Commit Graph

54 Commits

Author SHA1 Message Date
Hakan Bastedt
b534e3d4da Kludge with 'tone' of steppers. Gives ferror, but nice tone 2024-01-21 23:29:59 +01:00
Hakan Bastedt
1f97534133 New pwmCycleTime different from sync0CycleTime 2024-01-20 12:37:42 +01:00
Hakan Bastedt
e6bfe4f880 Tests for growling. That 10xcycle time seems to do it, but it's not right. 2024-01-19 21:31:15 +01:00
Hakan Bastedt
8f05f33e58 Both steppers work in simulated environment 2024-01-18 18:21:21 +01:00
Hakan Bastedt
9e311038c1 TIM8 replaced with TIM3. TIM3,chan4 works on PC9. TIM8,chan4 should work but doesn't. How should one know? 2024-01-18 17:21:12 +01:00
Hakan Bastedt
9ab4afabe4 New enable input. Add dc ability. Set only SM2 synchronization. Various to make lathe work 2024-01-18 14:26:51 +01:00
Hakan Bastedt
38825bbaf3 Trying to make stepper suing tim8 work 2024-01-18 14:25:19 +01:00
Hakan Bastedt
0a04124b35 Remove TIM8, TIM8 is stepper generator timer. 2024-01-17 00:29:06 +01:00
Hakan Bastedt
760944afe5 Various improvements. New pdo variable stepsPerMM. 2024-01-17 00:24:17 +01:00
Hakan Bastedt
044e8fd2c5 Removed unnecessary condition in timerCB, was aleady test on ´steps´ 2024-01-15 23:10:25 +01:00
Hakan Bastedt
697ea19dae Bug, digitalWrite of stepPin should be dirPin in timerCB. Reshuffling 2024-01-15 23:07:46 +01:00
Hakan Bastedt
20ca9d4974 Maybe works better now. Using G0 can not provoke the scrape sound. Jogging for a long while can provoke still. 2024-01-15 21:12:23 +01:00
Hakan Bastedt
c6857d0be2 Still problem with overlapping cycles in stepgen. TIM8 works for stepgen now. 2024-01-15 20:21:14 +01:00
Hakan Bastedt
deadd488dd Two stepgens active pending lathe test 2024-01-13 22:44:08 +01:00
Hakan Bastedt
25c2aa7b3c Some cleanup. 2024-01-12 14:37:25 +01:00
Hakan Bastedt
7a63d27303 Two encoders and two steppers declared. Encoder is 64-bit again 2024-01-11 21:10:27 +01:00
Hakan Bastedt
6176166b3a More encapsulation of timer info. 2024-01-11 16:44:28 +01:00
Hakan Bastedt
5df911296c sync0CycleTime static member of StepGen 2024-01-11 10:54:13 +01:00
Hakan Bastedt
a4488da9fa Cleanup encoder code 2024-01-11 10:42:29 +01:00
Hakan Bastedt
cd8388ea58 Reshuffle in main.cpp 2024-01-10 23:42:05 +01:00
Hakan Bastedt
6f29a0d492 Further cleanup and modularization 2024-01-10 22:27:04 +01:00
Hakan Bastedt
3bbf089d01 MyEncoder split into cpp and h file 2024-01-10 22:05:38 +01:00
Hakan Bastedt
2e4e768d5e Raw classified encoder. Compiles, untested 2024-01-10 21:58:11 +01:00
Hakan Bastedt
6c168d96d9 Test with different stepscale. ok 2024-01-10 20:25:26 +01:00
Hakan Bastedt
f8fc5fccd6 Max step-rate introduced - solves initial step storm 2024-01-10 20:08:30 +01:00
Hakan Bastedt
f6abd73c58 Split class into cpp and h file. Works at commands, but initial isn't ok 2024-01-10 18:59:16 +01:00
Hakan Bastedt
c5a95074dc Moved stepper into C++ class. Seems to work. Basic 2024-01-10 18:44:58 +01:00
Hakan Bastedt
8614d1dfd3 Moved stepper into C++ class. Seems to work. Basic 2024-01-10 18:44:40 +01:00
Hakan Bastedt
9adae08b98 It seems to work now, with reload in the timer_CB. Avoid micros() 2024-01-09 22:20:20 +01:00
Hakan Bastedt
0b9ce37200 Reloading in irq might possibly work 2024-01-09 10:39:56 +01:00
Hakan Bastedt
120b423f59 Restructure, makes pulses again 2024-01-08 21:13:34 +01:00
Hakan Bastedt
222e5857dc Reorganize 2024-01-07 19:38:22 +01:00
Hakan Bastedt
9dcc713fab Added a commandedVelocity variable for hal 2024-01-06 20:53:34 +01:00
Hakan Bastedt
ccde6ef15e definitely works better now with SM2 sync. handleStepper is back. 2024-01-05 21:37:55 +01:00
Hakan Bastedt
7e06b0ce68 ooops it was set to use DC sync, not the intention, SM2 sync was intended. Switch to SM2 sync 2024-01-05 21:33:06 +01:00
Hakan Bastedt
37d3ea6567 Solution was to NOT do the steps in the IRQ. Set a flag and do pulses outside of the irq. 2024-01-05 19:33:13 +01:00
Hakan Bastedt
ed168df64e It actually makes pulses at 1ms intervals here. Saving this moment. 2024-01-05 17:46:31 +01:00
Hakan Bastedt
1168f7e5ad Now it is occasionally updating correct 1 ms. wip. 2024-01-04 23:00:09 +01:00
Hakan Bastedt
93405efd63 Why is actualposition updated every 2 ms, not every 1 ms? 2024-01-04 22:23:52 +01:00
Hakan Bastedt
484c984e49 Basic function using SM2 sync inteerupt. Stepgen half functional 2024-01-04 21:38:40 +01:00
Hakan Bastedt
e6cd5356c9 tests with linuxcnc, DC sync is not the way to go, leaving that 2024-01-04 11:39:01 +01:00
Hakan Bastedt
5978ebec24 Pending tests. Merging this branch 2024-01-02 23:58:29 +01:00
Hakan Bastedt
d4a83eae80 Use frequency instead of pulse width - much better resolution at high freq. 2024-01-02 23:55:32 +01:00
Hakan Bastedt
2e8e938345 Bugs found. set duty cycle after overflow() 2024-01-02 23:51:58 +01:00
Hakan Bastedt
cc1ca73219 wip, local testing to start 2024-01-02 23:32:47 +01:00
Hakan Bastedt
b57bd740ab double float = fal_float size added to lcec. LCNC starts and works 2024-01-02 18:46:26 +01:00
Hakan Bastedt
8007a2ff6c Test with minium pulse delay 5 usec - can be set easily. 2024-01-01 22:52:14 +01:00
Hakan Bastedt
a21e932b5a Direction pin works in test 2024-01-01 22:48:50 +01:00
Hakan Bastedt
64d62a954a Basic test with puls sequence 2024-01-01 22:42:12 +01:00
Hakan Bastedt
f99d9bab77 Here are those changes... Worked in another repositiry hehe 2024-01-01 20:40:29 +01:00