Desalination water costs
are wrong

The costs you quoted in the March editorial for WA’s desalinated and piped water were surely not as shown? They correspond to about 50c and $1.50 per litre! This is as high as the bottled water racket charges on our supermarket shelves.

On the subject of loud DVD background music, my thanks to Graham Johnston (Mailbag, March 2005). I do not use an amplifier but I find that by connecting the centre audio channel to one of the TV speakers and making no connection to the other, nearly all the music is suppressed and the dialog is very clear.

Robin Stokes,
Armidale, NSW.

Comment: three people proof-read that March editorial – they must be blind! The figures should have related to a kilolitre of water!

Disagreement on
current transformers

David Millist wrote regarding current transformers (March 2005, page 6) and I wish to disagree with your comment on this matter. Referring to open circuit secondary voltages being dangerous, I enclose a copy of a relevant page from a text book used in gaining the Electrical Engineering Certificate from NSW TAFE during the 1960s, which included the Electrical Measurements option. It does indicate some danger to persons and equipment and the reasons behind it.

With regard to the one conductor through the core being considered as a full turn, many students question this but the lecturers were adamant that it was correct and that it definitely could not be treated as a "half turn". Unfortunately, age has blunted the memory and I am unable to elucidate the theory. Nor can I find any clear reference in the several texts which I have retained covering the subject.

In my work with current transformers used in test equipment, it was usual for there to be a link or simple switch provided on the transformer to facilitate short circuiting the secondary and the work instructions were quite clear "DON’T FORGET TO SHORT BEFORE OPENING THE SECONDARY CIRCUIT".

Peter Grout,
Forestville, NSW.

Current transformers
generate high voltage

I must agree with David Millist – current transformers (Instrument Type at least) must never have an open secondary as they generate high voltages.

This was one of the first things I was taught as an apprentice in the electrical generation/distribution industry some 35 years ago.

I have checked various reference books, such as Newnes Electronics Engineer’s Pocket Book 1996, and page 16 carries that very warning! It also explains why!

It is a shame your "comment author" did not check the facts before commenting!

Another electrical hand book clearly states: "The secondary must be short-circuited if the ammeter is removed for servicing". What a shame they have made such a basic uninformed mistake!

Michael Abrams,
Capalaba, Qld.

Current transformers
can be lethal

I read the letter from David Millist on current transformers and your reply. I am inclined to agree with Mr Millist. The secondary voltage of any transformer is (if I’ve got it right) actually proportional to the time rate of change of the flux – ie, d(f)/dt – which most of the time is proportional to the primary voltage.

The output voltage of a current transformer loaded with a resistor is proportional to the product of the secondary current and the load resistance, so in the extreme case when the load resistance becomes infinite, one might expect the secondary voltage to become infinite as well. Of course, this won’t happen in the real world as other effects come into play.

Furthermore, the burden of a current transformer will increase if it is not terminated, so the primary voltage will increase in such circumstances, and less voltage will be dropped across the load whose current is being measured.

So high voltages can be generated at the secondary terminals of a current transformer if it is not terminated with a low enough impedance. Whether they are lethal will depend on the particular circumstances. I reckon that in many cases the shock might be unpleasant at least. The source impedance of a current transformer (essentially the secondary winding impedance) is pretty low, so one would expect that they could deliver currents that might be lethal.

The issue is whether enough voltage can be sustained long enough for a lethal current to be delivered. Lethality of a current depends on its magnitude and the length of the exposure.

On the issue of full or half turns in a toroid, David Millist is quite correct. It is not possible to use a genuine toroid core with anything but an integer number of turns. The wire through the centre might appear as half a turn but the turn is completed in the outside world. Non integer turns are only possible in cores with more than one hole (such as the E-I form common in mains power transformers). Two holes can give half turns, three holes could give a third of a turn, etc.

Phil Denniss,
University of Sydney.

Comment: ahem, hmm (gulp – large slice of humble pie required to be eaten here). OK, since people have been so adamant about this, we had to go and do some actual tests. Never mind theory – what happens in practice?

We did a number of tests with toroidal power transformers to simulate the effect of a straight wire carrying 1A through the centre of the core. These tests were quite unequivocal – a straight wire is exactly equivalent to full turn, not a half turn. We repeated the tests with a clamp meter and again, the result was exactly the same. So yes, we were wrong!

On the question of current transformers though, we plead not (so) guilty. Our answer is still essentially correct. It appears that readers have assumed that we are talking about current transformers used in power distribution systems whereby very high currents are monitored by a moving iron ammeter in the secondary of the current transformer.

Certainly, these transformers can be very dangerous if the ammeter is disconnected while current flows in the primary. In fact, the voltages can be high enough to destroy the transformer itself. In really big current transformers, there is even the risk of a violent explosion if the ammeter connection in the secondary is inadvertently broken.

