The reason LV8729 are recomended for the extruder is quite simple: Higher microsteps means less torqe per microstep. As an extruder needs much force to actualy move the filament, on 256 microstepping you might get inconsistent extrusion (the well known issue 602 on prusas mk3). The difference between skipping steps and "skipping" microsteps due to low torque is: The torque of microsteps "accumulates". So on full step the full movement (normaly 1.8°) will be fullfilled. But on the 127/256 is there realy a movement 127/256*1.8°? Maybe not. Maybe the the 125/256*1.8° movement has succeded, but the next sucessful microstep is 131/256*1.8° (just an example). So the goal of using an LV8729 for the extruder is simply to find the microstepping with enough torque to actualy move the filament reliable for a constant extrusion, there every microstep succed in moving the filament. This might be 1/32 or 1/64 microstepping - but mostly not 1/128 and for shure not 1/256 microstepping like an TMC uses. Also the TMC only uses 1/16 as input and interpolates - and the interpolation might be wrong speed on the beginning and the end of an extruded line. As on the other side the extruder should use as many microsteps as reliable possible - you want a "smooth" movement of the extruder motor, no "pulses" of pressure. The movements of the gear(s) moving the fillament are very slow while extruding (as long you do not use huge nozzle diameters). So using 1/64 "native" (=not interpolated) microsteps are no problem even on 8-bit boards. So tuning in a LV8729 needs more testing. You have to find the optimal microstepping for not "loosing" microsteps to avoid incosistent extrusion, but also moving smooth. The option to do 16,32 and 64 microstepping give more options than a driver fixed to 16 microsteps or fixed doing 256 interpolation. Of course nowerdays this is mostly theoretical: Using a Titan or a Bondtech BMG with gear reduction from 3:1 you already reduce the possible inconsistance by a factor of 3; making it mostly invisible (as the molten plastic in the nozzle works as damper smoothing out partly the inconsitancy - on a bowden system also the filament in the bowden smooths out inconsistencies, so the issue is mostly relevant to direct extruders - the reason it got prominent at the mk3). On a remote direct drive (Nimble, flex3drive) with gear ratio of 30:1 / 40:1 its completly irrelevant (use TMC there, because extruder motor moves fast so 256 interpolated microstepping is usefull to reduce noise). So in summery: On a direct extruder with TMC and inconsistent extrusion you might try a LV8729 with 16, 32 or 64 microstepping. This might solve the problem. If you do not have an inconsistent extrusion issue, you do not need to solve it :-) so stay to the drivers you already have. But with an ungeared extruder having the issue, also check to switch to a geared extruder (like the Bondtech BMG/ Bondtech mk3 upgrade). It might be the only solution (without SMD soldering) in some situations anyway, as if an original Prusa mk3 is effected the TMC at the einsy board are soldered... Also you can reduce weight this way (pencake stepper) which is always nice (even if the bed on i3 style limits the speed, less weight is good for reducing stress on the mechanical parts). One comment for boards with integrated drivers: Even if you have a board with soldered drivers, you mostly can change to other drivers with only a little soldering skills. Look at the driver and track down the traces on the PCB to a point there you can easy solder wires for Enable, Dir and Step. I did that on the creality board - and it works fine. Mostly changing x and y is all you need anyway, so 2 boards for adding a stepper driver (de.aliexpress.com/item/3D-drucker-A4988-DRV8825-stepper-motor-drive-control-board-expansion-board-F-r-arduino/32922803521.html), two TMC2208 and some wires are all you need to make a creality board quiet. It is not mandatory to deactivate the soldered drivers - if there is no motor connected they will not disturb anything. While on a Creality board due to low flash and no free I/O a bigger board with stepper drivers in sockets is a little more expansive but give you other benefits, so track down traces might be a bit to big an effort, there are other printers (like the Anycubic Predator) with closed source 32 bit board and touch screen but loud soldered A4988, there tracking traces and soldering wires to external modules might be very profitable - changing the board would not only costs the price for a new board, but also problems finding the right settings (as you can not take a look to the original firmware) and also changing the touch display. Actually I ordered the Predator yesterday and tracking down Step/Dir/En will be one of the first things to do after assembling and doing some testprints (there of course all 4 drivers, as it has 3 fast moving axes and the extruder will also get fast as soon as the nimble arrives)
One additional reason to use LV8729 instead of the TMCs is that there are compatibility issues with Stealthchop mode when using linear advance feature in Marlin.
