For DC enthusiasts…

Is there a way to apply the “high voltage” principle to a DC layout? Yes, but it is not enough to create an AC voltage, which is very easy with a circuit such as the well known NE555 or CMOS logic gates, and then apply the voltage multiplication seen earlier. The main obstacle is the fact that the voltage in the rails is variable, from 0 to 12 V (theoretically).

What is needed is a device that creates an if possible constant and relatively high voltage, say between 24 V and 60 V, from the variable DC voltage, and this with a not too bad efficiency.

The step-up voltage converter

Well, there is such a thing, and it is called a boost or step-up converter. There are many such integrated circuits, which makes it difficult to choose one. Their operation involves an inductor which, like the capacitors in voltage multipliers, will act as an energy reserve.

Now, it is possible to find small, inexpensive and space-saving modules on certain sales websites, that seem to be perfect for this purpose.

Here are the main characteristics of the one I found the most interesting.

We can see that, by connecting the load circuit (LED) between the + and − outputs, we obtain an interesting voltage of 48 V, from an input voltage as low as 3.6 V (and even less, as we will see), and with an ample current, since a few milliamperes should be enough.

Converter module

On the other hand, the dimensions are very small, except perhaps the height, which depends on the most bulky component: the inductor, of 33 µH. It is calculated for the maximum output current. By limiting the output current to 10 mA for example, a much smaller SMD inductor could be used, which would reduce the height by almost 2 mm.

This device uses an XL6007 integrated circuit from XLSEMI. Given the characteristics of this IC, only the value of a resistor (R1 on the photo above, but R2 in the data sheet) needs to be modified to obtain the desired output voltage, up to 60 V. We can therefore easily consider increasing this output voltage from ±24 V to ±30 V, which brings us back to the DCC case!

Note: Why not use a 60 V single voltage — without a negative voltage, which limits the output current? The reason is that the capacitors used may not (and probably don’t) have a sufficient operating voltage.

Operating diagram

Starting with the same load as before, i.e. 12 LEDs in series, the DC diagram will be as follows. The rectifier is necessary to take into account the polarity inversion of the track. The capacitor C1 is not essential, but it is useful for filtering out bad contacts with the track and, in the case of a pulsed supply, for smoothing the voltage applied to the converter.

Operating diagram

Testing the module

For this test, I soldered pins on the terminals, so that the module can be installed on a breadboard.

Test conditions: 12 LEDs in series. Current in the LEDs set to 8 mA. DC input voltage varied from 2.5 to 15 V. LED circuit connected between the +24 V and −24 V outputs of the module. 100 µF / 63 V capacitor across the load.

Assembly view — very simple!

Test assembly

Some pictures taken for different input voltages.

Test at 3.5 V

Test at 7.5 V

Test at 15 V

The LEDs light up with the same intensity.


The power and efficiency are calculated from the voltages and currents.

VIN (V) IIN (mA) VOUT (V) IOUT (mA) PIN (mW) POUT (mW) Eff. (%)
2.5 20 36.5 0.6 50.0 21.9 43.8
2.8 62.0 41.4 3.3 173.6 136.6 78.7
5.0 106.0 48.0 7.9 530.0 379.2 71.5
7.5 74.0 48.0 8.0 555.0 384.0 69.2
10.0 57.6 48.1 8.0 576.0 384.8 66.8
12.0 48.0 48.0 8.0 576.0 384.0 66.7
15.0 36.0 47.9 8.0 540.0 383.2 71.0

It can be seen that the desired current is almost reached from 5 V onwards, but it is already sufficient at 3 V for the LEDs to light up visibly, which seems to me to be correct for a DC powered layout, as locos starting at a voltage below 3 V must be quite rare.

Curve giving the output current as a function of the input voltage:

Graph IO = f(VI)

Test in situation

I need a coach not yet equipped with a lighting strip, if possible with a “rustic” pickup, i.e. with thin strips, and easy to dismantle. I chose my old A4Dtuxj Hornby Acho, which has pickups for the rear lights.

I also need a strip with LEDs in series. I get the old strip from a B11 Est Roco coach, which had four groups of three LEDs in series, plus one by itself. I just cut the tracks that form the parallel connection, and connect everything in series. This gives me thirteen LEDs in series, with a 20 kΩ resistor. I complete this assembly with two 47 µF / 63 V capacitors, which is not a big capacity, is it?

Test assembly on a Hornby Acho coach

The converter output is connected to the strip (white wires), and its input to a rectifier (obliquely mounted component), to which the pickup wires are connected.

Test assembly, close-up view

To make things worse, I’ll use an old BB 16001 Jouef (as all my recent machines are DCC equipped), and a no less old Disjoncta transformer! I’ll pass on the not very fluid functioning of the whole, but the light strip lights up well as soon as the loco starts, and keeps a constant luminosity. Except that, when I set the speed to the best possible idle, the light tends to switch off.

Here is a video of the tests. Show video.

Checking with my laboratory power supply, which gives a much better performance, I notice that the LEDs only light up above 4.5 V (compared to 3 V on the table). Explanation: the rectifier adds a voltage drop of 1.4 V. To improve this, a Schottky rectifier should be used.

Finally, let us note (and I tested it) that such a strip also works in DCC, which makes it a “universal” strip.


With this simple and inexpensive module, the same advantage is obtained in variable DC as in DCC: little flickering of the lighting with a capacitor of moderate capacity, and, moreover, lighting of constant intensity over a wide voltage range.

Dual Mini DC-DC Boost Step Up Converter
About 2.40 € (+ 0.80 € shipping)
on ebay

Viking NL12KTC330
CMS 1812 inductor
4.5 × 3.2 × 3.2 mm
1.43 € per 10
at tme