But in our reply to David Millist, we did not think about such large current transformers. After all, not many electronic enthusiasts have access to such beasts! Our original answer relates to a letter on page 98 of the January 2005 issue and is about a flea-power transformer to drive a high impedance multimeter. Nor did we suggest the specific figure of 1000 turns, just many turns.

However, as part of our tests, we measured the voltage from the 240VAC primary of a 160VA transformer which actually had 1024 turns. With the same 1A wire through the centre of the toroid and using the 240VAC primary as the secondary, the output voltage was 18.5VAC into a 10MW load – ie, a digital multimeter – and 6.2VAC when loaded with a 10kW resistor.

From that, if the wire carried 10A, you could expect 185VAC into 10MW and 62VAC into a 10kW load. So yes, the voltage into an open circuit is much higher, as we would expect, but again even this is a much bigger current transformer than our original answer suggested.

In hindsight, we should have been more specific in our original answer. If we had suggested, say, 50 or 100 turns on the secondary, it would have been quite sufficient to drive a high impedance (10MW) digital multimeter.

Current transformers
operate in a different mode

In your reply to the letter on Current Transformers in the March 2005 issue, you state that you cannot "see how a current transformer can generate a dangerous voltage unless it has a significant voltage across its primary". The problem with current transformers is that they operate in a different mode to the basic voltage transformer that most people are used to – and the expected transformer ratio rules simply don’t apply!

For a current transformer, the core flux depends on the primary current (as per a voltage transformer) and also on the flux, generated by the secondary current that, according to Lenz’s Law, cancels out most of the primary flux, leaving the core with a low level of total flux, and a low generated EMF in the secondary winding. (CTs typically run with low core flux levels compared to voltage transformers, and so generate no more than 5 -15V at maximum primary and secondary current.) CTs with no load have a primary flux but do not have an opposing secondary flux, so the basic voltage transformer operation, as well as the normal transformer rules cannot be applied to them.

This occurs because, if you open the secondary and power up the CT primary, the primary current will drive the core flux to quite high values, as there is no secondary flux to oppose and reduce the primary flux, and these high flux levels in the core generate high EMFs in the secondary winding, which may have hundreds of turns in larger (commercial) CTs. In many instances, low ratio CTs – 50:5 (ie, 50A primary to 5A secondary), or 100:5 – can yield over 100V across the secondary terminals and in the case of high ratio CTs (2000:5 or 5000:5), several thousand volts are common. This can cause flashovers and in the case of oil-filled CTs used by the supply authorities, has led to explosions and fires on occasions.

Although the CT in the project mentioned is not capable of these effects, it may generate voltages that do exceed Extra-Low Voltage (32V AC) and so may present a hazard to electronic devices to which they connect or they may "bite" an incautious service person hard enough to cause dropped tools, with possible short-circuit problems and naughty sailor-type words from the lips of the startled technician.

So David Millist’s statement that CTs can and will generate lethal voltages with an open circuit secondary is quite correct. The "lethal" applies mainly to commercial CTs as you know but you cannot apply the "logical" voltage transformation rules to a CT to understand the reason for those high voltages!

Secondly, a "half-turn" through a CT core is effectively a single turn – if you draw a sketch of a single turn passing through a toroid, all of that turn’s flux expands out and enters the core, and generates an EMF. The permeability of the core, with no air gaps, gives up to 2000 times the flux in the core that a conductor surrounded by air would generate, so any conductor inside the core is very effective. But if you place the same conductor outside the core and sketch its flux path, the flux may enter the core on the side nearest the core but it must leave the core before passing all the way round and has to expand away from the core, through air, on the side remote from the core.

That means there’s a very large air gap in the magnetic path outside the core, which gives a low relative permeability (typically slightly more than unity) for the flux path and hence low flux generated in the turn outside the core. That, in turn, means that the side of the turn outside the core has almost no effect in generating voltages in a core with a "closed" flux path!

Only an E-I core or double C-core will use both sides of a turn and hence have "half-turn" capability. Toroids or single C-cores only use the part of the turn inside the core to generate useful flux, generally, so a "half-turn" counts as one turn, really!

Of course, this doesn’t apply to air-core inductors (RF transformers, etc) where the absence of the iron core means all core permeabilities are unity. Nor does it apply to RF and IF transformer "rod-type" ferrite cores. This is because this type of construction always has large air gaps around the coils, with the core inside the turns and no "easy" flux path. This means that all flux paths have about the same (low) permeability, and "partial" turns do have a proportional effect.

Brian Spencer,
Seaford, SA.

Comment: oh, well – see our previous comment.

Thanks from East Hills Girls

Just a short email to thank Ross Tester and the staff at SILICON CHIP for your excellent article on the PED-X project in the February 2005 Issue. The article has provided the greatest boost for both our students and the school community. It has show-cased what a small group of our students are capable of achieving with very limited resources, especially time. They were particularly proud to have their photo in print and also on the internet.