I must have missed the fine print that reads "1/256 microstepping must be used with tmc drivers " or that microstepping MUST be interpolated to 1/256. Either that or that is a completely specious argument. Perhaps it was next to the bit that says you must also always use steathchop thereby hitting incompatibility issues with linear advance. #define INTERPOLATE false
@@m3chanist To set a standalone TMC to 16 microsteps you have to put cfg1 and cfg2 to Vcc. As we are talking about standalone driver modules as replacement you can not do this by jumpers on most boards, only by soldering - as you can only set GND or "open" but the cfg pins on TMC are tristate... And yes, if you do so (or control a TMC via SPI or serial) you can use 1/16 microstepping without interpolation. But as you have to do soldering (or add spi/serial controll) it is not a simple drop-in replacement. This video series is about changing the drivers on a board with the common sockets, using ready to use steper modules; not about what you can do with self modified modules. Sorry, have to correct: You are right as RAMPS design are jumpers to Vcc, not to GND. So you can use TMC 1/16 microstepping without interpolation simply by using the jumpers; without soldering. Only if TMC are soldered to the board but used standalone (like on Creality's silent board) you might have a problem (as there are no jumpers to change the config). So the main reason for not choosing TMC on E is the price. But as sets of 4 TMC normaly are the same price or even less expansive than 3 single bought TMC + 1 other driver, it seems mostly irrelevant.
Here is the bug report on TMC Stealthchop incompatibility with the Linear Advance in Marlin (and other firmwares as well): github.com/MarlinFirmware/Marlin/issues/11825
I did have skipped steps on the extruder with the 2208s in stealthchop2. I went 2208's in stealthchop2 for XYZ and 2208 with the One Time programable spreadcycle for E . Since I did this upgrade I feel deeper in love with my Ender 3. Thanks for the great videos!!!
Hi Michael! Thanks for all your great instructional videos! I am in the midst of doing a MKS Gen L with TMC2130s. Doing the loom for SPI was daunting (great difficulty with the Dupont connectors seating in the jackets, and wires coming dislodged when trying to force them) but your guides have made the process easier for sure. Thanks!
Watched your stepper drivers videos I upgrade my ender3 with the mks gen L board with tmc2100, tft 3.2 inch touch screen, and dual extruders I followed your vids to a tee and had flawless performance.
the loud sound you hear with LV8729 is just the firmware going into quadstep mode because the MCU can't keep up, thus going down to 1/16 and maybe skipping some steps LV8729 are meant for 32 bit controllers. 32 bit + LV8729 is a much better solution than 8 bit + TMCs
There are two versions of LV8729 drivers, each of which requires a different vref. That's probably why there were skipping problems with them in the video. If the bottom of the driver has R100 resistors, then, as in the video, the vref is (A/2). However, if there are R220 resistors then vref is (A*1.1). There's a very good chance you actually needed 0.83V for XYZ and 0.99V for E.
