Here is a unit that can be used to control the on/off switching of remote devices. Designed primarily for the modelling radio-control fraternity, devices such as boat lighting, or model plane bomb doors, etc, can be operated from any channel on your R/C transmitter.
The circuit shown here uses a MOSFET output stage - part number IRF630. Specifications state that the continuous maximum current handling is 9 amps. But because I only need to switch relatively small currents, there is no need for a heatsink. The safe maximum current output without one is in the order of 6 - 7 amps. This is an improvement on an earlier version which employed a Darlington transistor, handling just 3 - 4 amps.

A few other small refinements have also been made:

• The inclusion of two protection diodes at the output.
• The LED is now routed to the positive supply rail. It will glow at the same time as the output load is ON, making setting up much easier.
• Mulitple output configurations are now possible via a 5-way screw terminal block.
• The option of a component overlay, giving the effect of a dedicated printed circuit board.

Begin by cutting a piece of veroboard, 22 holes x 11 tracks (fig.1).
Flip the board over, then make 10 breaks at the following locations.

Because two of them are located under the chip, it is important that the wire links are soldered first. Use single-strand telephone wire or similar.
But notice that the links shown in columns 13 and 19 are thicker. These carry the negative supply rail to the MOSFET, and the positive feed-thru from an external battery to the output load. And because high currents will flow, you'll need to use stout wire. For example: copper wire used for domestic house lighting ciruits, or 45Amp fuse wire.

The components can be soldered in any order of preference. But some things to be aware of...

• The servo signal wire is usually white or yellow. Solder it at location F1. Solder the pos. wire at H1 and the neg. wire at K1.
• The wiper leg of the trimpot should be at location G5. The other two legs at locations E4 and G3.
• The LED cathode (denoted by a 'flat' on the body) at location A4. The anode at B4.
• The 10k resistor at E2 and D5. The 470-ohm resistor at A12 and F12.
• Pin 1 of the CD4013 at location D7.
• Insert the MOSFET with its metal tab facing towards the 470-ohm resistor.
• Insert the two diodes that their anodes are at locations D17 and D18. Their cathodes are denoted by a coloured band at one end. Solder the cathodes at F17 and B18.

Probably a neater way of doing it would be to paste a parts-placement overlay to the board first. It's just a matter of then stuffing all the components where they are shown.
Although aligning the overlay to the board can be a bit tricky, there are perhaps two advantages:

1. Reduces the chance of soldering a component in wrong location.
2. Gives the board a neat PCB-type appearance.

• Turn VR1 trimpot fully counter-clockwise.
• Connect the servo plug to, say, the throttle channel at your receiver.
• Power-up the receiver and the transmitter.
• Push the throttle stick to a position where you want the switch to operate.
• Holding the stick at this position, tweak the trimpot until the on-board LED just turns on.

When you return the stick to neutral, the LED should extinguish. Push the stick forward again and the LED should start glowing from the point where it reaches your pre-set position.
If the output works in the opposite to how you want (i.e.: LED glows when stick at neutral), simply reverse the servo switch at your transmitter.

Note: Initial testing of the unit can also be carried out with the aid of the servo tester, seen elsewhere on these pages.

Figs. 4a thru 4d will alternate every 10 seconds between the veroboard layout and the parts-placement overlay. No reason for this other than to show that the off-board connections are the same, regardless of which version is built.

In order to power devices from your receiver battery, you'll need to use the two terminals shown here in 4a. If you are using polarised devices (in this case a LED), connect it as shown - the positive (+) at the yellow wire; the negative (-) at the blue wire. Non-polarity devices, such as incandescent lamps, can be connected either way round.
But bear in mind that you don't want to put too much drain on your Rx battery. A standard 2-volt LED, through 150-ohm resistance, will only draw a measly 20'ish milli-amps. Even ten parallel LEDs at 200mA will still not put too much strain on a fresh battery. But although the output can handle as much as 7+ amps, your receiver and servos will soon complain if you start to overdo it. A bit of common savvy will tell you that hanging a 5-amp motor off the end is overdoing it.

So in order not to prematurely drain the Rx battery, you can power your devices instead from a separate supply. Figure 4b shows how.
Here you see the same low-current load at the output, but note that the anode wire is now at a different terminal. This terminal is linked via link J8 to the positive wire of the separate battery. In other words, the load is now independent from the Rx supply.
A manual on/off switch will not be needed, because when the MOSFET is not conducting, off really is OFF.

If you're so inclined, you can hang more than 7-amp's-worth of LEDs here. But if you're using, say, a standard 9-volt battery, it just won't have the power to drive them. More juice is needed...

