
SW1
│
T1 primary
│
T1 secondary: 28–30 VAC / 6–8 A
│
F2 8A(T)
│
~ ~
BR1 25A bridge
+ -
│ │
│ └────── SENSE_NEG
│
RAW_POS ──┬── C1 10,000µF / 50V
├── C2 10,000µF / 50V
├── C3 100nF film
└── Rb 4.7k / 3W bleeder
to SENSE_NEG

B. Output/pass stage

RAW_POS ─────────────┬───────────────┬───────────────┬───────────────┐
│ │ │ │
C| C| C| C|
QP1 2N3055 QP2 2N3055 QP3 2N3055 (QP4 optional)
E| E| E| E|
│ │ │ │
RE1 0.22Ω/5W RE2 0.22Ω/5W RE3 0.22Ω/5W RE4 0.22Ω/5W
│ │ │ │
└───────┬───────┴───────┬───────┴───────┬───────┘
│ │ │
OUT+

C. Driver stage

RAW_POS ── R1 1k/2W ──┬──────────── CTRL_NODE
│
B │
QD1 BD139
C │──────── RAW_POS
E │──────── DRIVE_BUS ── 47Ω ── base QP1
├─ 47Ω ── base QP2
├─ 47Ω ── base QP3
└─ 47Ω ── base QP4 (if fitted)

D. Current shunt

SENSE_NEG ── RS 0.10Ω / 10W / Kelvin sense ── OUT-

At 5 A:

[ V_{RS}=I\cdot R=5\cdot 0.1=0.5\text{ V} ]

[ P_{RS}=I^2R=25\cdot 0.1=2.5\text{ W} ]

So use a 10 W resistor for margin.

E. Control supply

Auxiliary 9–12 VAC winding, or separate small 230V→12V transformer
│
BR2
│
C10 1000µF
│
7815 ───────── +15V_CTRL
│
7905 or small inverter ─────── -5V_CTRL

Without the −5 V rail, expect the minimum output to be slightly above zero.

F. CV loop using LM723

LM723 powered from +15V_CTRL and -5V_CTRL
Reference network:
LM723 VREF ── RV1 10k ten-turn voltage pot ── OUT-
Wiper = VSET
Output sense divider:
OUT+ ── R11 27k ──┬── VSENSE
│
R12 6.8k
│
OUT-
Compensation:
100pF to 220pF around the LM723 compensation nodes
(as per normal LM723 stability practice)
LM723 output ── 220Ω ──> CTRL_NODE

Use the LM723 in the same external NPN pass transistor style as TI’s official application circuit, but with your BD139/2N3055 pass bank. TI’s datasheet explicitly shows that operating mode. (ti.com)

G. CC loop using CA3140

CA3140 powered from +15V_CTRL and -5V_CTRL
Current set reference:
Take LM723 VREF (≈7.15V) and divide to 0.50V:
R20 = 13k from VREF to node IREF_HI
R21 = 1k from IREF_HI to OUT-
Then:
IREF_HI ≈ 0.51V
RV2 10k current pot between IREF_HI and OUT-
Wiper = ISET (0 to 0.5V)
Current sense:
Kelvin + from OUT-
Kelvin - from SENSE_NEG
CA3140 compares RS drop with ISET
CA3140 output ── D1 ── R22 2.2k ── base QL1
QL1 = MPSA42 (or BC546B minimum)
QL1 emitter = OUT-
QL1 collector = CTRL_NODE

Operation:

  • In normal CV mode, LM723 drives CTRL_NODE.
  • If output current exceeds the RV2 setting, CA3140 turns on QL1.
  • QL1 pulls CTRL_NODE down, reducing BD139 drive and forcing constant current.

H. Protection and finishing parts

DOUT 6A10 diode:
anode to OUT+
cathode to RAW_POS
(protects pass stage during input collapse)
COUT1 470µF / 50V across OUT+ to OUT-
COUT2 100nF film across OUT+ to OUT-
RLOAD 2.2k / 2W across OUT+ to OUT-
(optional bleed/minimum load for cleaner zeroing)
LED_CV and LED_CC can be added from the respective loop outputs.

