This is only a preview of the June 2022 issue of Silicon Chip. You can view 41 of the 112 pages in the full issue, including the advertisments. For full access, purchase the issue for $10.00 or subscribe for access to the latest issues. Articles in this series:
Items relevant to "Spectral Sound MIDI Synthesiser":
Items relevant to "Buck-Boost LED Driver":
Items relevant to "Arduino Programmable Load":
Items relevant to "500W Power Amplifier, Part 3":
Items relevant to "MOS Air Quality Sensors":
Items relevant to "Revised Battery Charge Controller":
Purchase a printed copy of this issue for $11.50. |
SERVICEMAN’S LOG
Ion with the wind
Dave Thompson
Servicing can be a strange industry. These days, much of what comes
through the door is not designed to be repaired. You can imagine how
that makes the job a bit of a challenge!
I understand that companies want to protect their designs.
Still, if someone wants to clone a product, unless it uses
cutting-edge technology, they can do it without too much
difficulty.
Making devices unrepairable usually has the most significant impact on the customer – someone that the company
making the goods probably should want to keep happy!
Someone, somewhere, always has the wherewithal,
resources and ability to ‘deconstruct’ or ‘reverse engineer’
something to find out how it works. If the mood or the
promise of commercial gain takes them, they will replicate it and sell it, likely at a lower price. Many countries’
economies are seemingly reliant on copying the ‘intellectual property’ of others.
Some of these ‘clone jobs’ are so shameless that they
replicate the external appearance of the original product,
down to the shape, the colours and even the font. They just
replace the original company’s name with their own and
sell it for a fraction of the price!
I think this is a basic human instinct, illustrated by the
fact that I (like many others who would be reading this
column) pulled many things apart when I was a wee fella
to see what made them tick. Dad had to put most of them
back together – that is, until I could do it myself.
My first guitar builds were attempts at making copies of
existing models as I tried to make an instrument I could
afford, hopefully as good (if not better) than those available
for what were, to me at the time, vast amounts of money.
Whatever the motivation behind copying others’ work,
it still happens a lot today.
That said, finding an epoxy resin-potted ‘module’ in a
commercial unit may have another more practical explanation than obfuscation.
Ionisers are positively great
Recently, a client brought in a device I’ve not seen for
many years; a commercially-produced air purifier/ioniser.
These devices were all the rage as far back as the 60s and
70s, in a jet-age, sci-fi sort of way, and were the ‘go-to’
gadget for a while. They were also popular as a project in
magazines back then.
It got me, too; the subject of purifying air using electronics fascinated me. The result was that I built many negative ion generators over the years, with varying success.
However, that didn’t mean all was well in the state of
negative ion generators. There have been hundreds of studies showing that negative ions have no real benefit to people or pets, while a similar number of studies have proven
that they are beneficial.
92
Silicon Chip
Some took them so seriously that hospitals utilised
them as part of their air-conditioning systems to minimise
or neutralise airborne infections. The recent pandemics
(SARS, COVID etc) have resulted in a considerable boost
in air ioniser sales.
As is typical these days, ‘mileage’ varies with any ‘health’
product. If using one makes one feel better, why not use
it? Regardless of the health and well-being implications,
making and setting them up is fun and educational. That
(for me) makes up for any of the controversies. Many commercial variants are still sold today for use in the health
and horticulture industries.
The premise of these devices is simple: apply a high
voltage to an array of sharp metal “emitters”, and a corona
or ion wind will stream from those points. Many airborne
pollutants are electrostatically charged by this wind and
are attracted to a nearby ground. So the theory of air purification by negative ions is sound.
I have seen this for myself; I built several ionisers in
the early 1980s for a friend who had a small greenhouse/
hydroponic setup for producing cabbages and cauliflowers.
This guy wanted to improve the air quality in his setup,
and when he heard me going on (as was my wont) about
this new-fangled method I’d been reading about, he was
keen to bankroll a couple to see how it worked for him.
I scaled up a project from an American magazine and set
them up in his greenhouse. They sat on a large baking-tray
type metal plate that was Earthed through the mains, and
sandwiched between that and the ioniser was a sheet of
white paper. After just two days, that paper was turning
grey, and when the ioniser was moved, there was a stark
white outline where it had been sitting.
