DIY Analog Variable Bench Power Supply W/ Precision Current Limiter





Introduction: DIY Analog Variable Bench Power Supply W/ Precision Current Limiter

About: 北部九州アラフィフ(Around 50)の波乗りオヤジ。波がないときは家庭菜園で有機野菜を作り、かみさん子供と庭でバーベQ。第三のビール風飲料と芋焼酎のロックが大好きなちょっとメタボなトドお父さん。玉子に薀蓄。タマゴかけごはん愛好会 会員番号 100012

In this project I will show you how to use famous LM317T with a Current Booster power transistor, and how to use Linear Technology LT6106 current sense amplifier for precision current limiter .This circuit may allows you to use up to more than 5A, but this time it is used for just 2A light load because I choose a 24V 2A relatively small transformer and a small enclosure. And I prefer output voltage from 0.0V, then I add some diode(s) in series to cancel LM317 minimum output voltage 1.25V. this spec. also allows you to short circuit protection.
Those circuits are combined to create an analog variable bench power supply which generate 0.0V-28V and 0.0A-2A with precision current limiter. The regulation and noise floor performance is pretty good in comparison with similer DC-DC converter based power supplies. Therefore this model is better to use especially for analog audio applications. Let's get started !

Step 1: Schematic and Parts List

I'd like to show you the whole schematic of this project.

I had divided the hole schematic in three parts for easy explanation.
① AC Input section、② Middle section(DC Control circuits)、③ Output section.

I'd like to continue explain the parts list for each section respectively.

Step 2: Preparing to Drill the Case and Drilling

We should collect the exterior parts and drilling the case (enclosure) first.

The case design of this project was done with Adobe illustrator.

Regarding the placement of parts, I made a lot of trial and error considering and deciding as a first photo shows.

But I love this moment cause I can be dreaming what shall i made ? or which is better?

It's like a waiting good wave. It is really precious tiime at all! lol.

Anyway, I would like to attach file and .pdf file also.

To prepare for case drilling, print the design to A4 size adhesive paper and stick it to the case.

It will be marks when you drilling the case, and it will be the cosmetic design for the enclosure.

If the paper got dirty, please peel it off and stick the paper again.

If you prepared for case drilling, you can start the drilling the case according to the center marks on the case.

I strongly recommend you to describe the size of holes on the sticked paper as 8Φ, 6Φ like that.

Using tools are an electric drill, drill bits, step drill bits, and a hand nibbler tool or dremel tool.

Please be careful and take enough time to avoid accident.


Safety glasses and Safety Gloves are necessary.

Step 3: ① AC Input Section

After finishing the case drilling and finishing, let's start making the electric boards and wiring.

Here are the list of parts. Sorry for some links are for Japanese seller.

I hope you can get similar parts from your nearby sellers.

1.Used parts of ① AC Input section

Seller: Marutsu parts
- 1 x RC-3:

Price: ¥1,330 (approx.US$12)

- 1 x 24V 2A AC Power Transformer[HT-242]:

Price: ¥2,790 (approx.US$26) if you like 220V input, choose [2H-242] ¥2,880

- 1 x AC code with a Plug:

Price:¥180 (approx.US$1.5)

- 1 x AC Fuse box 【F-4000-B】 Sato Parts:
Price:¥180 (approx.US$1.5)

- 1 x AC Power Switch(Large) NKK 【M-2022L/B】 :
Price:¥380 (approx.US$3.5)

- 1 x 12V/24V Switch(small) Miyama 【M5550K】 :
Price:¥181 (approx.US$1.7)

- 1 x Bridge rectifire diode(large) 400V 15A 【GBJ1504-BP】:
Price:¥318 (approx.US$3.0)

- 1 x Bridge rectifire diode(small) 400V 4A 【GBU4G-BP】 :
Price:¥210 (approx.US$2.0)

- 1 x Large condenser 2200uf 50V【ESMH500VSN222MP25S】:

- 1 x 4p Lagged terminal 【L-590-4P】 :

Sorry for inconvenient link to Japanese site, please search seller handling similar parts with referring those links.

Step 4: ② Middle Section (DC Control Circuit)

From here, it is the control part of the main power supply DC voltage.

The operation of this part will be explained later based on simulation results as well.

Basically I am using the classic LM317T with a large power transistor for large current output ability until 3A alike.

And to cancel 1.25V LM317T minimum output voltage, I added D8 diode for Vf to Q2 Vbe.

I guess Vf of D8 is approx. 0.6V and Q2 Vbe also approx. 0.65V then the total is 1.25V.

(But this voltage is depend on If and Ibe, so taking care is needed to use this method)

The part around Q3 surrounded by dotted line is not mounted. (for optional for future thermal shutdown feature.)

