Back
[00:00:03] <justan0theruser> I see
[00:01:24] <anniepoo_> alternatively you can use 3 springs, compressed with pull cords
[00:07:22] <justan0theruser> I asked the mfr of my planetary gear whether they sold torque limiters for it :p
[00:07:51] <anniepoo_> that's reasonable
[00:08:20] <justan0theruser> they've been very kind and helpful so far
[00:09:23] <justan0theruser> I assume the smartest way to do this if I wanted this to be sustainable, would be to have a tension sensor and stop signalling the motor if it reaches a certain threshold
[00:09:33] <anniepoo_> most manufacturers will be. Just please don't pester them, that'll make it hard on the next hobbyist. But manufacturers do understand that hobby things sometimes become commercial things and they end up with large orders
[00:10:03] <anniepoo_> well, you don't want software or anythign close to it in a loop that critical
[00:10:43] <justan0theruser> don't think I'm pestering, just asking if they sell the part
[00:10:55] <justan0theruser> they're chinese, so they seem pesterable. They sent me weird emojis
[00:11:11] <anniepoo_> cool
[00:11:51] <anniepoo_> I spent over an hour on skype with the guy who sold us our big laser one night, just talking about modern chinese history
[00:12:21] <justan0theruser> http://imgur.com/a/r94tO
[00:12:28] <justan0theruser> lol
[00:14:28] <Jak_o_Shadows> z64555, I don't know what you're talking about with RLC networks, it's just a mass spring damper isn't it?
[00:14:32] <Jak_o_Shadows> Everything is a mass spring damper
[00:15:38] <z64555> RLC is a resistor, inductor, capacitor
[00:15:43] <z64555> but basically, yeah
[00:18:52] <anniepoo_> the math is the same for many situations
[00:20:02] <anniepoo_> this second order differential equation appears in mechanics, electronics, and even crazy stuff like ecology and economics
[08:13:23] <justan0theruser> :/ aliexpress doesn't know what a torque limiter is
[08:13:33] <justan0theruser> the aliexpress merchant I used I mean
[11:25:38] <z64555> justan0theruser: try searching for slip-clutches
[11:25:59] <z64555> and overload clutch
[11:26:10] <z64555> https://en.wikipedia.org/wiki/Torque_limiter
[11:27:19] <z64555> huh, "magnetic particle clutch"
[11:32:04] <z64555> the hysteresis clutch is also fascinating
[12:24:06] <rue_house> ok I have to write a serial reciever thread for this cnc controller
[12:28:18] <anniepoo_> ??? the merchant doesn't know. Ok, maybe they don't know in english, or you're talking to somebody who's a middleman
[12:33:37] <anniepoo_> but AJava, can_unify(X,_) :- var(X). can_unify(_,Y) :- var(Y). can_unify(X,Y) :- X = Y.
[12:34:54] <anniepoo_> sorry, ww
[12:52:30] <beanbag-> CANUCK
[13:27:22] <rue_bed> ok,I think
[13:27:31] <rue_bed> I count pwm cycles
[13:27:43] <rue_bed> start with a counter at 0
[13:28:04] <rue_bed> at the end of the first pwn cycle I count it to 1
[13:28:58] <rue_bed> if the counter is 2, I copy all the adc values to a buffer, and trigger the peak detector reset circuit, then clear the counter to 0
[13:29:08] <rue_bed> nope
[13:29:36] * beanbag- bites rue_bed LTNS
[13:29:38] <rue_bed> I need to know the adc values happened after the first pwm cycle
[13:30:01] <rue_bed> flyback, do you remember why you left
[13:30:14] <beanbag-> nope
[13:30:38] <rue_bed> we all hate you, and at some point you realized it and went away, saying that we all suck
[13:30:50] <rue_bed> do we have to go thru it all again
[13:30:52] <beanbag-> ok I will leave then
[13:30:59] <rue_bed> or is this where you start saying you have changed
[13:31:15] <beanbag-> na ill just leave it's fine
[13:31:18] <rue_bed> anyhow
[13:32:04] <rue_bed> the adc values need two cycles to get all the samples
[13:32:36] <rue_bed> so, after the peak detector reset is finished, I suppose I set a counter to 1
[13:33:10] <rue_bed> but, the current values slope, so the first pwm pulse needs to be finished before the samples start
[13:33:32] <rue_bed> so,the adc shouldn't sample if the counter is 0
[13:33:47] <rue_bed> and I need a flag saying that all the adcs have finished sampling
[13:33:55] <rue_bed> beofre I copy them to the buffer
[13:34:46] <rue_bed> I need a waveform diagram
[13:35:17] <rue_bed> ______....-----_________.....------___________......-----_______
[13:35:47] <rue_bed> ....-----_________.....------___________......-----_______3 cycles
[13:36:09] <rue_bed> no samples in period 1
[13:36:19] <rue_bed> samples start at period 2,
[13:36:35] <rue_bed> by the end of period 3, all channels are sampled
[13:36:43] <rue_bed> hmmm
[13:36:48] <Snert> ... and there ya have it. Time travel in a nutshell :)
[13:37:04] <Snert> make that work for me bynoon tomorrow, ok?
