Normally for jaw crushers in mining industry earlier star delta starters and soft starters were
implemented for primary crusher [jaw type]
the load characteristics in the case of jaw crusher is continuously varying since the rock sizes coming in to the feed of various sizes.
offer the motor goes to overload peaks and comes down.
when parametrization the drive in v/f mode often the drive goes to fault mode, mainly due to overload. How about implementing Vector control mode with dynamic torque control?

For any high Torque applications like Jaw Crusher, V/F (being a scalar control) is never a right Torque control selection mode for drives. Variable frequency drive will obviously trip on over current faults whenever sudden dynamic response or Torque demanded is seen from the Load.
Also, if you operate a constant Torque application at low speed typically less than 30Hz in V/F control, VFD will trip on over current as it can not provide the required Torque demanded by Load.
V/F is best suited for variable Torque applications like Fan and pump.

The best selection in my understanding would be a modern Vector control drive operated in vector mode. Many modern drives have some best Torque Control features like FORCE (Field Oriented Control) that gives excellent Dynamic response, Torque control and are designed to run such application smoothly.
FORCE can provide control similar to that of a DC drive by separating Flux and Torque producing components and controlling them individually to achieve best Torque and speed control results.

Last day itself, i had a site where i had to commission a vfd in jaw crusher application
The drive was commissioned on vector mode with speed control
the drive is working satisfactory
but we did also try out v/f mode . we found that whenever the current crossed the motor rated current, the drive was unable to pump the required torque to the motor
When the drive was configured on vector mode the output voltage increased whenever there was requirement for torque, in trend function available in drive software we monitored the quadrature axis and direct axis component of output current. As per vector control theory, the quadrature axis component provides the magnetizing component while the other the load current. it was evident that the vector control is the best suitable one.
But whenever a situation occurs where the jaw crusher needs to be jogged from full loaded condition, many a time we had to jog the drive forward and reverse to breakdown the rock. the problem was later resolved by increasing the overload rating and the jog setpoint

Use Torque Control, Torque varies according to an analog measurement...and speed is not controlled...you can switch Torque control, speed control. I recommend an encoder for better performance. I used this on a mining application in Brazil, 2 inverters, one torque slave of the other. You may not have the master in torque, but the slave must be a torque slave.

What are you hoping to accomplish by varying the speed of a jaw crusher by the way? I don't think this is a viable application, as the change in speed does nothing in terms of changing the work done. A jaw crusher works by incrementally fracturing the rocks in a tapered space. The fracturing is done by the mechanical force exerted by the moveable jaw closing in and backing away. The kinetic energy to accomplish this is supplied by a massive flywheel that is kept spinning by a motor (or engine). Varying the speed of the motor does not change the annular space, all it will do is prevent the flywheel from recovering as fast, so the task will take longer to accomplish and no energy will be saved, in fact it may end up consuming more energy as a lower amount of m3echanical pressure on the rocks may allow it to be absorbed (as heat in the jaw material) rather than causing immediate fracturing, which means more repeated application with less work accomplished.

Flow control through a jaw crusher is typically done by controlling the vibrating feeder that dumps rocks into it. A typical feedback loop, if any, is to monitor the jaw motor current and as it increases, back off on the feed rate into the jaw. I have done it with a PID loop, but to be hinest, the caliber of people typically working in a crushing plant is such that they can't understand the process and a manual system tends to work better.

Direct Torque Control (DTC) is an optimised AC drives control principle where inverter switching directly controls the motor variables: flux and torque.
The measured input values to the DTC control are motor current and voltage. The voltage is defined from the DC-bus voltage and inverter switch positions.
The voltage and current signals are inputs to an accurate motor model which produces an exact actual value of stator flux and torque every 25 microseconds.
Motor torque and flux two-level comparators compare the actual values to the reference values produced by torque and flux reference controllers.
The outputs from these two-level controllers are updated every 25 microseconds and they indicate whether the torque or flux has to be varied.
Depending on the outputs from the two-level controllers, the switching logic directly determines the optimum inverter switch positions. Therefore every single
voltage pulse is determined separately at "atomic level".The inverter switch positions again determine the motor voltage and current, which in turn influence
the motor torque and flux and the control loop is closed. Used abb drive ACS880 is the most powerful drive in the market.

Jeff
My thoughts exactly, in fact I have a customer with precisely this application and in an effort to minimize his jam ups through operator error, I am looking at putting a soft starter on the crusher and using its current output to control the speed of the vibrating feeder.
Just one question... the existing motor has a S/D starter and fluid coupling. With all that inertia, I am concerned about the soft starter being sufficient as they are looking to remove the fluid coupling. I'm thinking they may need the fluid coupling as well. Its a 90kw (120hp) motor.