Category Archives: Power Supply

Add Secondary Windings to your Toroidal Transformer for Additional Voltage Outputs

So here’s the scenario… you have a power transformer with a voltage output of 25V + 25V AC.   Let’s say, you’re using this power transformer to supply power to a power amplifier project.  So it’s all good.

But you also need 5 Volts DC to power some logic circuits (a microprocessor, Arduino, PIC microcontroller or similar) within the same project.  Nowadays, incorporating microprocessors in DIY projects is common… thanks to the popularity of Arduino. You may be using this microprocessor to integrate an LCD screen, or perform temperature monitoring, or perform speaker overload protection.

What are your options?

1. You can use a voltage regulator and step down the high voltage DC output of your power supply to 5VDC. You can use a 7805 regulator, or an LM317HVT 3-terminal regulator. But this is wasteful, and will generate a lot of heat!  25VAC rectified to DC will be approximately 35VDC, give or take a few. If you regulate this down to 5VDC, you’ll be dissipating 30Volts in your regulator as heat! This will require a big heatsink.  In this example, we’re using a 25+25VAC transformer. But what if your transformer is bigger, say 50+50VAC output, you’ll need to bring down 70VDC (after rectification) to 5VDC.  Yikes, you need to dissipate 65Volts of energy.  Not to mention, this may be exceeding the Safe Opearting Area (SOA) of your regulator. (See datasheet on maximum voltage differential, Vout – Vin allowed by your regulator.)

2. You can install an additional small power transformer in your project. This is feasible, and one can do this instead. The smaller transformer can be another toroid, or even a small PCB mounted transformer. The cons of this though is the cost.  Buying another transformer for your project will be expensive. If this is a PCB mounted transformer, then you need to design a board to accomodate this transformer.  Not to mention, the additional transformer and PCB will take up valuable real-estate space in your already cramped chassis.

Hmmmm…. adding that Arduino to your power amplifier project seems to be a lot of hassle. Now, you’re thinking of giving up the idea  of having a cool-looking LCD display and push buttons and rotary encoders on your project.

But wait… there’s another solution!  How about just installing additional secondary windings to your existing toroidal transformer?

It’s very easy, let me show you how:

So let’s say this is your power transformer:



What we need to do is just wrap another layer of wire around this transformer. So you’ll need magnet wire (check out eBay for sellers selling magnet wires, so you don’t have to buy several thousand feet of spool.)

I used the same magnet wire thickness just to keep things simple.



Step 1: Wrap 10 turns of magnet wire around your power transformer.  I used masking tape to keep the wire in place.



Step 2: Strip the insulation from the ends. It’s coated with lacquer so you can either use a sandpaper or blade to scrape off the insulation and get to the bare copper. I personally used a thermal wire stripper since I have one.



Step 3: Measure the AC voltage you got from your 10-turn secondary winding.

In my case, I got 2.74Volts AC from my 10-turn winding.



Step 4. Compute how many volts per turn we got.

2.74Volts / 10 turns = 0.274 Volts per turn

Using this information gathered from our transformer, we can now compute how many turns we need for the voltage output we want.

I also measured the length of wire needed to make that 10-turn winding. For my power transformer size, it takes about 5.5ft of wire to make that 10-turn winding.

Step 5. I want to use 5 Volts DC to power my Arduino… and I want it regulated.  So we need at least 7 Volts DC rectified so we can get 5Volts regulated output.   So let’s shoot for 5Volts AC for our secondary winding add-on.

That means, we’ll need:

5VAC / 0.27 Volts per turn = 18.51 turns, round up to 19 turns

In this case, I decided to round up to 20 turns for a nice round number.

And we’ll need approximately 11 feet of magnet wire to wind 20-turns.

…. so go back to step 1, but this time winding 20-turns around your power transformer.

I had some extra wire left, so I just went for another extra turn… 21 turns total.   Doing some math, we should get around 5.67 Volts AC (21 turns x 0.27 Volts / turn).

