Homebrew-hints

Repurposing LEDs from commercial strips

I’ve used leaded LEDs with a series 1k resistor on boards to indicate DC power for a long time — for example, around T/R switching, or when switching bandpass or low pass filters in a multiband transceiver. After an electrician left a LED strip off-cut in a bin, I spotted an opportunity to reuse these high intensity surface mount LEDs as on-board indicators in my projects. They unsoldered easily. At 12V a 10k series resistor cuts the current draw down to a few mA whilst still emitting a satisfactory amount of light.

Repurposed LEDs indicate one of four control lines for a bandpass filter control overlay board, using a PCF8574.

Connecting boards and controls with 0.1″ headers

Getting signal and power lines on and off boards is typically dealt with using some kind of connector. I’ve standardised on 0.1″ sockets and headers, either soldered vertically or horizontally (as below). These are adequate for DC, control and RF paths (at HF), and allow the board to be physically disconnected or removed without un-soldering. It also lets you isolate a stage for testing, measurement or temporarily replacement.

DC and audio are not fussy, and can be brought up to a 0.1″ socket without much thought. If the DC current might exceed a few hundred milliamps or so, for example, to a QRP RF Power Amp stage, I might parallel two adjacent pins. Where an RF low-level signal is brought onto or off a board, a useful convention is to install a 3-pin socket with the two outer pins earthed and signal on the middle pin. The matching header is then reversible, which makes plugging in the coaxial flying lead literally a ‘no-brainer’.

Keeping plugin boards plugged in

Making your bandpass, low pass and crystal filters on small plugin boards, using 0.1″ headers and sockets, is a widely used technique. See , for example, the QRPLabs BPF and LPF kits and relay switching motherboards. In January 2019 I tested the idea to its limits after lugging my homebrew transceiver up and down SOTA summits over 50km in four days. Here’s what happened:

The 80m LPF (right) has worked its way out of its socket, and the 40m one (middle) is hanging by the electrical equivalent of a thread. The fix is simple — a small block of foam, shaped to fit the space between boards, while exerting gentle pressure on the boards.

Late changes to PCBs with custom overlays

Many homebrewers won’t start a project unless they are in possession of a professionally or machine-made PCB. Some use CAD and photographic techniques to fabricate PCBs. Others are using professional multi layered boards at ridiculously low prices out of China.

I still prefer the Old School method of working entirely by hand with pencil on paper, pen on copper, and ferric chloride. My reasons are that I enjoy re-living the high school drafting and art-room experience. Also, that it is immediate, I can design, draw, etch and finish a small board in an hour at home. I’ve found that with the right fine tip felt pen I can reliably draw lines and pads for SMD components, in fact, I now prefer it.

Another benefit is that if I change my mind during the evolution of a project, even after a board is built up, I can fabricate an ‘overlay board’ to support an alternate design, and simply solder the new layer on top of the old. Overlay boards like the ones pictured below have literally saved receiver and transceiver projects in the past, when I gave up on getting a stage to work satisfactorily. They have also allowed me to achieve a higher density when a compact transceiver build necessitates cramming it all into a tight space.

Left: irregular board for driver and IRF510, made for some free space on an exciter board, freed up by a change of mind. Right: board carrying PCF8574 and relay drivers, built to overlay a bandpass filter board.

To make an overlay board, I cut and trim a paper stencil to get the shape right. Then, cut the PCB to match, clean it up, and draw on the patterns of trace and ground plane. The overlay board is then built up and tested as usual. When debugged, it can be dropped onto its host board, and wires soldered to connect in signals, power and ground. On a small overlay board there’s no need for nuts and bolts, the link wires provide mechanical rigidity. A variant on this idea is to design the boards around 0.1: connectors (above).

Seeing a receiver’s audio bandwidth on your Smartphone

The best way to characterize a homebrew crystal filter or receiver bandwidth is with a spectrum analyzer. But most of us don’t have one lying around. There’s a lo-fi but effective alternative — any of the audio spectrum analyzer apps available in the app stores. Two to try are Spectroid and Frequensee. These apps provide a good approximation of the audio frequency bandwidth of your receiver. Hold the phone close to your receiver’s speaker. Remember that the spectrum you are seeing is a result of the IF bandwidth and any audio filtering after the detector (even simple designs often have some).

Spectroid plot of one of my homebrew rigs (Summit Prowler IV). The red line is the peak reading and gives a good indication of the radio’s audio bandwidth. Most of the energy in the top 5dB is between 300Hz and 2kHz. That confirms what I hear.

Mounting RF boards against angle aluminium partitions

Angle aluminium in standard sizes, for example, 40mm or 50mm, 1.5mm thick, makes a practical mechanically and electrically sound vertical partition and mounting facility inside an aluminium or steel chassis. If you employ a little bit of planning, you can design your PCBs to match your aluminium divider. So a 38mm x 150mm PCB sits vertically snugly against a section of 40mm angle, and the whole assembly can be easily bolted vertically onto the project chassis.

If the PCB is single sided, it can be lightly bolted flat against the angle, allowing the M2.5 or M3 bolts and nuts to secure a quality ground connection. If thru-hole or double-sided, the PCB may be spaced off the angle with steel or brass spacers in the usual way. If the board hosts trimpots or trimmers, you will need to design the board to place these on the outside or top edge. If you choose your components with a vertical board in mind, it’s usually not a problem.

I’ve used this technique successfully on a number of transceiver projects and found it to be practical and effective. It uses much less space than laying down the boards horizontally against the chassis. Of course you lose some physical access, but I’ve not found this to be too big a compromise. It is much more compact, and the angle piece can be arranged to form an RF tight compartment, valuable for housing gain stages such as IF strips and for shielding transmitter audio and IF stages from the RF power amplifier.

Plug-in 50 ohm pads

50 ohm pi-attenuators (or pads) are often used at amplifier outputs and on mixer ports to establish desired signal levels . These can easily be built up on a tiny scrap of Veroboard soldered to the top of a 3 pin 0.1 inch header, to mate with a 3-socket header soldered to the board. This allows individual ‘pads’ to be plugged it at will, allowing any value of padding from 0dB (a short circuit) to whatever you wish. 3dB and 6dB pads are used most often. If you make them up with signal on either end and the center pin earthed, they are not polarised (ie. they are reversible).

3dB pad on a scrap of Veroboard.

10dB plug-in pad with wired 1/4 watt resistors.

A 160m Low Pass Filter mounted on a 3×2 pin header. This filter is on the output of a 1.8MHz VFO for a transmitter. Inductors are wound on T30-2 toroids. There are twice the number of capacitors than there should be, as in all cases two junk box values were paralleled to get the desired pF. The filter sweep is below.

Cramming a LPF onto a 3×2 DIL header piece is not ideal RF design. The filter was swept to determine its bandwidth characteristics. Does not appear to have suffered from this somewhat experimental construction technique.