Signal injector, is it worth the effort ?

[clay shooter]

I am thinking of constructing a signal injector for fault finding in the circuits of my radio projects.
Is it worth the effort or would I just be better to use my digital multi meter ?

[paulsherwin]

A simple one is very easy to build - you only need two junk box transistors and a few other bits. They were often featured in radio mags as beginners' projects.

That said, I rarely use mine.

[clay shooter]

That's what I was thinking, however, it would make an interesting small project for the winter evenings

[paulsherwin]

If you have construction experience you can knock one up in half an hour. The last one I built just had birdsnested components fitted inside an old TV remote control.

[ms660]

I built one back in the day but to be honest it soon gathered dust, I found a service type RF signal generator far more useful, audio out for audio, RF out (CW & Mod) for RF and IF work.

You can't do any proper RF/IF alignment with a multivib signal injector.

That said if you're mainly doing audio work then you will probably find it a useful addition at a cheap price.

Lawrence.

[avocollector]

I reckon it's worth the effort especially if they are transistor circuits (or IC) - also a lot of fun and if you get it right the harmonics can check out the IF and RF sections. it does allow you to go from transistor base to base starting with the output driver transistor and working back to the rf amp. When the signal vanishes, you've found the faulty stage (or the battery has dropped out of the injector or any one of 101 fun things that can happen LOL).

[David G4EBT]

The short answer is yes - it would be very worthwhile and will do things that a multi-meter clearly can't do.

Circuits for injectors abound on internet, but ideally, you need a tracer too unless you already have one. For me, a signal injector/tracer is second ‘must have’ item on the list after a multi-meter - it speeds up the process of fault tracing no end. I can’t recommend too highly the Velleman K7000 combined injector/tracer kits which costs under a tenner, and often crops on the forum, as at this link, where there’s an example at post #4 of one that I built:

https://www.vintage-radio.net/forum/...Velleman+K7000

The kit can be found at several suppliers, notably ESR Electronics who stock the full range of Velleman kits. The kit consists of a PCB and all the board mounted components, including the volume controls for the tracer and injector, mounted directly on the PCB. In addition, you need a small speaker, switch, LED (if desired) sockets, and a box of some sort in which to house it:


https://www.esr.co.uk/velleman/k7000.htm

We all have our own approach to fault tracing is non-working radios.

My approach ever since my teens sixty years ago after looking for obvious faults such as loose connections, broken wires, doing voltage checks, continuity tests on speakers, OPTs etc, is to use the signal tracer/injector approach to test the audio stage to see if the fault lies there or in the earlier stages. (As an aside, from long experience of dead audio stages in transistor radios, the culprit is often an open circuit earphone socket. It's only supposed to silence the speaker when the earphone is plugged in, but the spring contacts weaken over time and often go open circuit even with the earphone plug removed).

By injecting an AF signal of about 1kHz (actual frequency unimportant) at the slider of the volume control, if there's a loud tone from the speaker, that shows that at least the audio stage is working. But suppose that the AF stage isn’t working - what then? If the RF/IF/detector stages are working, a probe from a signal tracer (basically, just a small amplifier), placed on the slider of the volume control will confirm that those stages are working. If so, tunable signals should be heard on the speaker of the tracer.

But what if there are no signals at the volume control slider?

By applying an RF probe (a diode detector attached to the tracer input) and working towards the front end of the set, the faulty stage can be found. EG, if there are no signals getting to the detector valve (or diode in a transistor radio) the second IF stage may be faulty, but if a signal from the injector is squirted into the grid of the IF valve, and the tracer probe applied to the detector, and a loud tone is heard, that suggests that the second IF stage is probably working and the fault lies closer to the front end of the radio.

To give an actual example of how a signal tracer/injector can save time, consider a dead Bush TR82 variant with AF117 transistors in the RF/IF stages. I dare say that due to ‘tin whisker’ folklore, many will tend to home in on those as being the culprit in a dead set and start poking around there. However, in my experience, just as likely to be at fault is the OA90 detector diode which is hidden from view in the can of IFT3. Apply a tracer probe to the input (pin 6 IFT3) then the output of the diode on pin 5. If you hear signals at the input and not the output of the diode, that means that the RF/IF stages are working, but the diode is dud. No need to dismantle IFT3 – just leave the dead diode in its ‘coffin’ and solder a new one across pins 5&6 of the IFT, and no need to muck around with the AF117s.