Used parts is as below,

0.1Ω 2W Akizuki Densho

heat sink 【34H115L70】Multsu Parts

Rectifier Diode (100V 1A) IN4001 ebay

LM317T Voltage Control IC Akizuki Denshi

General Purose NPN Tr 2SC1815 Akizuki Denshi

U2 LT6106 Current Sense IC Akizuki Denshi

Pitch convert PCB for LT6106 (SOT23) Akizuki Denshi

U3 Comparator IC NJM2903 Akizuki Denshi

POT 10kΩ、500Ω、5KΩ Akizuki Denshi

Step 5: ③ Output Section

The last part is Output Section.

I like retro analog meters, then I adopted analog meter.

And I adopted a Poly Switch (resettable fuse) for output protection.

Used parts is as below,

Resettable fuse 2.5A REUF25 Akizuki Denshi

2.2KΩ 2W bleeder registor Akizuki Denshi

32V Analog volt meter (Panel meter) Akizuki Denshi

3A Analog volt meter (Panel meter) Akizuki Denshi

Output Terminal MB-126G Red and Black Akizuki Denshi

Universal Bread Board 210 x 155mm Akizuki Denshi

Terminal for bread board (as you like) Akizuki

Step 6: Finish Asembling and Testing

So far, I think that the your main board was also completed.

Please proceed with wiring to parts attached to the case like pods, meters, terminals.

If you finished making the project.

Final step is testing the project.

This analog power supply basic specifications are

1, 0~30V output voltage coarse adjustment and fine adjustment.

2, 0~2.0A output current with limiter ( I recommend to use under transformer spec.)

3, Output voltage changing switch on back panel for reducing the the environmental loss

( 0~12V, 12~30V)

Basic Testing

Testing the circuit work.

I used a 5W 10Ω resister as a dummy load as shown on the photo.

When you set 5V, it provide 0.5A. 10V 1A, 20V 2.0A.

And when you adjust current limit to your favorite level, the current limiter works.

In this case, output voltage is getting lower according to your adjusting output current.

Oscilloscope waveform Testing

I'd like to show you oscilloscope waveforms also.

The first waveform is voltage rising waveform when you turn on the power of the unit.

CH1(Blue) is just after rectifier and 2200uF capacitor approx. 35V 5V/div).

CH2(Sky blue) is output voltage of the unit (2V/div). It is adjusted to 12V and reduced the input ripple.

The second waveform is enlarged waveform.

CH1 and CH2 are now 100mV/div. CH2 ripple is not observed due to LM317 IC feedback is correctly working.

Next step, I'd like to test at 11V with 500mA current load (22Ω 5W ). Do you remember Ohm's low I = R / E ?

Then CH1 input voltage ripple getting bigger to 350mVp-p, but no ripple observed on CH2 output voltage also.

I'd like to compare to some DC-DC back type regulator with same 500mA load.

Large 200mA switching noise is observed on CH2 output.

As you can see,

General speaking, Analog power supply is to suit for low noise audio application.

How about that ?

If you have further question, please feel free to ask me.

Step 7: Appendix 1 : Circuit Operation Details and Simulation Results

Wow, so many reader's over 1k was visited to my first post.

I'm simply grad to see the numerous view counter.

Well, I'd like to return to my subject.

Input Section Simulation results

I have used LT Spice simulator to verify the circuit design.

Regarding how to install or how to use LT Spice, please google it.

It is free and good analog simulator to learn.

The first schematic is a simplified for LT Spice simulation and I'd like to attach .asc file too.

The second schematic is for input simulation.

I defined a voltage source DC offset 0, amplitude 36V, freq 60Hz, and input resistor 5ohm as comparative specs for the transformer. As you know, transformer output voltage is displayed in rms, then 24Vrms output should be 36Vpeak.

The first waveform is voltage source + (green) and bridge rectifier + w/ 2200uF(blue) . It will go to around 36V.

The LT Spice couldn't use variable potentiometer, I'd like to set fixed value to this circuit.

Output voltage 12V current limit 1A like that. I'd like to proceed to next step.

Voltage Control Section using LT317T

The next figure shows LT317 operation, basically LT317 works as so called shunt regulator it means that Output voltage pin to Adj. pin is always 1.25V reference voltage regardless for input voltage.

It also means a certain current bleeds in R1 and R2. The current LM317 adj. pin to R2 also exist, but too small as 100uA then we can neglect it .

Until this far, you can clearly understood the current I1 which bleed in R1 is always constant.

Then we could made the formula R1 : R2=Vref(1.25V) : V2 . I choose 220Ω to R1, and 2.2K to R2,

Then the formula is transformed V2= 1.25V x 2.2k / 220 = 12.5V. Be aware of real output voltage is V1 and V2.

