So, im asking to see how i could connect rotary encoders to my 5160s as apparent they support that, if anyone knows whats going on feel free to drop in, heres documentation from trinamics website
#5160 Rotary encoder support
young sandal
undone fossilBOT
So, im asking to see how i could connect rotary encoders to my 5160s as apparent...
young sandal
wow chatgpt is good at this stuff
ionic edge
Yeah, seems like some manufacturers call index channel N, I thought that it's bad enough that index could be called either Z or I
ionic edge
I wonder what's the Index channel for here
young sandal
i see ENCB_DCEN is encoder B and ENCA_CDIN is encoder A right?
just dunno what index channel is
ionic edge
Yep
ionic edge
just dunno what index channel is
It's separate channel with 1 pulse per revolution in case of rotational ones and it can be used eg for homing if your device operates in rotational domain
young sandal
ah ok
oh
index is DC0 apparently
you also mentioned a relay when splitting the signal from the encoder? could you please elaborate on that
ionic edge
Now, encoders can have channels and also inverted (with – above the letter), inverted channels are worth it as then you can do differential pair on the wiring and limit the noise
ionic edge
you also mentioned a relay when splitting the signal from the encoder? could you...
Buffer if without inverted channels, comparator if with inverted channels
young sandal
ah, is that dependent on the encoder i choose?
ionic edge
You can use buffer with one with inverted channels but it defeats the point of using inverted channels
young sandal
what would be a good omron encoder for this application? if price was not important
and what would go into choosing one
ionic edge
Check if TMC is limiting pulse rate or CPR first
young sandal
what does “software controlled” correction of step loss mean
meaning set in the register?
ionic edge
Hmm, I'd have to read into it, alternative version is to use cheapo encoder and just test
young sandal
i think im getting ahead of myself
dc step is a “mode” or communication interface?
also found this in regards to the encoder poll rate
125us
so hears what i understand
comparator makes signal integrity better if i use inverted channels, but if the 5160 expects uninverted i need to use an inverter?
or repeater if normal channels for good signal integrity?
so i can use a repeater/comparator before i go into the wago and before it goes into the chip?
oh no i get it now
comparator converts differential signals into single ended signals, repeater amplifies signal
so i need a comparator on A1 and A2 to make A going into 5160, and B1 and B2 going into a comparator to make B
Index is just its own thing snd i need a repeater
seems like a job for a pcb!
sad waterrot didnt hook up these and i needa solder directly to the chip
ionic edge
Index is just its own thing snd i need a repeater
Index isn't needed at all according to datasheet screenshots
young sandal
would it be better with?
ionic edge
But yes, you use differential pair with inverted channel to have better signal integrity and right next to the input of the chip you put a comparator to filter it and make it into single signal
ionic edge
would it be better with?
Nah
ionic edge
ive landed on omron encoder e6b2-cwz6c 1800p/r
What's the max speed you're expecting on the printer?
young sandal
What's the max speed you're expecting on the printer?
1.5m/s around
but fast accelerations
its under its max of 100r/s
ionic edge
Ok, so 1500mm/s with 32mm/rev gives ~49rev/s, with 1800 p/r you get 88200 Hz, so whole cycle is 11.3us, pulse is 5.6us
ionic edge
Then 7us pulse length
young sandal
yeah
ionic edge
That's a bit lower than 125us
young sandal
too fast?
young sandal
so i need a rotary encoder with less p/r?
wont is be less accurate?
around 100p/r
for ~125
ionic edge
Find a section that specifically tells about encoder timing as the one you sent with 125us pulse is for STEP in DcStep
ionic edge
8MHz gives you 0.125us so you'd be fine with 1800 ppr
young sandal
8MHz gives you 0.125us so you'd be fine with 1800 ppr
the clock speed being 8MHz?