If the tracer probe is applied to the grid of the first IF valve of a set and tunable signals are heard, that proves that the mixer/oscillator (frequency changer) is working. However, if no signals are heard, that stage is faulty. If the signal injector probe is applied to the grid of the first IF valve (or base of the first IF transistor in a transistor radio) and the tracer probe is placed on the slider of the volume control, and a loud tone is heard, that's a good indication that the IF stages are working, thus confirming that the fault lies in the front end. It could be the valve not oscillating due to the valve itself or the associated circuitry - open circuit coils, faulty wave-change switch connections etc.

If you place the RF probe at the grid of the hexode section of the frequency changer valve, you are basically turning the signal tracer into the equivalent of a crystal set by connecting the probe just to the tuned circuits of the radio, and if you don't hear any signals, there must be an open circuit in the coils, such as a frame aerial, aerial coils, ferrite rod or whatever. In fact if you just connect the RF probe of the tracer to an aerial and earth with no radio attached to it at all, you'll hear several stations all on top of one another, just as you would with a crystal set!

It may take a little time to get the hang of adopting this technique to anyone not already familiar with it, but once adopted it saves hours of head-scratching, peering into rats nests of wires and components, wondering just where that elusive fault might (or might not) be.

A bit of a ramble, but I hope it's helpful.

[clay shooter]

Well thank you for your indepth explanation, it is appreciated. I am now inspired to build /buy one.
I have considered building one before, then though would it be just accumulating dust at the back of the "bench" with my capacitance /inductance meter, the last " must have" that I bought and have never used !

[Sideband]

Originally Posted by clay shooter

then though would it be just accumulating dust at the back of the "bench" with my capacitance /inductance meter, the last " must have" that I bought and have never used !

Interesting. My L C R meter is one of the most useful instruments I have. Not used daily or even monthly but certainly several times a year.

[TonyDuell]

Apart from a multimeter, which I think most people would agree is essential for electronic design, construction and repair (and which is used almost all the time), I feel that what other test gear you need depends on what you are working one, and the way you like to trace faults.

An example of the first point is that I don't use a 'scope that often (although I am glad I have several), but a logic analyser gets almost daily use. I suspect for most here it's the reverse, if you even use a logic analyser at all.

An example of the second point is that although I have 4 valve testers, I rarely use them. I prefer to check the stage containing the valve (say an IF amplifier) as a whole, looking at imput and output signals, valve operating voltages, etc. I can normally tell if the fault is due to the valve (e.g. a ridiculously low anode current) or if said valve is working well enough for that circuit, even if it doesn't meet the original specification.

I also divide test gear into general purpose and special purpose. The former are things like the multimeter, 'scope, logic analyser. The latter are things (looking round the room like a transistor curve tracer, CRT tester, valve tester, telephone line simulator, telephone test set, telegraph distortion measuring set, etc. Special purpose test gear can be great fun, and very useful if you happen to be working on that sort of device, but useless for other applications and you can often manage without it anyway. OK, I collect odd test gear, but if you are more normal, you should probably go for general purpose instruments at least at the start.

A signal injector/tracer is general purpose. It's also easy to use with minimal setting up (unlike say, the many settings on a 'scope). It won't find all faults (nothing will!), but it will sort out a lot of them. It's useful for quicking finding out which area of a radio (or whatever) is faulty. If you can get a signal from the input of the second IF amplifier through to the speaker, but not from the input of the first IF amplifier to the speaker then it a good idea to check said first IF amplifier. I think you should get one (heck, you have just convinced me to get that Velleman kit, for all I don't work on radios that much!)