That proof was good enough for me. Those ionisers ran
for the next 15 years until the guy moved, and I was sold
on the idea.
The last one I built was for a person here who suffers
from a seasonal complaint we call “Nor’ West Syndrome”.
We get a very hot, dry, gusty wind during the spring and
summer months, prevailing from the northwest. It is loaded
with pollens and dust picked up by roaring over the nearby
Canterbury Plains.
As it blasts through Christchurch, it dumps that pollen
in buildings, on the ground and anywhere the air reaches.
It looks like yellow, granular dust and is sometimes everywhere. Those with hay fever or any sensitivity to pollen
or dust can have serious health impacts due to this phenomenon.
When this person complained to me about it, I suggested
an ioniser might be the answer, especially if set up near
Australia's electronics magazine
siliconchip.com.au
Items Covered This Month
•
•
•
Ion with the wind
A nomadic TV antenna
Repairing a microswitch in a washing machine
Dave Thompson runs PC Anytime in Christchurch, NZ.
Website: www.pcanytime.co.nz
Email: dave<at>pcanytime.co.nz
Cartoonist – Louis Decrevel
Website: loueee.com
their bed. I initially built one ion generator for them, then
another for their home office; they swear by their use now.
Unrepairable but not unbreakable
The one that came into the workshop recently is a commercial unit, made in China, and is quite small compared
to others I’ve encountered. It runs from a 9V battery and
is designed to sit on a bedside table or similar.
After determining the battery was good, the only real
option was to pull it apart and have a look. There were no
screws; it was clipped together, so I got the client’s permission to (literally) crack it open.
Inside was a solid mass of black potting compound.
There was absolutely nothing I could do with it repairwise. Many such devices are potted like this inside, not
to mask the circuitry (although that might be a useful
byproduct) but more likely to prevent corona bleeding or
arcing between all the components that are in very close
proximity.
Most of these devices work the same way, though there
are different lines of thought regarding the power input.
Some apply a relatively low DC voltage (usually from a
plugpack supply), via a simple switching circuit, to the primary of a transformer. The stepped-up transformer output
then goes through a voltage multiplier circuit.
The resulting HV output terminates at the pin array,
though sometimes it can just be a single-pointed electrode/antenna.
The other method is to use mains power directly into the
multiplier circuit, sans transformer (although some use a
1:1 transformer for isolation). But I doubt that is legal here;
the whole circuit would run at mains potential; perhaps
another good reason to pot everything solid!
The output voltage (around -3kV to -10kV depending on
the circuit) is safe for anyone touching it (the output pins
may be accessible through holes in the case), yet it’s high
enough to cause an ion wind to stream from the points.
This wind can usually be felt by placing a (wet) finger close
to the emitter array, but note that some ionisers employ a
fan to fudge/boost the ion output.
You might think getting too close to the pins could be
dangerous; sometimes, a tiny, weak arc to a fingertip may
be visible in very low light. But a line of high-value resistors in series with the HV feed to the pin array limits the
available current to a safe level.
The idea is that the negative ions from the emitter pins
charge any muck in the air, and this finds its way to the
nearest ground. Most commercial units don’t come with a
ground plane to sit on, or a handy connection to add one,
siliconchip.com.au
so we can only assume the particles find something else
grounded enough to be attracted to.
Also, some of these commercial devices are tiny, about
the size of a small mobile phone, so it is hard to see how
they can be as effective as those boasting a decent-sized
pin array.
Either way, this device wasn’t running, and there was
nothing I could do. They are pretty cheap to purchase off
the internet, so I wasn’t sure this would be a feasible job
anyway. I talked to the client about it (good communication is essential to a serviceman). In the end, they commissioned me to make one of the models I’d produced before
(and likely harped on a bit too much about!).
Do it yourself!
While not overly complex in the models I build, the actual
electronics involved are pretty interesting from a theoretical point of view and are good fodder for the home experimenter. Anyone who likes watching big arcs and making
circuits with the potential (!) for lots of high-voltage experimentation can use a very similar layout to power up the
likes of Tesla coils and Jacob’s Ladders.