Then the 13.75V is appears on LM317 output pin and GND. And also aware to when R2 is zero, 1.25V output


Then I used simple solution, I just use the output transisitor Vbe and diode Vf to cancel 1.25V.

General speaking Vbe and Vf is around 0.6 to 0.7V. But you also have to aware to Ic - Vbe and If - Vf charactoristcs.

It shows that a certain bleeder current is needed when you use this method for cancel 1.25V.

Therefore I add a bleeder register R13 2.2K 2W. It bleeds approx. 5mA when 12V output.

Until this far, I'm a little bit tired to explain. I need lunch and lunch beer.(lol)

Then, I'd like to continue to next week gradually. So sorry for your inconvenience.

Next step I would like to explain how current limiter works precisely, using LT Spice load parameter step simulation.

Current Limiter Section using LT6106

Please visit Linear Technology Site and see the datasheet for LT6106 application.

I'd like to show the drawing to explain Typical application that describes AV=10 for 5A example.

There is a 0.02 ohm current sense register and sensed output from out pin is now 200mV/A then

the out pin would rise up to 1V at 5A, right ?

Let's think about my application with this typical example in mind.

This time we would like to use current limit under 2A, then the 0.1 ohm is suitable.

In this case out pin rise 2V at 2A ? It means the sensitivity is now 1000mV/A.

After that we have to do, just turn ON / OFF the LM317 ADJ pin with the generic comparator

like NJM2903 LM393, or LT1017 and generic NPN transistor like 2SC1815 or BC337?

which cut off with the detected voltage as the threshold.

Until this far, circuit explanation is over, and let's start complete circuit simulations !

Step 8: Appendix 2 : Circuit Step Simulation and Simulation Results

I'd like to explain so called step simulation.

Usual simple simulation simulates just one condition, but with step simulation, we can change conditions continuously.

For example, step simulation definition for load register R13 is shown next photo and below.

.step param Rf list 1k 100 24 12 6 3

It means R13 value shown like {Rf} varies from 1K ohm ,(100 , 24 , 12 , 6) to 3 ohm.

As obviously understood, when 1K ohm current drawn to load R is ①12mA

(because output voltage is now set to 12V).

and ②120mA at 100 ohm, ③1A at 12 ohm , ④2A at 6 ohm , ⑤4A at 3 ohm.

But you can see the threshold voltage is set to 1V by R3 8k and R7 2k (and voltage for comparator is 5V).

Then from condition ③, the current limiter circuit is supposed to work. The next drawing is simulation result.

How about that until this far ?

It may be a little bit difficult to understand. because the simulation result might be difficult to read.

Green lines show output voltage and blue lines show output current.

You can see voltage is relatively steady until 12 ohm 1A, but from 6 ohm 2A voltage decrease to 6V to limit current to 1A.

You can also see the DC output voltage from 12mA to 1A is a little bit dropped.

It is almost caused by Vbe and Vf un-lineality as i explained in former section.

I'd like to add next simulation.

If you omit D7 on simulation schematic as attached , the output voltage results would be relatively stable.

(but output voltage is getting higher than previous, off course.)

But it is a kind of trade off things, because I'd like to control this project from 0V even if the stability is a little bit lost.

If you start using analog simulation like LT Spice, it is easy to check and try your analog circuit idea.

Ummm, eventually it seems like I have finished complete explanation eventually.

I need a couple of beer for weekend (lol)

If you have any question about this project, please feel free to ask me.

And I hope all of you would enjoy good DIY life with my article !




    • Clocks Contest

      Clocks Contest
    • Casting Contest

      Casting Contest
    • Make it Move Contest

      Make it Move Contest

    We have a be nice policy.
    Please be positive and constructive.




    Please note that D6 and D9 are drawn with inverted polarity!

    Please also note that the output on/off switch doesn't switch off immediately, because output capacitors are not immediately discharged. You should put the switch after output capacitors.

    1 reply

    Thank you for your kind advice.

    I checked the schematic and as you mentioned D6 and D9 direction was wrong.

    I have fixed the schematic as real my circuit board.

    But regarding on/off switch, actual circuit is same as the schematic.

    It may not be a problem because output cap 470uF relatively immediately

    discharged by R13 2.2k bleeder register.

    But I don't intend to restrict circuit change for better result.


    Thank you very much for your comment.

    Those days are digital age, therefore analog circuit explanation
    is hard to find. How to add current limiter?
    That is motivation for this instructables.
    Actuallly, I would like to add explanation for the circiuit explanation.
    But for us Japanese, describing in English is very hard.

    I'd like to add explanation gradually at weekend.