YO3BN - Linear Power Supply 1.3V-15V 1.5A

Posted at - 28th May 2017

Linear Power Supply 1.3-15V 1.5A

Because I wanted to replace my old, bulky, heavy, crappy power supply, I have designed a new one aiming the following:

simple design
as lightest as possible
compact size
good ripple rejection
main noise filtering
15V at 1.5-2A max.

I plan to use it as a general power supply, charging accumulators, providing current for my QRP transceivers, etc. From my experience, I found cheap switching power supplies very noisy regarding RF emissions especially for HF spectrum, therefore for this power supply I'll use a 50Hz power transformer. A toroidal transformer was chosen over a classical E-I one, because of its high efficiency and weight/power ratio.

After an hour of looking on TME about components and their characteristics, finally I found the right transformer 18V 1.7A 32VA 480g.

The Schematic

The transformer have an initial voltage about 19.5Vrms, so when rectified, this becomes 19.5 * √(2) = 27.5V, so care must be taken for the first electrolytic capacitor regarding its voltage rating, in my case I used a 10000uF rated at 35V. The rectifying diodes are schottky low voltage drop, about 0.15V, 3A at 40V reverse.

The LM317 claims to have overload thermal protection, current limiting protection, very good ripple rejection, 1.5A, adjustable voltage. So, it was very well suited for my project. The D5 and D6 provide low impedance discharge path for C4, C5, C6 and C3, protecting the LM317 circuit. The C3 is used to further reduce the ripple rejection. RV1 and R1 are required to set the output voltage. According to LM317 datasheet, the output voltage can be calculated using the following formula:

Vo = Vref * (1 + RV1/R1) + (Iadj * R1)
where,

Vref is 1.25V
Iadj is 50uA

So, for a 5K potentiometer in order to deliver 15V maximum output the R1 must be around 450 ohms. Before calculating R1, the potentiometer RV1 must be measured first!! Because RV1 could not be exactly of 5K ohms, the R1 can be a trimmer pot in order to exactly match the wanted output voltage without messing around with R1 values. Also, don't do measurements without filtering capacitors C2, C6 because the voltage will be much lower than expected!!

The filter is used to cancel out the noise coming from main line acting as a common current choke. It is formed from two twisted conductors of 1mm diameter wound on a KEMET ESD-R-25D-8 toroidal core.

The last electrolytic cap of 22000uF/25V is used to further smooth the current. I have intended to use a 22000uF for C2 too, but I didn't had one with big voltage rating at that time, so 10000uF/35V was used.

Construction Details

The LM317 was insulated from the alu radiator.
The source is protecred by a temporised fuse, not shown in schematic. Its value can be calculated using transformer VA value divided by 230V, 32VA / 240V = 0.133A, so taking into account the transformer's efficiency too, the fuse value should be around 150-200mA.

Measurements

The power supply can provide adjustable voltage from 1.3V to 15.0V at 1.5-2A max. The voltage remains steady for short periods (minutes) of 1A consumption, but when used for long periods of high current drawn, the radiator becomes hot approx. 60-70° C and the voltage slightly drops about 0.3V, which actually isn't a problem.

When used to power my 80m QRP 5W transceiver it does not introduce any hum or noise, nor for transmission, neither for reception, which is exactly what I wanted.

It weighs approx. 1.04Kg.

Kicad Files

power-supply-15V-1.5A

73 de YO3BN