Always send M502 (load default values = the values from fw) followed by M500 (save values to eeprom) after flashing fw. Just to make sure to load the new values and save them to eeprom to make everything in sync. You will not need to do manual changes like you do in the video (sending M92...). The eeprom settings will be used before the settings in the fw (and flashing will not change eeprom) so making them the same after config changes could remove some unnecessary headache
I'm running Lerdge branded LV 8729's that have the 220 ohm resistors and not the 100 ohm ones on them, just like 4988's there is a different VRef calculation depending on the resistors these drivers have, and I have them in 2 of my printers, one on a mega/ramps 1.5 that i shoe horned in a cr10 and the other I just recently fitted to a RE ARM /ramps 1.5 in my 'franken kossel' which in the kossel are running @ 64 steps Vref is @ .5v and Marlin 2.0 bugfix branch, I just put these drivers in the to check the ramps & and my then untested ReArm combo was ok before risking the 2130's in it, but I was in for a surprise, they worked perfect right off the bat, I get no skipped steps or layer shifts or that noise you seem to be getting here, the unexpected bonus was the motors are running cool, even during an overnight print the motors were only just barely warm to the touch, even at 120mms, have also run them @ 160 mms and there is no pausing in the print to indicate a gcode processing issue but at that speed the kossel's arms hammer way too hard for my liking, I previously had 2130's on the trigorilla board in the kossel with hardware SPI and seemed to be not much quieter, but when the 2130's were fitted had to drop the acceleration down quite a bit as the 2130's would layer shift, I had suspected that a 24v conversion would fix this as I dont get this problem with my ender 3 with 2130's in it, these 8729 drivers work so well on my kossel that I'm not putting the 2130's back in it, as the heated bed heats up quick enough anyway it will stay at 12v, on the my cr10 these 8729's are definitely noisier but they are running at 32 steps and the sound a bit whiney, I reckon 32 steps @ sensible print speeds on a printer like the CR10 with an 8 bit board seem to be the limit, one thing I did notice is the motors are warmer on the cr10 @ 32 steps than on the kossel at 64, maybe its the load? I suspect if you try these @ 64 steps on a 32 bit board your going to get a much quieter result, the thing to think about is, with a board like the Re ARm when you have fewer pins to play with I reckon these drivers are a good choice, not as quiet as my ender 3 with its 2130's but plenty quiet enough
Great video, thanks for all the information you provided! It would be great if you could conpare other drivers too. St820, Toshiba tb67s109, and other not-so-well known alternatives
FWIW - in my experience, the TMC2130s are fine in extruders. I run them exclusively, the only issue is they use the first-gen Interpolation algorithm which means they 1) could be up to a single full step behind the other axes at any point with it enabled, and 2) can in some rare cases miss steps on reversing. Extruders reverse A LOT in normal operation between presssure/linear advance and extracts, so this is probably where some people run into issues. YMMV, but I run with it (Interpolation) disabled on extruders.
Marlin has a way of dealing with insufficient step routine interrupt rate, it can actually send multiple STEP pulses within a single invocation. So on the 8-bit hardware, for every move section, Marlin chooses an interrupt rate such, that it's below 10 KHz, but it'll multi blip the drivers as needed; but the driver might end up being confused and doing something not-so-nice. They actually have some timing tuning for the feature, so you should specify the type of driver you're using in the config! On a typical printer with 16 microsteps and less than 200mm/s movement speed, you can assume that these double and quad stepping modes will not be activated.
Do you have any videos on flashing the latest version of marlin to a MKS Gen L board, I can't find any that show how to do it comprehensivly, and I know you go into great detail, and are very informative. Great video, and thanks!
Hi, congratulations for the always very precise and effective explanation, I would like to make an upgrade on my sidewinder x1 it is possible to replace the original drivers with LV8729 both for the axis motors and for the extruder and what are the best settings to change the firmware . Maybe you can make a dedicated video. thanks
Thirded!. Not that long ago I upgraded to an MKS-GEN-L and TMC2130 drivers after watching Michael's tutorials......and now we have what looks like an even better option of the SKR 1.3 and TMC2208! Doh!! I might just upgrade the mainboard as they look to be stupidly cheap, and keep the 2130 drivers. I'll wait and see if Michael does this first though - keen to see what benefits, if any in practice, there are to having the 32 bit board.
5 ปีที่แล้ว +9
So a 32-bit board like the SKR V1.3 should be able to handle 1/128 stepping without any issues? How would that compare to stepping interpolation on TMC2208?
I just put TMC2100’s on an MKS Gen L board on my Ender 3. I’m having some issues with homing, which are think are due to missing numbers for bed coordinates in Marlin, though I still haven’t tracked the issue. I was having lots of trouble with missed extruder steps on the original Melli board, which prompted me to make the change. Now I’m hoping to fix the homing issue, so I can continue to enjoy the Ender 3.
Great roundup Michael. As you know, I've been pondering which board and drivers to upgrade the Borg with. At the moment I think I'm going to go with the TMC 2208's and since I'd like to have a go with 32bit, it might be worth having a go with the SKR 1.3 board, do you have any plans to test the SKR 1.3? I'd love to hear your views on it first.