thats why its 0.125us
ionic edge
No, 8MHz encoder counting frequency
young sandal
how would i find that
ionic edge
how would i find that
It means that as long as your encoder output frequency is below 8 MHz it'll manage
And encoder output frequency is ppr × rev/s
ionic edge
1800 is perfectly fine as with 37.5 rev/s you get 0.068 MHz which is lower than 8 MHz
young sandal
ah
even with 60 rev/s its still far under 8MHz right? so it should be fine
in that case more resolution more better?
i can get upto 2500p/s
also, do you know of a repeater pcb that also splits into 2 outputs
ionic edge
I'm not sure if this style encoder is the way to go
young sandal
what would you suggest?
ionic edge
Some magnetic ones get output with AB, have 12+ bit resolution and are cheaper, mostly Osram AMS
young sandal
price doesnt really matter. Would they be more accurate?
Also why would you recommend them over this
1800 is perfectly fine for me, its already a 8th microstep
ionic edge
Also why would you recommend them over this
Easier to use, they don't wear out as fast
young sandal
how much longer would they last over an optical encoder like this?
ionic edge
Optical ones usually are limited by bearing lifespan and this depends on how well you mount them
young sandal
would i expect more than 10k hours on this one if mounted correctly?
ionic edge
would i expect more than 10k hours on this one if mounted correctly?
This rating is usually specified in millions of rotations
You'd need a longer shaft or double sided shaft to mount though
young sandal
im planning on using a coupler
ionic edge
Which coupler?
young sandal
Which coupler?
undecided yet
ionic edge
minimum life expectancy
Yeah, these are usually used in different scenarios
young sandal
whats the one they slap at the back of steppers? or in servos?
ionic edge
undecided yet
Couplers introduce kinematic inaccuracies which means that if you use wrong ones or use them wrong, you'll introduce errors
young sandal
ah ok
ionic edge
whats the one they slap at the back of steppers? or in servos?
Ones I know are 2 part ones so they have codewheel attached directly to the shaft and encoder unit attached to the body, thanks to this they don't require coupler, no external hardware, bearing lifespan is lifespan of the servo bearings and these are often beefy
young sandal
hrmm interesting
also note i plan to not put these encoders directly on the motors
they will be belt driven
ionic edge
Oh no
young sandal
they are for rough positioning 
ionic edge
1800 PPR isn't rough positioning
ionic edge
Then why would you plan on belt driving them?
crisp fiber
Why do you want to do this ?
ionic edge
Worry not, you'll need machined parts to mount it correctly as you won't hit this spec with prints
young sandal
Why do you want to do this ?
to close the loop at high speeds, also just for fun
apparently 5160s can support rotary encoders so i wanted to ask about it since i already had some
crisp fiber
to close the loop at high speeds, also just for fun
But why do you need a closed loop? Your printer shouldn’t be dropping steps
It’s generally a myth that closed loop is better for 3D printers
young sandal
for a certain fast boat race
ionic edge
for a certain fast boat race <:vaulthah:597787241812852736>
Are encoders allowed these? It kinda makes it into servo and these are no good
young sandal
rules state “must be stepper powered” no limits on closed loop
it is powered by stepper motors
crisp fiber
I still don’t know why you need a closed loop stepper - it’ll either step or it won’t. Closing the loop won’t help
undone fossilBOT
young sandal
I still don’t know why you need a closed loop stepper - it’ll either step or it ...
if it loses steps you can detect that and move it back, in most benchy attemps it happens only once or twice during the run
when the rotor lags behind
the way dcstep works is it automatically finds the highest speed it can step back at without skipping and steps back to the desired position
the reason i set the limit at 1.5m/s is because thats the point where the phase resistance and back emf overcomes 48v and no longer makes torque (with my motors)
the way the steppers step is they appromimate a sine wave to get a motion
sometimes when the wave is off or there is a sudden back emf spike it will be too much to overcome and the motor will “skip”
see this simulation
when the peaks get to low the motor will lose steps, thats why i want to put a rotary encoder in there t
thats why the motor wont lose steps 100% of the time and just some of the time
