A Little "Pantry" Transmitter for the IC-allergic

[mictester]

A couple of years ago, I put a simple MW transmitter circuit up for people who wanted something cheap and simple, that doesn't use exotic components.

Recently, I was asked for a PCB layout for it - several years ago, when I first built it, I made a few PCBs by hand.....

Here's the little board I made. Remember - none of the parts are really critical. I used 2N3904 transistors throughout, only because I had a huge box of them (about 6000 of the little blighters!), but you can use almost any small signal NPN silicon transistor .

There are three attachments - circuit diagram, tracks (it's single-sided), and component layout:







The supply is labelled as 12V, but I've used an unregulated "12V" DC supply that actually reads about 14.6V on load, without any issues. I added a 470µ electrolytic across the supply (it's not on the original circuit diagram). You'll also see that there's an unfilled hole in the PCB labelled "TP" - this is the test point for setting the modulator bias - adjust the preset for ½Vsupply at the test point. I put a "Veropin" there for testing.

I built the originals in tobacco tins, wired point-to-point, but I did make a few using the layout above. My "aerial" is about 6ft of dangly wire, but the length isn't too critical. I can receive this little transmitter throughout my house and garden (with just one exception - next to the boiler in the garage), and the audio quality is fine. I've added a simple compressor to the audio input of mine - if anyone wants the circuit, I'll put it up here.

Have fun!

[kalee20]

It looks not unlike something I built 12 years ago (still in use!). I can imagine it working well.

One tip, go round with a portable SW receiver, harmonics get radiated much more efficiently from a short wire, even allowing for the pi filter at the output. You really don't want to attract attention...

[mictester]

Experience shows that there are no harmonic products that can be received more than about ½m from the aerial! I think that the crude lowpass filter at the output works well enough.

[mictester]

Further circuit:

This is the simple audio processor that I used for my Mini MW Transmitter. I used a TL074 quad op-amp, but you could use a TL071 and a TL072. Using the quad, there's one redundant op-amp in the package. You could use it for an "active mid-rail" supply, but I find a pair of resistor as drawn is plenty good enough. Sorry that I used op-amps, but this allowed the use of a minimum number of parts. To achieve the same result with just transistors would be rather complex!

The first bit combines left and right stereo audio to mono, then applies a little passive "top lift" to "brighten" the sound. This is followed by an active filter at ~6kHz, to prevent interference to adjacent channels, and to help "concentrate" the audio you want. The last two op-amps and the FET form a simple compressor. This allows the modulation to be consistent whilst preventing overmodulation. You'll find that this helps your tiny signal to be audible above all the background noise on the Medium Waveband.



As usual, none of the components are particularly critical, but you'll notice that I've put two 1n5 capacitors in series - this could be a 750pF capacitor (if you can find one). The electrolytic timing capacitor in the compressor section is the right way around - the rectified audio fed to the FET gate is negative-going to vary the resistance shown across the FET. The capacitor and resistor from gate to drain of the FET helps to reduce the second harmonic distortion that a FET normally introduces. Shortly, I will put up a PCB layout that I used, and a Veroboard version too. One small change that some people might want to try, is to increase the 47k resistor in the time constant circuit (in parallel with the 10µ capacitor) in the compressor (try 82k or even 100k) if you think that the compressor is acting too rapidly.

[kalee20]

First, good that there's nothing receivable on harmonics, you can use it and sleep easy

Next, that's a nice audio processor! A couple of queries:

1) Why the 6kHz top-cut? You're not going to be worried about audio spilling into adjacent channels, because the only person in the reception area is yourself (unless you've built several of your transmitter, operating close together in frequency) and you might as well get the best audio frequency response from your whole system.

2) The feedback resistor around the 2N3819 JFET - why not 470k same as the series gate resistor? Then you'd have an audio null point halfway along the conduction channel, which would (assuming the FET is symmetrical) give you the best signal-strength/distortion performance.

The whole thing is simple and elegant. I like it!

[Hybrid tellies]

Of all the pantry transmitter designs I have seen I do like the look of this circuit.