Great video Michael. So when you doing the Bigtreetech SKR 1.3 upgrade? lol. I've just upgraded my Anet E10 with SKR 1.3 board and Malin 2.0. Running TMC 2208s on the X and Y axis. It's so much quieter. I was hoping to run them on my dual Z axis also but 1 of the drivers is faulty so I'm running A4988s on the dual Z and extruder. Do you think I'd be ok running a TMNC2208 and a A4988 on my Z axis?
I just changed from TMC2208 to LV8729 for my extruder because i had constant issues with TMC2208 in stealthchop mode for the extruder. Changing the TMC2208 to another mode needs soldering and my soldering skills need training bevor I would try that. No my ender 3 runs perfectly with an MKS Gen 1.4 and three TMC2208 for x/y/z and one LV8729 for the extruder.
Why did you disconnect the USB cable? On the Gen-L the logic voltage is 5v either way, and the vref is a current mirror connected to the logic voltage, not the motor voltage. And as someone else said, the calculation depends on the resistor value which will vary from vendor to vendor for the lv8729. And on a 32-bit board the logic voltage will be 3.3v so the vref value will be different, and if the value of the resistor is too high you won't be able to set enough current. As others have said, the 8-bit controller can't really do the higher microstepping modes so the firmware goes into a mode where it sends several pulses at once, essentially reverting to 1/4 step mode.
Fyi, issue with 2208 in stealthchop on an extruder is with linear advance on low step/mm extruders. A tmc bug will put the driver in shutdown. Geared extruders (over 280 steps/mm in my testing) avoid the bug as does spreadcyle.
Just setting up an Anet A8 with the 2208's in UART mode on a Skr 1.3. all vrefs can be done via gcode in Pronterface also no ugly extra wiring to do. just solder the jumpers on the 2208 board. The board will also run the MKS tft32 as well as the std ones like Reprap full screen ect. Board in UK cost around £20 and it's 32 bit
I have tested ST820, it's strong, smooth ultra silent in low speed, but in fast travel modes it's little noisy. S109 is really strong, smooth but noisy driver, good for extruder or CNC. I have felt LV8729 in 1/16 small jumps, It's not so smoth, I don't want it even on extruder. After all this experiments I'm back using TMC2208, but in UART mode in spreadcycle and I'm looking for TMC2209.
Please forgive if you have already covered this but I am replacing my Ender 3 V2 board with the BTT Octopus Pro 1.3 board. However, while I think I want the TMC2209 drivers, I am not sure how to know if that I can just choose the driver I want or if there is something I need to match on the printer. I hope that I suscenct stated my issue. If not, please forgive me. Thanx
Have you ever considered trying out the A4988 'fix' by shorting the resistor next to the cooling element? If you're interested I can give you some background information. What this does, it changing the decay setting on the A4988. I would really like to see your comparison since you use a Decibel meter.
@@KayakingVince Here is more info: reprap.org/forum/read.php?1,465761,851804 Sadly TeachingTech never replies to any comments. Would be worthwhile to try this out with a good decibel meter and vibration setup.
I was trying to get eh LV8729 to work on teh E0 on my SKR 1.4 Turbo. I get zero movement out of the steppers. Anyone have an idea what could cause the issue? All Marlin 2.0 the TMC 2209's worked fine. I disabled stealthchop in the firmware for E and set the driver in the config file to LV8729. Still no movement.
Awesome work as usual! I went with the 2208s and mks gen l. But I still need to install the card reader. The difference is worth it, it makes much less noise. Thank you
Hey. Great video. Could you please tell me (if you know) wich is the best driver to replace in the Artillery sidewinder x1 without changing the firmware?
Thanks to your videos, I think I'm going with 2208, printer is in the room and 4498 are really loud, and on the other side, not really into sensorless homing. Already have MKS board and screen, but ordered wrong drivers...
I am running X/Y with TMC2130, extruder and Z with A4988. I have spare TMC2130 and DRV8825 one piece each, should I change my extruder and Z with them and which go to which ?