The main difference with this circuit is that we want to
keep arcs out of the equation as much as we can and feed
all that juicy HV energy to the pin array. If we get a flashover at any of the connections on the multiplier, not only
do we create a fire and shock hazard, we lose any semblance of decent ion emission at the output.
The circuit I use consists of three sections: a driver/
oscillator section, a multiplier/output section and the pin
array. Each section has its own circuit board.
Anyone familiar with such things would recognise a
Cockroft-Walton arrangement of high-voltage capacitors
and diodes in a characteristic criss-cross ladder outlay.
This is actually a line of half-wave rectifiers connected in
series, with each stage boosting the previous stage’s output ever higher.
Theoretically, you can just keep adding stages. As long
as your components can handle the resulting voltage (and
the spacing between components and stages
is enough to prevent arcing between them),
Australia's electronics magazine
June 2022 93
you can get some seriously high voltages out of such a setup.
Practically though, it’s a different story.
One significant consideration is what to assemble these
components on. I could use veroboard or perf-board, but
I’d spend most of my time stripping tracks and arc-proofing
it. A better option is to use a printed circuit board with the
correct spacing already worked out.
Fortunately, because I throw nothing away, I have a spare
set of boards from the run of ionisers I made back in the day.
These are very much home-made, meat and potatoes PCBs,
not like the multi-layer works of art we see today, but they
still look good and do the job, which is all I want from them.
The driver board boasts a board-mounted barrel jack
for DC input (typically from a 9-15V plugpack) powering
a 555 timer configured to run as an astable multivibrator
that, in turn, drives two NPN power transistors in a pushpull configuration.
The relatively square-wave output from the transistors
switches current through the primary of a custom-wound
transformer. A few other ancillary components ensure
everything runs as it should.
The secondary of the transformer runs off to the multiplier board. If this sounds familiar to some readers, that is
because the April 1981 issue of Electronics Today International (ETI) (which we offer scans for siliconchip.com.au/
Shop/17) featured a very similar circuit. This circuit is also
quite close to most oscillator-based ioniser circuits from
around that time and even some I see on the web today.
I changed it slightly to upgrade the transistors, included
heatsinking and added two extra stages to the multiplier,
altering the PCB artwork to suit.
The rest of the build involved making the transformer
and the pin array, both tasks which can put constructors
off. But while they both seem to be quite tricky to make,
neither is overly difficult.
The transformer uses a standard ferrite pot core and bobbin, and it is just a matter of winding the coils carefully and
94
Silicon Chip
neatly, insulating well between each layer. I made a jig years
ago for such things and keep it handy, just in case. While
the wire is still wound by hand, holding the assembly is
much easier, and the resulting windings are much neater.
This component is the most critical, especially when adding extra stages to the multiplier because the transformer
will break down internally – usually before anything else
– should the HV output go too high.
The pin array is a little trickier, but I used a 0.5mm PCB
drill bit and a small manual hand drill to carefully bore
the holes for modified (head cut off) dress pins through
a 120mm length of 4mm brass tube. Getting them all in
line is probably the worst part of it. Still, given that they
can be easily bent into shape if they are off-angle a little,
it isn’t too onerous.
Soldering the pins in is also a bit of a challenge, but I use
liquid flux and lots of solder, and it seems to stick them
in OK. I also fill the ends of the tube with solder and sand
that and each pin end smooth to ensure there are no sharp
edges – ions will ‘leak’ from anything sharp.
This is also why the capacitors and diodes in the multiplier must be soldered to the PCB with their leads cut very
close and what I would usually consider too much solder; a
nice round blob will be less likely to arc to another nearby
joint or bleed ions. We want all the ions coming from the
sharpest parts – the pins.
Once it is up and running and mounted in a nice case,
the question is how to know it is working. Other than sticking your hand in front of it and hoping to feel the corona
wind (no, not that kind of corona!), the ion output can be
measured using a simple detector.
That ETI version I mentioned included a tester that supposedly lit up when ions were present, but I could never
get it to work. That could be because the PCB I made for it
wasn’t good enough, the components were not quite right,
or perhaps the ioniser I built at the time didn’t work well
enough to trigger it.