After doing this, I measure the output voltage and I got 5.8Volts.  Very very close to our computed 5.67 Volts.



Perfect…  so this means our 5.8VAC rectified to DC will be about 8.2VDC, which is more than plenty enough for our project.

Now, all we need to do is rectify this AC voltage to DC, and regulate this down to 5VDC. And we’re done.


High Voltage, High Current Power Supply

I plan on building a solid state power amp, but don’t know which one yet. In the meantime, I figure I can go ahead and start making the PSU for my future amplifier.

I tried iTead Mall to place my order for (5) PSU boards. So far, I’m pleased with iTead’s quality, though there was some miscommunication with them. They didn’t update my order status for 2 weeks, it still says “Processing” on their website, when in reality, they’ve already shipped my order but didn’t email me a tracking #. Anyway, once I submitted a ticket, they’ve been very responsive and apologetic. Then by coincidence, I got my PCB in the mail the day after I complained. I will be using them again in the future.

So basically, I want the PSU to be able to cater to higher voltages, like +/-80 or 90VDC… so I made sure there’s enough room on the board to allow the use of bigger capacitor, 100+V rated or more.

The PSU board can accept either spade connectors, or terminal blocks.

I’m using Ultra Fast Recovery Rectifiers, bypassed with Panasonic Safety Class X2 polyester film caps (250V rated). I have heatsinks on the rectifiers but they don’t need it. They’re cool to the touch.

There are green LED indicators under the fuse on the DC output, so if the fuse blows, the LED also goes out and you can quickly see the problem. There’s also a couple of AC fuses.

It’s basically just a simple power supply, albeit using high current/high voltage parts.

Step 1: Bare PCB … ready for population and build.


STEP 2: Insert 0.1uf / 100V ceramic caps. These are bypass capacitors for the DC output. If you need to add more bypass caps, you can solder them underneath the board across the capacitor terminals.


STEP 3: These are the current-limiting resistors for the LED power indicator. The 3mm LEDs are located under the fuse on the DC output.


STEP 4: Fuse holder clips…. 2 fuses for the AC primary side, and 2 fuses for the DC side. Depending on the VA size of your power transformer, pick the appropriate/correct fuse rating.


STEP 5: These are the 4x Ultra Fast Recovery Rectifiers, rated 600V Reverse Voltage, 10Amps. Added TO-220 heatsinks just to be sure. Yes, I know… it’s overkill.


STEP 6:  I made a mistake of trusting the datasheet… the component size is correct but the lead spacing is wider than I expected. So here’s a temporary fix to make it work with the PCB.HVPSU-Step5

STEP 7: These are the (4x) Safety Polyester Film Capacitors, rated 250Volts, Class X2. Each capacitor is connected across each Rectifier Diode.


STEP 8: Terminal Blocks with screws. Heavy-duty, rated up to 32Amp, and can handle up to AWG#10 gauge wire. AC side accepts dual secondaries from power transformer. DC side outputs split voltage, V+, V- and GND.


STEP 9: 10,000uf capacitors for DC Filtering. Since this is a split power supply, there’s 20,000uf for each V+ or V- voltage rail. Use the appropriate voltage rated capacitor matched to the power transformer, and DC output you need. The higher the capacitor voltage rating, the more expensive it gets. There’s enough room on the board to accept bigger diameter/higher voltage capacitors up to 100Volts DC rated or more.


Another view of the finished High Voltage Power Supply.


Size comparison of finished High Voltage, High Current Power Supply compared to an iPad Air. PSU weighs in at 11.1 ounces.


Power Supply Testing.


Ripple voltage is 11.5mVolts peak-to-peak, 25V+25 AC input, 35.4VDC + 35.4VDC output.




The above screenshot for ripple is wrong. I’m basically measuring noise in the environment, instead of the actual PSU ripple. Here are updated screenshots.

There you go, that’s more like it… 2.5mVpp ripple.



And we know it’s the ripple from the PSU, and not some environment noise because the frequency is around 120Hz…. ripple from a full wave rectifier.