[TonyDuell]

Actually, having looked at the Velleman one, I am not convinced it is ideal. It appears to be audio only. The ones I am used to have a square wave source, rich in harmonics, so even though the nominal frequency is in the audio band, they can be used to inject an IF signal. And the tracer has an (optional) diode detector on the input so you can use it to listen to AM modulated IF and RF signals.

To inject a signal I merrily poke a screwdriver at a connexion, you get crackles from (almost) DC to (again almost) light. If I was limited to just one bit of test gear it would be a 'scope (coupled with brains and a few components it will measure anything). After that it would be a multimeter with a proper scale.

Before any of those a small set of tools, soldering iron, cutters (get a good pair), small pliers and a selection of screwdrivers. I would put a signal injector at the lower end of my list, fun to make and use I agree, but hardly essential.

[Radio Wrangler]

I built a little signal injector to a magazine design (in one of those aluminium tubes that contained a yard or two of cored solder) almost 50 years ago, and in all that time I never used it after trying it out, then it was lost and never replaced.... I never felt the need.

Different people use different things. Screwdriver crackles are very broadband!

David

[paulsherwin]

If I were the OP I would just build one. It will be a useful exercise, and may be very handy, depending on his method of working. If it doesn't see much use then there's no great loss.

Here is a typical circuit. The transistor types and component values aren't at all critical. Personally I wouldn't use that 10uF electrolytic to couple the output if it's going to be used with valve gear - a 0.1uF 400V plastic film cap would be a much better idea.

[G0HZU_JMR]

When I was a student one of my lecturers used to give us (often just me) faultfinding tests once a week or so. This would be after a lecture covering some circuit theory on the device he would later put a fault on. This was on a variety of RF based gear and I had to do it using just an AVO8, a homebrew diode based RF probe (that connected to the AVO8) and also a signal tracer.
So I have fond memories of using a signal tracer. However, I think that a well constructed wideband RF noise source is also worth making and I use a noise source quite often for all kinds of design/dev work.

[WaveyDipole]

I seem to recall in my younger days that I built a injector/tracer from a circuit in Everyday Electronics. Since I couldn't afford an oscilloscope, next to a multimeter this was a useful tool. I seem to recall that the 'probe' was made from a felt pen barrel using a fine nail as the point. I also recall it had a diode that could be switched in for 'RF mode' and would allow tracing back along the IF up to a point. I believe it was based around a 741 op amp. Im intrigued G0HZU by how you might have used a diode based RF probe with an AVO?

To answer the op, I think it is worth building in addition to owning a multimeter even for educational purposes. The Vellman kit mentioned by David looks ideal.

[G0HZU_JMR]

The RF probe I used back then was just a classic pair of signal diodes arranged as an envelope detector with a couple of capacitors. The AVO was set to the 10V DC range and the probe was used to sniff for RF signals. It effectively rectifies an RF voltage to a DC voltage but not very efficiently. It was mainly used to look for local oscillator signals in a receiver or transmitter or for the various signals within a frequency synthesiser. It couldn't measure very small signals but it could easily detect something like a local oscillator signal even up at VHF.

[G0HZU_JMR]

I dug out a 'modern' pencil version that I made quite a few years ago and this one is designed for use with a high impedance DMM on the DC V range. It can detect down to about 100mVpkpk with a modern DMM and you can see I've written the calibration data on it in terms of detector volts vs RF pkpk volts. There's another scale on the back that goes up to 5V pkpk. It uses a pair of modern schottky diodes in a SOT23 package and it works quite well up into the UHF region if I use the little ground spike on the tip rather than the croc lead to ground it. You can see that the probe has both options.

You can see it is just made from a skinny piece of FR4 PCB material and cut into the shape of a pencil probe. It is surprisingly easy to use like this.

However, I don't use this probe anymore as I've got a couple of modern RF probes from Marconi that cover up to 200MHz and one goes to 1GHz and these can be used with a spectrum analyser. But the probe I used with the AVO 8 back in my student days would have used germanium diodes and it was just an ugly built thing on a tiny square of veroboard. I think I've still got my RF detector from my student days 'somewhere' as I made a copy of the one the lecturer provided for me. But I haven't seen it for ages.