Page 2 of 2

Re: TCC load hysteresis

Posted: Wed Nov 06, 2013 8:41 am
by Bernard Fife
Tobmag,

I am not sure I fully understand. I think the existing code should do what you want. Using the values from your example above, if we had settings of 80 and 10, then the TCC would:
- lock if below 70 (80-10)
- unlock if the load climbs to 80.
- between 70 and 80 the TCC should remain in it's current state. So if it was unlocked (load falling), it should stay unlocked, if it was locked (load rising) it should stay locked.

At least that's the way it is supposed to work. It's certainly possible that I have missed something, of course!

Ultimately, though, there will always be conditions under which the operation is somewhat unstable (due to the on-off nature of the TCC). That's why we have the time hysteresis that prevents rapidly cycling.

The indicator colors are set in a line in the INI like this:

Code: Select all

   indicator = { lock_TCC            }, "TCC Off",       "TCC Applied",    green,   black,    green,   blue;
near line 2734. The four colors are: "off background ", "off text", "on background ", "on text". I have changed these to:

Code: Select all

   indicator = { lock_TCC            }, "TCC Off",       "TCC Applied",    green,   black,      red,  white;
in the attached INI, but you can edit them to whatever you like (there are plenty of examples in the INI itself).
GPIO_MShift_4134.ini
(217.63 KiB) Downloaded 582 times

(Note that 'TCC applied' is shown whenever the PWM isn't zero.)


Lance.

Re: TCC load hysteresis

Posted: Thu Nov 07, 2013 1:33 am
by tobmag
Lance.
- lock if below 70 (80-10)
- unlock if the load climbs to 80.
- between 70 and 80 the TCC should remain in it's current state. So if it was unlocked (load falling), it should stay unlocked, if it was locked (load rising) it should stay locked.
Yes you are right, it is working correctly it was mee not testing this in the right way on the bench.
Did some more testing last night and it works as you stated so all is good.

BR//Tobmag

Re: TCC load hysteresis

Posted: Thu Nov 07, 2013 8:24 am
by Bernard Fife
Tobmag,

Okay, great, thanks for confirming that. But if there's anything that comes up that strikes you as strange, please let me know.

Lance.