Q&A Forum
Hello, all.
What rpm error (brake rpm - target rpm) would we expect for a Telma unit with 300mm diameter rollers? during a power run or rpm run?
(using a YourDyno eddy current PS for the PWM to eddy brake)
We are seeing, intially a slight instability (40-70rpm) followed by a lag at the upper end of the rpm range of 200-300 rpm.
settings: P= 0.4, I=0.25, D=0.05. update rate = 15hz. (tried 10 and 20Hz)
run: 5000 to 9000 rpm, 400Hz/sec.
To me the 'I' term is too slow. doesn't seem respond over the final run to correct the steady state error, or the initial I factor is too high for the start of the run.
If we increase the P term the system goes unstable. if we increase the I term the system goes unstable at the lower rpms.
The Telma is quite a large unit, we are seeing 55% brake @ 280 BHP.
my question:
should the system be capable of controlling better than this? considering it's an electrically controller Telma unit.
Do we need to use the "PID parameter scaling to apply a higher PID parameters at the higher rpms? or is there another trick?
thanks
Ian (UK)
Hi there,
A good starting point for many dynos is Kp = 0.5, Ki = 0.5 and Kd = 0.05. You can plot the different PID parameters vs Time to see which one is causing oscillations, if any.
Note that if Ki is too low, then Kp will often oscillate. Normally you can keep Ki as high as Kp or even higher.
If you send me a csv file of a run then I can take a look. Send also the saved config file and a couple of raw logs.
The exact overrun/underrun that is to be expected will vary with moment of inertia, the engine and the PID setup. What is more important though is a smooth run with no oscillations. For example the Load Control regulator by design allows some overrun (i.e. acceleration is higher where the torque is higher), and this gives a more natural run than a pure hard PID.
Hi,
emailed you the CVS and Config files
emailed you the CVS and Config files
Hi, emailed you the files