#robotics Logs

Jul 02 2018

#robotics Calendar

12:51 AM rue_shop3: yay, I made a second serial DAC current controlled motor driver
12:54 AM rue_shop3: so much faster after the first one
12:54 AM rue_shop3: it'll be even faster once I have serial DAC modules
01:10 AM rue_shop3: so this will be an interesting difference
01:11 AM rue_shop3: :( I wanted to tig up the plastic injection moulder I started yesterday
01:11 AM rue_shop3: ok, the idea with the servo is this, we have two positions
01:11 AM rue_shop3: we want to apply force to make the positions match
01:12 AM rue_shop3: the force is applied to both servos at the same time, in the same amount
01:12 AM rue_shop3: I'm the only one left alive in the world
01:12 AM rue_shop3: and each servo applies the force in the opposite direction
01:12 AM rue_shop3: so, I can, for this, parallel the serial lines to the motor drivers
01:13 AM rue_shop3: I suppose, for the first test, one master avr can take both positions as analog values
01:13 AM rue_shop3: but I'll want to divide it in two peices later
01:14 AM rue_shop3: is there any point to waiting for anyone else on the internet? its been like 16 years now
01:14 AM rue_shop3: the day you give up is the day someone comes along
01:14 AM rue_shop3: do you really think I want to play fates game like that?
01:15 AM rue_shop3: dosn't the possability make you want to hold on
01:15 AM rue_shop3: not with the remaining odds
01:15 AM rue_shop3: two controllers
01:15 AM rue_shop3: yea, ok
01:15 AM rue_shop3: so, each is doing its adc for position feedback
01:16 AM rue_shop3: so, maybe 3 points?
01:16 AM rue_shop3: one on each side that takes force and gives back position, one in the middle that takes the two positions and sends the level of force
01:17 AM rue_shop3: the servo loop calcs cant be split between the two ends?
01:17 AM rue_shop3: wouldn't that make it more complex?
01:19 AM rue_shop3: well, if both sides do their own math
01:19 AM rue_shop3: and they do it the same
01:19 AM rue_shop3: then your set
01:19 AM rue_shop3: hmm
01:21 AM rue_shop3: pin count
01:23 AM rue_shop3: RX, TX, ADC, DRIVE_C, DRIVE_D, DRIVE_S
01:23 AM rue_shop3: not quite small enough for a t13
01:40 AM rue_shop3: I need some big-ass pins for the mecha joints
07:43 PM rue_shop4: OK who can *quickly* remmeber if tig welding is EP or EN? (torch + or -)
07:45 PM rue_shop4: ok I think its EN, I'm going for it
07:45 PM rue_shop4: zhanx, Tom_L ?
07:45 PM rue_shop4: kat?
07:45 PM zhanx: torch -
07:45 PM zhanx: wait
07:46 PM zhanx: shield or no shield ?
07:46 PM zhanx: or is it inverted wave?
07:46 PM zhanx: or
08:06 PM rue_shop3: DC, w/shield gas yea, cause its tig
08:07 PM rue_shop3: I'm gonna weld an extra thing cause thats fun
08:09 PM Tom_L: does polarity depend on the metal?
08:10 PM robotustra: it is
08:11 PM robotustra: and on type of weld
08:11 PM robotustra: if you adding metal from electrode its usually EP
08:12 PM robotustra: if you just melt metal with tungsten electrode it's usually EN
08:12 PM rue_shop4: yea, aluminum is opposite of steel
08:13 PM rue_shop4: I'm making one of those silly plastic moulding things
08:14 PM rue_shop3: ok, the stm32 is pwm'ing at 9khz
08:14 PM robotustra: I installed the version of t-flex with cnc module and read docs how to generate trajectories
08:40 PM rue_shop3: :/ at ~10Mhz, the output of the stm32 dont look too pretty
08:44 PM rue_shop3: yea, hmm
08:45 PM rue_shop3: ringing like all hell, recon it needs to be terminated somehow to run that fast
08:50 PM rue_shop3: so, one challange is to run the pll at 50Mhz instead of 72
08:50 PM rue_shop3: so I can divide it down properly
08:55 PM rue_shop3: ah yea
08:55 PM rue_shop3: and there's the rub
08:55 PM rue_shop3: you cant go nicely down from your max freq
08:55 PM rue_shop3: the next step down from 10Mhz is 8.3Mhz
08:55 PM rue_shop3: thats not cool
09:06 PM rue_: so then, where did I put those variable capacitors?
10:05 PM rue_shop3: so colpitts or heatley
10:05 PM rue_shop3: heartley
10:07 PM rue_shop3: heartley indeed
10:07 PM rue_shop3: tapped inductor
11:29 PM rue_shop4: nope, the variable cap I have just dosn't have the range
11:30 PM rue_shop4: I'll have to build a cap with a better ratio
11:30 PM zhanx: sand paper?
11:38 PM rue_shop3: variable cap