#robotics Logs

Apr 29 2019

02:02 AM rue_bed: zhanx, so, I cracked open a 4x4 ABS controller and there was a 1 axis IMU that has analog output
02:07 AM rue_bed: 5V power, hooked it up to the scope, works like a charm
02:07 AM rue_bed: Tom_L, need to know more about the balancing bot
02:07 AM rue_bed: was the trick to maintain angle or not be falling over?
05:55 PM gaze__: you guys have a preferred way of building stuff out of prefabricated parts? Lego? Meccano? something else?
05:55 PM gaze__: I like 80/20 rail too but I usually have to cut that to preferred lengths
05:58 PM zhanx: gaze__, lego for me
05:58 PM zhanx: i have a ton of them
05:58 PM gaze__: lego seems like a good way to go
05:59 PM gaze__: I wonder if you can 3D print parts which are precise enough to snap to legos
05:59 PM zhanx: well i use the lego nxt really
06:03 PM gaze__: hmm interesting
06:23 PM Mattx: Anyone familiar with LTspice around? I want to simulate a 50Hz noise on my voltage source (thought it's DC) and input signal (a wire connected to your arm), what are good parameters for this?
06:24 PM Mattx: I guess the way to go is replace the dc voltage with a ac voltage source at a certain offset, but what amplitude?
08:01 PM zhanx: working on my robot controller
08:01 PM zhanx: i borked up the system getting the 2.8 tft working
11:00 PM Mattx: rue_mohr, ping. have a min?
11:01 PM rue_mohr: I'm here
11:01 PM Mattx: Hey, hi
11:02 PM Mattx: I'm doing some simulations on ltspice, are you familiar with it?
11:02 PM rue_mohr: I stop at simulation
11:03 PM Mattx: The thing is I'm simulating the circuit I linked yesterday, remember?
11:03 PM Mattx: but I'm not sure how to add noise on the signal
11:03 PM rue_mohr: simulations are generally useless, as much as people argue to support them, they do not make real world results
11:03 PM Mattx: I thought about adding 50Hz noise but then it's removed in the first step, which is a diff amp
11:03 PM rue_mohr: well, you cant use a PN juntion to make white noise, cause its a simulation
11:03 PM Mattx: as the noise is exactly the same in both inputs, it's canceled
11:03 PM Mattx: makes sense right?
11:04 PM rue_mohr: its a simulation
11:04 PM rue_mohr: if you want to make it, make it
11:04 PM rue_mohr: dont simulate it
11:04 PM Mattx: yeah, it's useful for me believe it or not, I can test multiple gains and have a better idea of where it clips the signal, etc
11:05 PM Mattx: maybe you run the simulation on your head, I'm just learning and don't have that module on mine :P
11:05 PM rue_mohr: they will give you ballpark, and then you need to build it
11:05 PM rue_mohr: because the real world stuff does NOT come up in the simulator
11:05 PM Mattx: I have the circuit built but it's not working
11:06 PM rue_mohr: ah, ok what signal test source are you using?
11:07 PM Mattx: no signal source other than the electrodes, I don't have a signal generator around
11:08 PM Mattx: so it's difficult to test as you can imagine
11:08 PM rue_mohr: how about a 555
11:08 PM rue_mohr: or an arduino
11:09 PM rue_mohr: you just need to use a large votlage divider
11:09 PM Mattx: you mean using a PWM source?
11:09 PM Mattx: I can do that I guess
11:09 PM rue_mohr: or just a pin that you totggle high and low
11:10 PM Mattx: yeah, could be
11:10 PM Mattx: anyway, the simulations are not useless for now, I'm learning quite a lot
11:10 PM Mattx: if the noise was the same on both inputs it should be canceled on the first diff amp, is that correct?
11:11 PM rue_mohr: if its common mode noise
11:11 PM Mattx: what do you mean with common mode?
11:12 PM rue_mohr: common mode means that both inputs go up and down at the same time
11:12 PM rue_mohr: diffrential means the same signal isn't on both inputs
11:12 PM Mattx: oh, well, I'm guessing it does. the 50hz noise from the mains should be equally present on both inputs, right?
11:13 PM Mattx: maybe not the rest, but that at least?
11:13 PM Mattx: (I know the answer is no, otherwise I wouldn't need a notch filter)
11:14 PM Mattx: I'm just missing why it's not that straightforward
11:20 PM Mattx: rue_mohr, in other words, I wonder why the 50hz noise from the power line is not "common mode" noise
11:22 PM rue_mohr: yes and no, it depends on the impedence between the inputs
11:22 PM rue_mohr: but with the notch filter its not supposed to matter
11:23 PM rue_mohr: if you have 50Hz signal across you, then it turns up as differential siganal on the inputs
11:30 PM Mattx: ignoring the notch filter, only considering the first stage which is a differential amplifier...
11:30 PM Mattx: it should have the same noise on both its inputs, so the output should be clean of 50Hz noise
11:30 PM Mattx: that's what I'm saying
11:41 PM rue_mohr: mhm
11:41 PM rue_mohr: in the real world you pick up more 50Hz ponential on one electrode thant he toher
11:43 PM Mattx: makes sense, then a good way to simulate the noise is adding a 50Hz signal on one electrode and a slightly different one on the other
11:43 PM Mattx: the amp will amplify that difference and the notch should clean it
11:58 PM MLE_Online: Hey rue_mohr !
11:59 PM rue_shop1: hey!
11:59 PM rue_shop1: so its working!