Regardless, I ended up making a little hand-held meter
with a couple of transistors and a whip antenna that picked
up negative ion emissions quite well. I use that now – not
that I build many of these things – but when I do, it’s nice
to have the test gear to show they are working.
Australia's electronics magazine
siliconchip.com.au
The client was happy and claimed my version worked
way better than their original ever did. That’s always gratifying to hear, especially with something as ‘controversial’
regarding health benefits as air ionisers.
The nomadic TV antenna reception
G. C., of The Gap, Qld had a lot of trouble with TV reception in his caravan on a trip to Far North Queensland. He
traced it to a rather apparent electronic fault...
On the first night at Tewantin, we could not receive a single television channel. This was somewhat unexpected as
our caravan had been in storage for several months. Before
that, we had received television stations without difficulty
from the Mt Tinbeerwah transmitters near Tewantin. These
UHF transmitters broadcast with vertical polarisation.
Most modern caravans use fractal antennas that can
receive both horizontally and vertically polarised signals,
but our 16-year-old caravan does not have such luxuries.
It has a simple Winegard antenna (sometimes called a
batwing) with folded dipoles for VHF and UHF signals.
It can be raised off the caravan’s roof and rotated in the
direction of transmission by a mechanism on the ceiling.
Inside the antenna radome is a masthead amplifier that is
power-fed through the coaxial cable.
The two struts that hold and control the position of the
antenna head are in a parallelogram configuration. This
enables a simple modification to be carried out that involves
raising one of the struts, causing the antenna head to rotate
to the vertical position, which is needed at Tewantin.
Bypassing the PVR (personal video recorder) and connecting the antenna fly-lead directly to the TV did not fix the
problem. Neither did changing the fly-lead or connecting it
to the second antenna socket in the van. I checked that there
was 12V at the “F” connector at the antenna, and it was fine.
As a last resort, I removed the 694MHz filter (to block
mobile phone signals) on the drop-side of the power-feed
module and, surprisingly, I could then tune in all the channels. It seemed unlikely that a passive device had failed,
and I wondered at the time whether its insertion loss was
the straw that broke the camel’s back.
All went well until we arrived at Rolleston, where the
caravan park had a community antenna with coax distribution. Usually, we connect the audio output of the television
set to the auxiliary input of the caravan’s CD/radio stereo
system. However, with the external coax feed connected to
the van, mains hum from an apparent ground loop made
listening to the TV virtually impossible.
When we reverted to using the television speakers, there
was absolutely no sound. We had that problem on a previous trip, and we fixed it by doing a master reset on the
TV. But that did not do anything this time, and it wasn’t
until a couple of days later in Charters Towers that I had
time to dismantle the TV.
I was surprised to find that the voice-coils of both speakers were open-circuit. I couldn’t replace them at the time,
and when I later got home, I discovered that Sharp no longer sells replacement parts for this set.
Eventually, we arrived at Georgetown for a few days
exploring. Once again, we could not pick up the single-
channel broadcast from the transmitter just over 1km away.
This ABC transmitter was in the VHF band and transmitted at 4W. Even manually tuning the PVR and TV, I could
not find the ABC signal.
siliconchip.com.au
I decided it was time to check the amplifier in the antenna.
It was not easy to get to, as we did not bring a ladder. But by
lowering the antenna quite a bit, it was possible to access
the F-connector and pins securing the antenna to the struts
by sticking my head out of a hatch.
These antennas do not appear to be designed to be
repaired; umpteen diabolical plastic clips held the two
sides of the radome together, plus several plastic dowels from one side to the other, which were ultrasonically
welded. When I finally got to the printed circuit board, it
was obvious why it was so temperamental – it was severely
corroded, presumably due to moisture ingress.
I would generally clean a board in this state with isopropyl alcohol, but the best I had on board was plain old
methylated spirits. After cleaning it up, there were open-
circuit tracks that needed replacement. After re-assembling
the antenna and re-installing it, it was happy days again
when we could receive the ABC.