I've gone for the 2208s and MKS Gen L combo in my Ender 5. Going for silence, so the PSU is getting replaced along with moving it and the Gen L outside of the bottom enclosure. Big quiet fans and a new hotend cooling assembly with quieter fans there should do the job I hope.
@@akselanthonisenribe9984 The TMC5160 and TMC2209 have some slight differences. The TMC5160 has: SPI as it's communication protocol which is slightly faster than the TMC2209's UART dcStep - load dependent speed control The TMC2209 has stallGuard4 while the TMC5160 just has stallGuard2 Besides that, they're mostly the same.
After watch this perfect tutorial and changed my Lv8729 to 64 micro steps I hit a strange but. When doing retraction the extruder pulls back the filament. The next thing in the gcode is to return the fillament back, but the extruder decides to pull back one more time (second retraction) Using the same slicer as before and the safe firmware (just updated the esteps) Marlin 1.1.9. Any ideas?
Hi Micheal. Another great video, as usual! I just noticed one thing that I missed in previous videos, you changed the hotend fan from Noctua? Any specific reason for this?
@@deceitive3338 definitly right. Filament jam mostly leads to grinding down the filament, not to skipped steps (on a good extruder). Maybe a dust detector for the produced dust may detect it ;-) But seriously: Prusas old mk3 sensor was a good solution (laser sensor). Sadly they changed it as it got 2 problems: a) Reliability. Prusa uses black PET-G for printing the printhead. But PET-G is highly reflecting, even if you use black PET-G. I think they did the prototypes of the mk3 in ABS (like the mk2 was printed in ABS), which is less reflective. So they get it never reliable. I solved this issue by applying black non-glossy paint inside the detection chamber, never ever having false results. b) transparent high reflective filament (PET-G transparent or PC). Ever moved a mouse on a glas table? Can not work. Simple solution: run the filament though some gears first, leaving marks on it. Like sanding the glasstable if you want to use the mouse directly: Not elegant but solves the problem.
Did the 2208 upgrade little over a month ago. Much quieter now but didn't really help with some interesting vertical patterns or zebra striping Issues I have. Wondering if my table is too wobbly or is I should turn up the v ref on one or two of my drivers. What do you think? Thanks as always for your effort and excellent info.
@@deceitive3338 So I used the formula from his previous video and the stat sheet from the drivers I bought. Id have to check again to know exactly. I can see the layers, assume that's normal, but I seem to have vertical stripes on the walls of some prints. Not as noticeable on others and I know they aren't seems.
I have all LV8729s at 1/32 and it is quieter but fast X travel makes the same ringing noise as seen in your 1/64 example. Might just get 1 TMC driver for X
check out squash ball feet i put a set on my cr-10 with stepper motor dampers(had them of for a while now) and my printer is almost silent with the a4988 drivers.
Hi, just curious, I am using the Bigtreetech skr v1.3 board on my CR-10S, along with the LV8729 drivers. So my question is this, where do I locate the vref values for the 10S.
I'm getting an BTT SKR 1.4 Turbo for my Tarantula Pro, I want to put in updated drivers, but I need it to be somewhat easy, I'd like to just drop them in. Is 2209s easier than 2208s? How about the higher series with no pot adj, SPI? Any suggestions? Thx
Hi, I would like to make an upgrade on my sidewinder x1 it is possible to replace the original drivers with LV8729 for both the axis motors and the extruder and what are the best settings to change the firmware. Maybe you can make a dedicated video. thanks
this is marlin v1.1 ? if yes use v1.1.9 good ? i see marlin 1.0 not show #define X_DRIVER_TYPE in configuration.h LV8729 use with Marlin 1.0 ? i use marlin 1.0
Hi, could You by chance do a Video on howto do wiring of TMC2208 (UART Mode) with an MKS SGen 1.0 Board? Seems there are some Videos with the SBase but the SGen is somewhat different.
I woudl agree that 2100 / 2130's shouldn't be used on extruders. 2208's shouldn't really be used in bowden or geared, however if you have done the direct drive mod and your retracts are relatively short (1mm) and the speeds are relatively slow (30mm/sec) it seems to work fine and be silent. Switching to a smaller hobb can help as well.