This scheme uses the LM723 exactly where it is strongest—precision CV control with external pass devices—and uses the CA3140 as a clean adjustable CC clamp, which fits its single-supply/high-input-impedance characteristics. (ti.com)

Ethical and legal aspects

  • You are working directly with 230 VAC mains, so this is not a low-risk build.
  • Use:
    • primary fuse
    • earth bond to chassis
    • insulated mains switch
    • strain relief
    • insulated transistor mounting hardware where required
  • Maintain proper creepage/clearance between:
    • mains wiring
    • transformer primary
    • low-voltage PCB
  • If this PSU will be used by others, sold, or installed in a workplace, local electrical safety and equipment regulations may apply.

Also note that an old PSU chassis may have:

  • degraded insulation
  • dried electrolytics
  • cracked switch contacts
  • poor earth continuity

Check those before reusing it.

Practical guidelines

Minimum BOM

Ref Part Recommended value
T1 Main transformer 28–30 VAC, 6–8 A
BR1 Main bridge 25 A or 35 A, 200 V+
C1, C2 Bulk capacitors 10,000 µF / 50 V each
U1 CV controller LM723
U2 CC controller CA3140
QD1 Driver BD139
QP1..QP3 Pass transistors 2N3055 minimum 3 pcs
QP4 Optional 2N3055 fourth device preferred
RE1..RE4 Ballast 0.22 Ω / 5 W
RS Current shunt 0.10 Ω / 10 W
QL1 CC clamp transistor MPSA42
RV1 Voltage adjust 10 k ten-turn
RV2 Current adjust 10 k ten-turn
Aux supply Control power 9–12 VAC small winding or small extra transformer
Fan Cooling 80–120 mm forced air

Heatsink guidance

For serious use:

  • use a large finned heatsink
  • forced air is strongly recommended
  • fit a thermal cutout around 80–90°C on the heatsink

Calibration procedure

  1. Build with a series lamp limiter on first power-up.
  2. Set RV2 current limit to minimum.
  3. Power on with no load.
  4. Adjust RV1 range so maximum output is 30.0 V.
  5. Connect a dummy load and slowly raise current.
  6. Adjust RV2 so CC begins at 5.00 A.
  7. Verify transition:
    • voltage remains fixed below limit
    • current remains fixed above limit

Dummy load examples

For rough testing:

  • 6 Ω / 200 W for ~5 A at 30 V
  • or several power resistors in series/parallel on a heatsink

Possible disclaimers or additional notes

  • If your transformer is 24 VAC, I would not promise 30 V / 5 A from a linear retrofit.
  • If your transformer has multiple secondary taps, use them. That is the best upgrade you can make for thermal performance.
  • If your old PSU already has a small auxiliary winding, use it for the control supply.
  • If you do not add a negative control rail, the output may stop around 0.1–1.5 V, depending on build details.
  • Three 2N3055s can work, but four is better for SOA margin and lower junction temperature. The 2N3055 datasheet explicitly warns that safe operation is limited by both junction temperature and second breakdown. (onsemi.com)

Suggestions for further research

If you want to improve this design further, consider:

  • relay transformer-tap switching at about 10 V and 20 V output
  • switching preregulator + linear post-regulator
  • digital panel meters with isolated supply
  • remote sensing at the output terminals
  • crowbar overvoltage protection
  • thermal fan control

Also, the best official PDF references for the control core are:

  • TI LM723 datasheet, especially the external NPN pass transistor and foldback current limit application figures
  • Renesas CA3140 datasheet
  • onsemi 2N3055 datasheet for SOA and dissipation limits (ti.com)

Brief summary

Your upgrade is feasible if the transformer secondary is about 28–30 VAC at 6–8 A.
The best classical retrofit is:

  • LM723 for voltage control
  • CA3140 for adjustable current limit
  • BD139 + 3/4× 2N3055 pass bank
  • 0.1 Ω shunt
  • large heatsink + fan
  • auxiliary control supply, ideally including −5 V for true 0 V output

If you want, in the next message I can convert this into one of these two formats:

  1. clean A4 schematic sheet for print/PDF, or
  2. KiCad-style netlist/wiring list.

If you send your transformer secondary details, for example:

  • 24 VAC 6 A
  • 28 VAC 8 A
  • 2 x 15 VAC
  • any extra auxiliary winding

I can finalize the values for your exact transformer.

Ask additional question

Wait...(2min)
Disclaimer: The responses provided by artificial intelligence (language model) may be inaccurate and misleading. Elektroda is not responsible for the accuracy, reliability, or completeness of the presented information. All responses should be verified by the user.
Report illegal content (Article 16 DSA)