But our joy was short-lived, as after about 90 minutes,
there was a complete signal blackout again. The next day,
I risked life and limb to again remove the antenna to access
the PCB. I checked every plated-through hole with a multimeter and replaced another corroded link. Again, there
was apparent success, and the missus could watch her BBC
programs. All was well.
At Atherton, we could pick up the VHF broadcasts from
the powerful Mt Bellenden Ker transmitters but not the UHF
The masthead amplifier PCB and
components were badly corroded.
Australia's electronics magazine
June 2022 95
signals from the nearby transmitters at Hallorans Hill. On
passing through Mareeba, I was pleased to be able to buy
PCB lacquer at an electronics parts reseller and, when I
applied it, the board looked much better. But in hindsight,
it may have been better to have waited a bit.
At Redlynch, Cairns, there is a local broadcast site with
all channels in the UHF band, but we couldn’t pick them
up. When we arrived at Wongaling Beach just south of
Mission Beach, my suspicions that the antenna amp had
failed again were confirmed when we could not pick up
any UHF channels from the broadcast site on Dunk Island.
It was just 8km away, and we could see it clearly.
When I again removed the antenna and accessed the
PCB, I had a close look at the circuit. The signal split into
two paths from the balun: a VHF amplifier with one transistor and a UHF amplifier with two transistors. I quickly
sketched out the circuit of the UHF section.
At this stage, I got on to the internet to see if there was
any information about fixing these boards. There was information about replacing corroded PCB tracks and plated-
through holes, but someone had figured out that the UHF
transistors were BFR93A types.
I connected a 9V battery to the output coax connections
on the board so that I could make DC voltage measurements. It soon became evident that the second UHF transistor was not conducting at all, even though the base bias
voltage was correct. I removed this transistor, which wasn’t
easy in a caravan without a fine-tipped soldering iron etc.
The missus was the theatre nurse and held a magnifying glass and LED torch so I could see what I was doing,
which definitely helped.
No wonder we were having so many issues with the
antenna – the collector of this transistor was missing! It
had totally corroded away. Where do you buy a low-noise
RF transistor at Mission Beach? All I could do was bridge
the base track to the collector track and see how well the
UHF amplifier would work with just one transistor.
In practice, it worked surprisingly well, and for most
localities, either the PVR or TV signal meter displayed a
strength of around 70% and signal quality of 100%. There
was only one place when the antenna was pointing at dense
vegetation that the level of pixellation was so severe that
watching the television was untenable.
By the time we got home, an online retailer had delivered a few of the transistors as well as some SMD ceramic
capacitors. After soldering in a new transistor, I replaced
a filter capacitor that had a corroded end. I also strung an
MKT capacitor across the power supply electrolytic.
When re-assembling the radome, I siliconed both halves
together except for segments to allow egress of any water
entering when the antenna is in either the horizontal or
vertical plane. The antenna is now working as well as it
can, but the real question is: for how long?
service manual and parts manual from the Fisher & Paykel
website and printed them. Fisher & Paykel must be only one
of the few companies left that gives out service manuals.
The service manual shows how to enter diagnostic mode,
which gives codes for the last eight cycle errors. It also has
procedures on how to test the out-of-balance microswitch,
the pump and the water valves. The error codes are displayed on the eight wash progress LEDs, with the right-most
“spin” LED being the least significant bit and the left-most
“long wash” LED being the most significant bit.
The code that it came up with was 00101011 binary or
53 octal. The binary decoding table indicates that 3 octal
selects the 4th column while 5 octal selects the 6th row
down. This points to error code 43, which means that the
fault is that the out-of-balance microswitch is permanently
on or the harness to it is disconnected.
Since I made my repair, the Fisher & Paykel website
now has a service diagnostics manual that makes it easier
to read the error codes.
I activated the diagnostic mode to test the out-of-balance
switch. On manually activating the out-of-balance lever
under the top deck, the short wash LED did not turn on,
which indicated that the microswitch was not working.
To get to the microswitch, I removed two screws at the
top of the back panel. I could then lift up the top section
with the eight wash progress LEDs and remove a screw
that holds the grey control module. I then lifted the control module to reveal the microswitch, an SPST-NC type.
