Insane Hydraulics
Bold and audacious blog about fluid power
This troubleshooting episode is a good example of how modern electronic control systems can be easily messed up by things that would never affect good old thick-skinned (and bullet-proof) relay-based circuits.
Last summer I had to troubleshoot a malfunctioning scrap-metal press, that had a nasty come-and-go failure - the main pressing cylinder wouldn't extend occasionally. Sometimes it would work fine for weeks, and sometimes it would fail almost every hour. The technicians were able to track the malfunction down to a pressure switch, connected to the pressing cylinder bottom port. It appeared that the pressure switch would stay in the "ON" position all the time, giving the faulty signal and not allowing the machine's logic to extend the cylinder after it's retracted. The client concluded that the problem was caused by pressure trapped inside the manifold the switch was mounted on. The pressure switch had already been replaced several times, but the malfunction remained.
The pressing cycle was controlled by a small 24 VDC Siemens PLC, and the pressure switch was supposed to be doing what all pressure switches do - give a signal when the set pressure level was reached - a very simple on/off arrangement, with no proportional signals or pressure sensors.
Pressure readings showed that there was no "trapped pressure" problem, and tests confirmed that the pressure switch was functioning perfectly fine. The wiring seemed intact and the electrician double-checked there were no circuit interruptions or any extra connections between the pressure switch and the PLC. So... where could the "bad" signal be coming from?
It turned out that the faulty "ON" signal came from the wiring itself!
If you look at the pictures, you will see that the wiring wasn't protected from the elements and, to top it off, the plugs were of very poor quality, allowing rain and condensation water along with other contaminants to enter inside the plug and the wiring. Modern PLCs, as a rule, have high impedance digital inputs (this one had the impedance of around 5 kOhm), and generally consider any voltage above half of the supply voltage a "1", or "ON" signal, so a small leakage current of about 3 mA flowing through the liquid contamination was enough to signal the PLC and cause the faulty operation of the press.
Normally, the rainwater isn't conductive, but it was a scrap metal press, which meant that all sorts of acid and alkaline liquids and solids could get in contact with the wires that were situated right next to the pressing tunnel. When it rained - the machine failed more. When sunny weather settled - the wiring would dry out and the malfunction would miraculously disappear.
An instant and cheap solution that didn't require replacing the "intricate" wiring (I did replace the horrendous plugs) was applying a relay next to the PLC so that the pressure switch triggered the relay, which in its turn gave the 24V "ON" signal to the PLC input. The relay required 50 mA to trigger, almost 20 times as much as the PLC input, which made the leakage current of the wiring irrelevant. Furthermore, a protective metal structure was created above the pressure switch and the wires to protect them from the harsh environment.
This type of malfunction is not uncommon. The use of PLCs to control automated hydraulic machinery is a "must" these days, and the fact that these gadgets can be triggered with low current signals means conducting liquid contamination of the wiring can jeopardize the normal operation. Therefore special care must be taken when installing and/or checking the PLC signal wiring to make sure it's as insulated and protected as possible.
This also shows that troubleshooting modern hydraulic systems is impossible without an understanding of basic electric circuits.