...and in order to supply that juice, a large-capacity battery is called for. The battery in fig.4c is supposed to represent a hi-capacity pack. It's not drawn to scale, so a little imagination is needed. This could be a 12-volt gel-cell... a 1.2-volt, 4-amp ni-cad... Nickel-metal hydrides... nuclear reactor... you name it. And instead of a LED, a heavy-duty load is now shown. The wiring connections are the same as 4b, only this time a fuse is shown in series with the battery positive wire. The rule-of-thumb is to use a fuse which is rated approximately 10% higher than your anticipated maximum loading.
As you probably know, relays and motors are notorious for causing unwanted radio-frequency interference. They also create massive voltage spikes whenever the current to their coils stop flowing (sometimes known as Back Electromotive Force). This back EMF can run into hundreds, sometimes thousands, of volts, causing MOSFETs and transistors to frazzle in a split second. The inclusion of the two diodes at the output prevent this from happening. And if anything untowards does happen (i.e: accidental power reversal), then the fuse will blow, protecting not just the load but the MOSFET also.

It seems the best type of fuse to use these days are those designed for automobile use. Some would say it's because the 'spade' connections offer superior metal-to-metal contact over those of the 'in-line spring' types. My personal choice would be to opt for the auto' types, for several reasons. One being that they're smaller. But experience-wise, neither type has caused any problems in the past. The choice of fuse boils down to your own personal preference... or whatever might be available in your spares box.
As an aside, a brand-new ESC was ruined simply because I got lazy and omitted a fuse. Thirty-nine pounds wasted before I realised that I'd accidentally stuffed 14 ni-cads-worth of volts where the motor should have been. A simple fuse would have prevented that.

Here the output is shown connected to a standard glowplug. Most internal-combustion engines (especially four-stroke'ers) tend to run more smoothly at idle with the glowplug connected. Here the trimpot is adjusted that power is turned off at the glowplug when the throttle stick is advanced just past the idle position.
Again, note the inclusion of the fuse. Although fig.4d shows just a single cell, it's current capacity is still capable of causing wires to burst into flames if the terminals should short-circuit, or if the glowplug were to short.
With hindsight, it's easy to see the importance of fuses. They cannot be over-emphasised.

The image shown on the right is obviously too large in the real world. It has to be reduced in size first. So once you've copied it to your drive, and in order to have the printout match the actual board dimensions, you need to set your printer reduction to 54%.
When you get your first printout, hold it, and the board, against the light, then align the holes nearest the corners. If you're lucky they should be pretty much spot on. If not, try adjusting your printer setting just one per-cent plus/minus either side. It's a case of trial-and-error for each individual printer. But starting with 54% will put you in the right numbers.
Once you're armed with the correct setting, make a final printout on gloss photo paper. Roughly trim it to size, but leave approx 5mm around the perimeter. Paste the board and the printout with Pritt-Stick, then align / stick them together. Try not to get glue on the printed side, because it sticks like you-know-what and looks pug-ugly.
Trim off the excess overhang.

Pushing them skinny legs of resistors, etc, through plain-paper overlay is easy. Pushing them through gloss paper isn't. Jab a needle through there first.

Here's a photo showing a near-complete unit.
Prior to inserting the MOSFET, the three holes should be reamed a little larger. Refering to the pictorial at the top of the page you'll notice how each leg of the MOSFET is wider near the body. So making the holes larger first will allow you to sit the body nice and tight against the board.

A terminal pin was soldered in each of the servo input locations (not shown here). I thought it would make it easier when it came to soldering the servo connections. Easier, maybe. But not as neat as I'd like. Also not too impressed with how easy it could be to pull the wires loose. I reasoned I could counter the likelihood of that happening by sleeving the joins with heat-shrink tubing. Didn't have any. Instead, the wires were soldered directly to the board, then secured with a judicious dab of hot-melt glue.

The photo below shows the fully-assembled unit. It's now possible to see where the hot-melt has been applied. Although just a simple solution, it's quite sufficient in preventing the wires from working loose.

Note also how the MOSFET is sitting firmly on the board. Doing it this way makes for an altogether sturdier unit. And once the unit is mounted in the model it reduces the chance of component failure due to possible on-board vibration.

In order to fix the unit in the model 'twas decided to use double-sided tape -- the same stuff as carpet-fitters use. But first, the underside of the board has to be reasonably flat. And since each solder blob is anything but, they have to be filed flat first. Okay, okay - this is not the usual done deal, but with a little care quite good results can be achieved.

Take a fine-tooth file and gently remove all the 'high spots'. Then, with a damp cloth, remove any remaining swarf that gets clogged between the copper tracks. Now apply a piece of double-sided tape and trim any excess that may overhang the perimeter. Follow this with a layer of foam or rubber. In this case I used a thin'ish sheet of cardboard. Once in place and trimmed to size, a final layer of double-sided tape is added.