The normally open (NO) contact that would have made it
an SPDT switch had been cut off.
I guess that they used an SPST-NC switch to ensure that,
during assembly, the quick connector could not be placed
onto the wrong terminal. After disconnecting the quick
connectors, I used a multimeter to measure the resistance
of the normally-closed contacts. This showed that the contacts were open and that the microswitch was faulty.
Opening the microswitch revealed that the contacts had
become severely oxidised after many years of service in an
environment with water and steam. I made a trip to our
local electronics store to purchase an SPDT microswitch,
and having installed that, the machine worked again.
Some time later, after we moved to a new house, it failed
with error code 43 again. I knew what to do, so I purchased
Washing machine microswitch repair
R. W., of Mount Eliza, Vic has discovered that sealed
components are needed for harsh environments, including
the inside of a washing machine. At least he’s had plenty
of practice replacing the failing part. He can probably do
it in his sleep by now...
I repaired a Fisher & Paykel GW709AU washing machine
that was around 17 years old. The symptom was that it
would not start. I began by downloading copies of the
96
Silicon Chip
An internal view of the GW709AU out-of-balance
microswitch assembly. Replacement microswitches are not
available from Fisher & Paykel.
Australia's electronics magazine
siliconchip.com.au
another microswitch and installed it. Due to a busy lifestyle, I did not realise that it was only around two years
since I had first replaced it.
Well, two years on and the washing machine stopped
during the spin cycle. This time, the “final spin” LED and
“current spin speed” LED were both flashing. On restarting
the washing cycle, it would start and run then stop with
the same error code. Section 10.5 of the service manual
indicated that the fault could be any one of eight causes,
mostly related to the machine being out of balance.
One of these mentioned, “Check that the switch operates correctly and the contacts measure less than 2 ohms”.
I made the mistake of thinking that, as it was not that
long ago that the microswitch was replaced, it was unlikely
to be the problem. I also thought this because the washing
machine would start and then run before stopping.
I started a cycle and lifted up the lid a bit to see what
was happening without activating the lid switch. I could
see that the out-of-balance lever was not being activated,
but the washing machine was still stopping. This indicated
that out of the eight possible causes, it could only be the
microswitch at fault.
I measured the normally-closed contacts and got a reading of 8MW, not less than 2W as specified. It appears that
the fault was intermittent. Sometimes the switch would be
closed but then incorrectly open during a spin cycle. So
the microswitch was faulty once again.
Opening up the microswitch, it looked OK with no apparent oxidation. So what was wrong? An internet search
did not reveal a data sheet for this device. I thought this
type of switch might be OK when switching high-voltage,
high-current loads but perhaps was no good at switching
small currents in a wet and steamy environment.
I started looking for a better microswitch. The Fisher &
Paykel Parts Manual lists the microswitch part number as
436597 but they no longer sell that part.
The replacement is the sealed OOB (out of balance)
assembly, part number 420313. That includes a different
switch, bracket, lever and two-wire connectors. At ~$70,
this is considerably more expensive than just a microswitch. It looks as if the switch is now sealed and requires
a different bracket and lever because it is a different size.
Searching for the original part number on eBay showed
two sellers with photos of their microswitch that had
Omron part numbers D3V-6-2C24 and V-16-2C25 on them.
The Omron data sheets indicate 30mW and 15mW contact
resistance, respectively. The D3V-6-2C24 data sheet also
shows a graph for the micro load D3V-01 series at currents
as low as 0.16mA.
I think that a D3V-01 series microswitch might be suitable in this environment. But at the moment, the washing
machine is working with another generic microswitch from
my local store, so we will wait to see how it goes first. Next
time, the solution will be to use either that Omron device
or the Fisher & Paykel replacement part if the OOB micro
switch fails again.
I now realise why a normally closed switch was used
rather than normally open. If the contacts oxidise and the
out-of-balance lever operates the switch, the circuit would
not be closed, and the washing machine would not stop –
it would be hopping and jumping around in the laundry.
With an NC microswitch, the washing machine would just
stop working with a faulty switch.
SC
siliconchip.com.au
Australia's electronics magazine
June 2022 97
|