Troubleshooting SCR Firing Problems
User Rating
Score
Noisy motors, overheating transformers - it all adds up to SCR mis-firing: The Black Dog Problem
The Black Dog problem
Fig 1: Perro Negro II
Back in the late 1970's this phrased was coined by Hill Graham service engineers to describe a problem first seen on the Saipem Rig, Perro Negro. The rig was having problems with noisy motors driven by SCRs and these were also getting hot. As well as the motors being affected, the auxiliary distribution transformers were also getting overheated. The type of noise from the motors would have you convinced that there is a mechanical problem with the motor - perhaps a bearing failure, or faulty brushgear or commutator. You might not link the problem with other equipment overheating at all.
The source of the problem turned out to be a faulty thyristor in SCR 4, and the timing of the SCR firing pulses on the other SCRs which had drifted out of tolerance.
The Importance of Timing
You can think of an SCR as a 6-cylinder engine. Like an engine, if the timing of the firing of each cylinder is out, the engine will sound a bit rough, and put out-of-balance stresses on the crankshaft.
A thyristor bridge will not necessarily be adversely affected itself by this mis-firing, but the equipment it is driving will be. If one SCR is misfiring, the DC motor gets either an increased or decreased current pulse once every cycle, and is therefore driven by an uneven DC current (the current is a bumpy anyway - that's normal - but the bumps should be even). In the worst case this gives rise to excessive noise and overheating.
Furthermore, the 600V supply to the SCR is also subjected to a series of current pulses. These current pulses distort the voltage, which in turn imposes harmonic currents in other wound components connected to the same supply, such as transformers and AC motors, causing overheating. .
Causes of Misfiring
Fig 2: SCR Current Waveform: Unbalanced
Fig 3: SCR Current Waveform: Missing Pulse
Fig 4: SCR Current Waveform: All Good
The most common cause is likely to be the calibration of the controller (DC Module) which should be sent for repair and recalibration by a company like CONTRELEC. The cause of the misfiring may not be calibration - there could be excessive ripple on a power rail, or some other contributing factor. Controller problems can be easily isolated if the controller can be swapped out with a spare (provided the spare is in good order).
It has been known for a thyristor to fail in a mode that has it working as a diode, i.e. as soon as it becomes forward biased it starts conducting. The SCR bridge continues to be able to exercise some control because the return current path is through controlled thyristors, but the effect is easily observed on the SCR current waveform.
There could be other causes: snubber components, firing pulse transformer and pick-up in the thyristor gate leads. Make sure all components in the firing path arre in good order, and in particular, pay attention to small burden resistors connected close to the gate leads which may become dislodged, but are important for reducing spontaneous firing through pick-up
Include within the class of misfiring the possibility of one SCR not firing at all. The causes of this are many, including blown SCR fuse (which may not have triggered the fuse microswitch), faulty SCR, faulty firing transformer, faulty controller firing pulse output, disconnected gate leads (primary and secondary side of the pulse transformer). The effect is exaggerated on the current waveform because the current regulator compensates by increasing the firing on the following pulse to compensate for the missing one.
Checking for Misfiring
Checking for misfiring can only be done by checking the SCR current waveform with an oscilloscope
This can be done with a DC current probe (HED), but there may be a more convenient way. On Hill Graham and Ross Hill SCR systems the DC Module current feedback is derived by summing the output of 3 CTs on the AC supply to the SCR. This signal is made available for monitoring via test points on the module front plate, or you can monitor it at several places in the wiring to the module (pin 131).
Example waveforms are shown on the right.
It is worth checking this waveform periodically because it is an excellent indicator of the health of the SCR. Remember that a small misfire, or even a quite significant one, won't stop the SCR working. However it may have a degrading affect on other equipment which may not be apparent immediately.
Comments (0)