[13:37:05] <rue_bed> :)
[13:37:43] <rue_bed> what I dont like is that it takes 3 periods to sample the data for a feedback loop
[13:37:54] <rue_bed> there IS another factor
[13:38:07] <rue_bed> the adcs sample current, but also position
[13:38:20] <rue_bed> I dont so much care when the positions are sampled
[13:39:02] <rue_bed> BUT I'm not in as much control of when the sampling is done because of 10 channels being done by a different chip asynchronously
[14:09:25] <rue_bed> waaaaait a sec
[14:10:14] <rue_bed> while (IsLow(syncBit, syncPort)); while (IsHigh(syncBit, syncPort));
[14:10:14] <rue_bed> USIDR = byte;
[14:10:18] <rue_bed> one one avr
[14:10:42] <rue_bed> ---
[14:10:44] <rue_bed> / wait for last byte
[14:10:44] <rue_bed> syncHigh();
[14:10:44] <rue_bed> Delay(300);
[14:10:44] <rue_bed> while(!(SPSR & (1<<SPIF)));
[14:10:44] <rue_bed> //save
[14:10:45] <rue_bed> result = SPDR;
[14:10:47] <rue_bed> // pulse the sync
[14:10:49] <rue_bed> syncLow();
[14:10:51] <rue_bed> Delay(300);
[14:10:53] <rue_bed> // start next transfer
[14:10:57] <rue_bed> SPDR = 0xDA; // write another byte to get another byte
[14:10:59] <rue_bed> on the other avr
[14:11:24] <rue_bed> the delay dosn't need to be 300 after the high
[14:12:26] <rue_bed> er, after you go low, wait like 4 cycles and .... oh you have to wait for the ... no
[14:12:27] <rue_bed> wtf
[14:12:33] <rue_bed> I need paper
[14:13:36] <rue_bed> you capture the comm loop while the sync is low
[14:13:59] <rue_bed> then...
[14:15:17] <rue_bed> you put it high to get it to transmitt....
[14:15:23] <rue_bed> this is all messed up
[14:15:26] <rue_bed> arg
[20:02:32] <rue_shop3> USB CONNECTORS: the first two orientations your try will ALWAYS be wrong...
[20:02:43] <rue_shop3> !?!?!
[20:08:08] <Tom_L> rue_shop3 why won't winscp let me in?
[20:10:37] <rue_shop3> root is turned off in the new distro
[20:10:48] <rue_shop3> you have to google how to enable root login
[20:12:15] <Tom_L> i can login root on the pc
[20:19:31] <Tom_L> ty
[20:21:47] <Tom_L> i can no longer login as the user i originally created during install...
[20:21:56] <Tom_L> no fault of what i just did..
[20:41:59] <rue_shop3> the new ssh disables root login
[20:42:11] <rue_shop3> yo have to go change a file to make it so root can login over th network
[20:42:41] <rue_shop3> google it, you changed something from not_typed to enabled or something
[20:42:47] <rue_shop3> change..
[20:43:00] <rue_shop3> its just network login that root is disabled for
[20:44:26] <Tom_L> i got the root thing fixed
[22:38:55] <rue_house> lightdm rhing?
[22:45:48] <justan0theruser> b
[23:48:29] <rue_house> b is obsolete, use c
[23:48:54] <rue_house> did you sheer anything yet?
[23:49:04] <justan0theruser> pls stop ignore evading
[23:49:14] <rue_house> I'm not, there really are 8 of me
[23:49:25] <justan0theruser> interesting, please let me know the nicks of the 8 of you
[23:49:31] <rue_house> I know more than your giving me credit for
[23:49:47] <rue_house> if you cant work that out, I cant help you with anything
[23:49:48] <justan0theruser> clearly spelling isn't in that knowledge set
[23:50:41] <rue_house> there is a care set too ya know
[23:51:36] <justan0theruser> rue_house: I'll give you a shot. find me a 100 lb-ft torque limiter
[23:51:49] <rue_house> hah
[23:51:56] * rue_house hands justan0theruser a 3 amp fuse
[23:52:14] <justan0theruser> sigh
[23:52:39] <rue_house> oh, you gonna tell me torque isn't proportional to current?
[23:52:50] <rue_house> for a dc brushed motor?
[23:52:58] <rue_house> with a backdrivable gearbox?
[23:53:20] <justan0theruser> oh you're gonna tell me you think a motor can handle more torque than it is rated for through removing current
[23:53:39] <rue_house> ?
[23:54:15] <rue_house> if you dont want a one-time safety limit, use a current controlled driver
[23:54:46] <justan0theruser> what happens if I apply 200 lb-ft of torque to a gearbox rated for 100 lb-ft
[23:55:01] <justan0theruser> and where does the fusebox come in
[23:58:49] <rue_house> sorry, trying to fin a video, and I got distracted
[23:59:43] <rue_house> https://www.youtube.com/watch?v=IXXLvoa9BXk
[23:59:54] <rue_house> this is a motor under constatnt current control