Repair of a Gate
Quadro Comando ZK1
|
|
I wasn't sure how to entitle
this repair. The problem was damage to the circuit board used
to control the opening and closing of a pair of gates. Because
the board is located outdoors it suffers from potential damage,
not only from the elements, but creepy crawlies. In this specific
case probably one or more snails has met its end when it shorted
out the mains power supply and in its demised state severely
messed up the circuitry resulting in a lot of burning. Several
resistors caught fire and the fibreglass board material was badly
burnt. In fact, examining the board showed previous repairs from
similar causes. |
| |
|
Fortunately the designers
had elected to mark the board with resistor and capacitor values
so replacing some burnt resistors was straightforward, but alas,
board damage had erased the values of two resistors. A little
bit of evidence existed, being a somewhat blurry image of the
worst affected area, but this wasn't enough to precisely define
two of the resistors. I imagined I could trace and reverse engineer
the area affected but here there's a problem. Later versions
of the board use a microprocessor but this example uses a hard-logic
design. Basically this means that the designers incorporated
lots of diode logic making elements such as AND, NAND, OR, NOR
and EXOR functions. Nowadays these funcions are carried out by
firmware but in this example I counted well over fifty 1N4007
diodes. This makes reverse engineering a bit longwinded. |
| |
|
I decided first to restore
the tracks carrying power. I could then make measurements to
help decide on two unknown resistor values. These were located
in the burnt area at the middle of the board. I'd initially thought
the three worst to be 15Kohm, 4.7Kohm and zero ohms, but after
cleaning away soot I found the board under the first resistor
was still marked "150Kohm", so (because of this size)
is the middle one 47Kohm rather than 4.7Kohm? This resistor feeds
the base of a BC337 transistor with the 150Kohm connected to
the emitter. The power rail appears to be 24 volts (the relay
coil ratings) so the base will sit at around either 0.7V for
4.7Kohm or 7.5V for 47Kohm. However the two resistors do not
define the BC337 base voltage by way of simple potentiometer
action. I found instead that the 47K or 4.7K is connected to
a pair of diode cathodes, the anode of one being defined by a
12 volt zener diode and two resistors 470 and 4.7Kohms. In other
words the BC337 is being turned on by the voltage applied by
one or other of the two diodes (roughly 10 volts). This then
is a logic circuit no doubt subject to perhaps a few assorted
parameters. If say 10 volts is present at the diode cathode this
will result in a voltage at the BC337 base, turning it on. The
current flow into the base will be multiplied by NPN transistor
action at its collector )maybe wired to a relay coil?). For a
4.7Kohm resistor the 150Kohm from base to emitter will have very
little effect when the BC337 is turned on so the current flow
will be about 10 volts over 4.7Kohms or about 2mA. If the resistor
was 47Kohm the current would be 0.2mA but because this would
bring into play the 150Kohm resistor reducing the driving voltage
by a third making the current as low as 0.13mA This may not be
enough to turn on the BC337, so a value of 4.7K is more likely
to be correct.
Why did this 4.7Kohm resistor
burn up? The rating is only maybe a quarter watt so excessive
heating from being over-run must be the answer. Say the resistor
dissipated a watt. This sort of power would require a potential
of 216 volts. The board is clearly designed to run at 24 volts
because that's the rating of the relay coils, however one can
see signs of poor design practice in that there are two mains
voltage capacitors presumably for starting a motor or two and
the tracks feeding these are mixed up with tracks used for low
voltage DC.
Under normal circumstances there
isn't a problem... but take into account snails and even fingers
and 240 volts AC can easily get mixed up with ground or 24 volt
circuitry.
Below are pictures of the damage.
|
 |
|
Besides the group of three
burnt resistors left of centre there were also a burnt 4.7Kohm
adjacent to the marking "QUADRO COMANDO" and a burnt
470 ohm below this. Further checks revealed a short-circuit 12
volt zener diode marked "13", a bad capacitor visibly
burst top left, and a short-circuit BC337 to the right of that
burnt 4.7Kohm. The preset potentiometer above the three burnt
resistors had been melted and was unserviceable. I measured the
resistances to confirm its setting and restored these proportionally
in a new 1Mohm preset.
Numerous circuit tracks were
open circuit as can be seen repaired with white wires in the
penultimate picture below.
|
 |
 |
 |
 |
 |
|
Above, the site of a group of
destroyed components three of which have been already removed,
470ohm, 4.7Kohm and a BC337 but included that 12 volt zener diode
which was short-circuit... centre. |
|
|
Now two views of
the repaired circuit board. Oddly the three fuses, bottom left
were all intact.
Whether the board will work
when tested with the gate remains to be seen...
Well it failed to work and I
first suspected the connection to the transformer or even a thermal
fuse in the transformer winding. I was pretty close as I soon
found a resistor connected to the transformer input. This had
red and purple rings and measured 270Kohm. I'd measured it previously
but then I noticed a marking underneath it.. "2.7".
It had a gold third band not a yellow band. I fitted a new 2.2
ohm resistor and fingers crossed the board will now be OK...
You can see the bad 2.7ohm resistor top left below that blown
electrolytic. It's not burnt so hadn't registered like the others.
How about the mains motor start
capacitors?
I didn't check these because,
prior to the incident both gates had been working, but I'll check
these as both might be degrading.
|
 |
 |
|
|