I have a question could you do the same stepper review but with the smoothers as i have found that adding the smoothers does make things considerable quiter, i have the pmaxilion 320 printer and using the a4988 steppers and after adding the smoothers its is about 80% quiter
About the Vref on the LV8729, I think you got the R220 resistor drivers so the right formula would be: Vref = I x 1,1 as far as I know this is how it should be measured: R100: Vref = I x 0,5 R220: Vref = I x 1,1 Can you confirm this?
My skr v1.3 will be coming soon. I ordered 2208's and to be sure a lv8729 for the extruder. Why do people advice 8729? Do they have more torque? And can I run them 1/128 with the 32 bit board?
Because someone had a problem and didn't know how to diagnose what was happening, ignorance basically. They then told a friend who then told a friend who then told a friend, and so a big bright new baby bs fact was born. And all those who didn't know any better swallowed it as "good advice" and passed it on knowingly. In other words, the standard method.
The reason none of your settings take effect after uploading marlin is because you still have settings stored in EEPROM. You need to do an M502 to reinitialize eeprom with values from firmware, then all your settings you set in firmware will be in place. Then an M500 to store in EEPROM and you're good to go.
on a skr 1.3 i tried both tmc2209 and lv8729. put the lv8729 @ 1/64 and they crush the tmcs. they are both quieter AND have more torque AND don't need fiddling with stealthchop/spreadcycle/hybrid modes and the like
So LV8729 relies strictly on the micro controller for creating steps and risks skipped steps at really high accelerations, but how do TMC stepper drivers work with 8-bit because they take 1/16 micro stepping pulses and interpolate internally 1/256 micro steps and take less computing resources from the 8-bit controller I hope I am understanding this stuff right because there is a lot of here say and snake oil on the internet
Hey! I think the website that you used as reference got the Vref formula wrong. In the official LV8729 datasheet (www.mouser.com/datasheet/2/308/LV8729V-D-278452.pdf) the formula is Iout = (Vref/5)/R. The formula that you used was propably for DRV8825 driver. That is propably the reason why you have experienced some skipped steps before you upped the Vrefs. Just wanted to let everyone know to save some troubleshooting time :)
Per his own words in the video, The 8 bit board had a hard enough time processing the 1/64. He encountered skipping on the fast X movement. It would be presumed that 1/128 would be even worse for skipping steps.
13:20 "Challenge: Build a printer that uses four different drivers" Wouldn't surprise me if a Chinese manufacturer has already completed your challenge through sheer incompetence.
Usually those came with a Melzi-like board with A4988 soldered in. The only pinout compatible chip is maybe Hero HR4988, which supports higher microstepping modes (configurable). At 16 microsteps, it's a little less annoying sounding than A4988, but i don't know about its behaviour quirks. Straps for 16 microsteps are the same between A4988 and HR4988, but for all other modes they differ.
The reason LV8729 are recomended for the extruder is quite simple: Higher microsteps means less torqe per microstep. As an extruder needs much force to actualy move the filament, on 256 microstepping you might get inconsistent extrusion (the well known issue 602 on prusas mk3). The difference between skipping steps and "skipping" microsteps due to low torque is: The torque of microsteps "accumulates". So on full step the full movement (normaly 1.8°) will be fullfilled. But on the 127/256 is there realy a movement 127/256*1.8°? Maybe not. Maybe the the 125/256*1.8° movement has succeded, but the next sucessful microstep is 131/256*1.8° (just an example).
So the goal of using an LV8729 for the extruder is simply to find the microstepping with enough torque to actualy move the filament reliable for a constant extrusion, there every microstep succed in moving the filament. This might be 1/32 or 1/64 microstepping - but mostly not 1/128 and for shure not 1/256 microstepping like an TMC uses.
Also the TMC only uses 1/16 as input and interpolates - and the interpolation might be wrong speed on the beginning and the end of an extruded line.
As on the other side the extruder should use as many microsteps as reliable possible - you want a "smooth" movement of the extruder motor, no "pulses" of pressure. The movements of the gear(s) moving the fillament are very slow while extruding (as long you do not use huge nozzle diameters). So using 1/64 "native" (=not interpolated) microsteps are no problem even on 8-bit boards.