I shall probably use the NPN BF*** transistors and also use the audio processor circuit to improve sound quality with that 6Khz filter which also reduce side band splatter on adjacent channels.

[Hybrid tellies]

Just two points
As it is with receivers I would imagine it would be a good idea to use a linear 12v PSU rather than a noisy and nasty SMPS wall wart thing.

Second point, In view of the Long Wave station closures could the capacitors around the oscillator be changed in value to make it work across the Long Wave band? I would think there might be some changes required on the output feed to the aerial.

[David G4EBT]

A very neat design.

Those among us who have spent their careers as professional electronics developers certainly have the edge!

Given that for maximum portability, most small pantry transmitters tend to be powered by a 9V PP3 rather than from a wall wart, presumably with slight changes in value to the resistors in the modulator, it would operate on 9V? (The RF part of the circuit operates on 5V via the 78LO5, which would be fine with a PP3).

I also wondered why a ferrite toroid for the 47uH inductor.

Just curious - it's not a stumbling block; £2.99 post free from a UK supplier:

https://www.ebay.co.uk/itm/224371157...Bk9SR6aYsv7oYQ

(I note that the OP hasn't been on the forum since March 1st).

[Chris55000]

Does anybody recognise what the OP did the layout diagram with?

I've been wanting a tool to do that sort of layout diagram with colour–coded resistors for years and can't seem to find one!

Is the artwork diagram full size? – I can use it in my Sprint Layout Software to make the Gerber Files if it is!

If the OP reads this, I'd love a matching artwork/layout diagram for the Audio Compressor Module!

Chris Williams

[wave solder]

It looks a very interesting circuit. However it DOES have an IC in it, a 7805. 😁

[TonyDuell]

I wonder if you could replace that by a 5v6 zener and resistor together with another 2N3904 as an emitter follower.

[David G4EBT]

Originally Posted by Chris55000

Is the artwork diagram full size? – I can use it in my Sprint Layout Software to make the Gerber Files if it is!

If the OP reads this, I'd love a matching artwork/layout diagram for the Audio Compressor Module!

Chris Williams

Having printed the component overlay and PCB mask as it is, it prints out (in Photoshop) at 10cms wide x 9cms tall.

On the assumption that the pre-set tuning capacitor C1, and the pre-set pot VR1 are standard 5mm pitch items, and that likewise, the disc ceramic caps such as C3, C4 etc are also 5mm pitch, the PCB would need to be reduced in size to 6.5cms wide x 6cms tall. The pitch of the resistor holes would be 8mm, and the body of a standard 1/4Watt 1% metal film resistor is 6.3mm long, so when the wires are bent at 90º they would fit the 8mm pitch.

I've re-sized the image to 6.5 x 6cms, printed if off and placed the above components on the PCB, to verify that they match the pitch, but don't anyone take my word for it, do check for yourself. My only reservation were I to etch a DIY PCB would be that when reduced to that size, the PCB pads would be only 2mm diameter and from experience, anything less than 3.75mm diam presents a challenge when drilling holes, but since most constructors these days, both on the grounds of minimum cost and maximum convenience, source their PCBs from China, that ceases to be an issue.

It would of course be much better if the OP were put in an appearance, when he might be encouraged to note the interest in his neat design, given that his thread has now had more than 1,000 views. Pantry transmitters have always excited the interest of forum members, as evidenced by the 'mini-mod', which is to pantry transmitters, as vacuum flasks are to Thermos, and vacuum cleaners are to 'Hoover'.

[ThePillenwerfer]

The forum messes up the size of images unless they are put in ZIP.

I realised the size was out and re-sized it to 5” x 4⅝”. Like that the holes for the IC and transistors are 0.1” apart. The outer ends of VR1's track are 0.2” apart and the wiper 0.4” from those, which matched the 'skeleton' type I usually use. The beauty of discrete components is that legs can be bent to fit where holes are either way.

A version at that size is attached.

[ThePillenwerfer]

Sorry everybody — please disregard my previous post. The board should be 2 1/2” x 2 5/16”. After all these years I should know Paint Shop Pro's Hefferlump Traps.

Really correct-size version attached.

Hopefully a helpful moderator can amend my above post.