This year is also the first time that the school has run a year 9 electronics class. It is proving to be quite a success. We were surprised just how many girls are involved in electronics in one form or another, outside of school.

Steve Sharp, Head Teacher,
Information & Communications Technology, East Hills Girls
Technology High School, NSW.

Better method for setting amplifier quiescent current

I wish to comment about the setting of quiescent current (Iq) in audio amplifiers. I never set it up by the book – ie, adjust the trimpot for XmA or YmV across the emitter resistors. I have found that to get it right, you connect a dummy load to the output of the amplifier, connect your CRO across the load, connect your signal generator to the amplifier input and set it to say 30kHz.

With an output of full screen on the CRO with the attenuator set to 0.5V/DIV, probes on X1 setting (a total of about 4V p-p), you then adjust the Iq pot for no crossover distortion. With amplifiers that I know will run OK with more Iq, I give them say, an extra 10-15mA – it depends on the heatsink size and thermal feedback arrangements. Leaving them to soak for an hour or so and then giving them a power/cool down test usually lets you know if it’s set too high.

Many years ago, a Leak Delta 70 came across my bench. When repaired, set up as above and returned to its owner, I received a phone call from him. "Oh-oh what’s wrong?" was my initial reaction. As it turned out there was nothing wrong. He rang to say that he had had the unit repaired before and when he received it this time it "sounded" so much better. He had not really been happy with the sound since new but he was really pleased this time.

Brad Sheargold,
Collaroy, NSW.

Comment: we are not keen on your method for setting Iq. In our experience, at 30kHz, secondary crossover distortion becomes quite significant and usually is not greatly affected by the amount of quiescent current. In any case, trying to judge crossover distortion just by looking at a sinewave is very difficult.

Most good amplifiers will show no visible crossover distortion, even at high frequencies and low power. The only sure way to judge the setting is with the benefit of a harmonic distortion analyser which can display the distortion products on an oscilloscope.

That is not to say that some amplifiers might not benefit from a small increase in quiescent current. However we would not recommend increasing the quiescent current setting for any of our designs, unless the builder has access to the above test equipment.

Positive feedback
on audio projects

This is just some feedback about very positive results I have had from several SILICON CHIP projects. The first is the 175W power amplifier (SILICON CHIP, April 1996) built for my brother who is a musician. I matched it to a version of the JC-80 loudspeakers (October 2003). I used the same drivers and crossover but changed the shape (not the volume) of the speaker box.

In the setup, a balanced microphone signal is fed into a preamp, (SILICON CHIP, April 1995) and then fed to the amplifier via a 5-Band Graphics Equaliser, (SILICON CHIP, December 1995). The results are very pleasing, with plenty of power and excellent quality sound.

The other success is a Sub Bass Processor described in "Electronics Australia" (September 1999), feeding into an SC480 amplifier (SILICON CHIP, January & February 2003), driving the subwoofer described in March 2003. The Sub Bass Processor is connected to the speaker output lines of a Pioneer stereo hifi amplifier (about 30W RMS per channel). The results were very pleasing indeed, especially playing older tapes and LPs that could only supply a rather poor bass signal.

Joe Kelly, via email.

Ratting old headlamp globes

Thanks for that great Salvage It! article on pages 90 & 91 of the February 2005 issue. Headlight globes usually fail when the low-beam filament blows before the high-beam filament. This leaves us with a globe that can’t be used in a vehicle.

In the circuit on page 91, a resistor is used to limit the current to the maximum of the plugpack. By replacing the resistor with the high-beam filament of an old headlamp globe, we can utilise a globe that would normally be thrown out.

Not only does this recycle an old globe but it also acts as a constant current source, while still allowing the battery voltage to rise with time.

Let us assume that a high-beam filament is rated at 100W. Using the formula R = V x V divided by P, then the resistance of the hot filament would be about 1.44W. Assuming a 13.5V DC plugpack and that the battery is at about 11V when flat, there would only be a maximum of 2.5V across the globe, so it would never get hot.

The resistance of a headlight filament varies with current, thus acting as a constant current source while still charging the battery. In the case where a headlight globe was passing too much current, an 18W stoplight globe could be used.

Chris Potter, Kilsyth, Vic.

Cheap checker for
remote controls

Using the Remote Control Checker kit (January 2005) or a video camera (February 2005) to check if your remote control is working are both effective. If you can’t afford the above methods, try my (poor man’s) version: grab the nearest broadcast trannie, tune between stations, increase volume and place the remote control within a couple of centimetres of the end of its ferrite rod antenna and press a button.

Norman Ratcliffe,
Toorbul, Qld.

Comment: your method using an AM broadcast radio works extremely well. Thanks.