So tuning in a LV8729 needs more testing. You have to find the optimal microstepping for not "loosing" microsteps to avoid incosistent extrusion, but also moving smooth. The option to do 16,32 and 64 microstepping give more options than a driver fixed to 16 microsteps or fixed doing 256 interpolation.
Of course nowerdays this is mostly theoretical: Using a Titan or a Bondtech BMG with gear reduction from 3:1 you already reduce the possible inconsistance by a factor of 3; making it mostly invisible (as the molten plastic in the nozzle works as damper smoothing out partly the inconsitancy - on a bowden system also the filament in the bowden smooths out inconsistencies, so the issue is mostly relevant to direct extruders - the reason it got prominent at the mk3).
On a remote direct drive (Nimble, flex3drive) with gear ratio of 30:1 / 40:1 its completly irrelevant (use TMC there, because extruder motor moves fast so 256 interpolated microstepping is usefull to reduce noise).
So in summery:
On a direct extruder with TMC and inconsistent extrusion you might try a LV8729 with 16, 32 or 64 microstepping. This might solve the problem. If you do not have an inconsistent extrusion issue, you do not need to solve it :-) so stay to the drivers you already have. But with an ungeared extruder having the issue, also check to switch to a geared extruder (like the Bondtech BMG/ Bondtech mk3 upgrade). It might be the only solution (without SMD soldering) in some situations anyway, as if an original Prusa mk3 is effected the TMC at the einsy board are soldered...
Also you can reduce weight this way (pencake stepper) which is always nice (even if the bed on i3 style limits the speed, less weight is good for reducing stress on the mechanical parts).
One comment for boards with integrated drivers: Even if you have a board with soldered drivers, you mostly can change to other drivers with only a little soldering skills. Look at the driver and track down the traces on the PCB to a point there you can easy solder wires for Enable, Dir and Step. I did that on the creality board - and it works fine. Mostly changing x and y is all you need anyway, so 2 boards for adding a stepper driver (de.aliexpress.com/item/3D-drucker-A4988-DRV8825-stepper-motor-drive-control-board-expansion-board-F-r-arduino/32922803521.html), two TMC2208 and some wires are all you need to make a creality board quiet. It is not mandatory to deactivate the soldered drivers - if there is no motor connected they will not disturb anything.
While on a Creality board due to low flash and no free I/O a bigger board with stepper drivers in sockets is a little more expansive but give you other benefits, so track down traces might be a bit to big an effort, there are other printers (like the Anycubic Predator) with closed source 32 bit board and touch screen but loud soldered A4988, there tracking traces and soldering wires to external modules might be very profitable - changing the board would not only costs the price for a new board, but also problems finding the right settings (as you can not take a look to the original firmware) and also changing the touch display. Actually I ordered the Predator yesterday and tracking down Step/Dir/En will be one of the first things to do after assembling and doing some testprints (there of course all 4 drivers, as it has 3 fast moving axes and the extruder will also get fast as soon as the nimble arrives)
spot on, this is why some folks leave 4988's on the extruder
One additional reason to use LV8729 instead of the TMCs is that there are compatibility issues with Stealthchop mode when using linear advance feature in Marlin.
I must have missed the fine print that reads "1/256 microstepping must be used with tmc drivers " or that microstepping MUST be interpolated to 1/256. Either that or that is a completely specious argument. Perhaps it was next to the bit that says you must also always use steathchop thereby hitting incompatibility issues with linear advance.
#define INTERPOLATE false
@@m3chanist To set a standalone TMC to 16 microsteps you have to put cfg1 and cfg2 to Vcc. As we are talking about standalone driver modules as replacement you can not do this by jumpers on most boards, only by soldering - as you can only set GND or "open" but the cfg pins on TMC are tristate... And yes, if you do so (or control a TMC via SPI or serial) you can use 1/16 microstepping without interpolation. But as you have to do soldering (or add spi/serial controll) it is not a simple drop-in replacement.
This video series is about changing the drivers on a board with the common sockets, using ready to use steper modules; not about what you can do with self modified modules.
Sorry, have to correct: You are right as RAMPS design are jumpers to Vcc, not to GND. So you can use TMC 1/16 microstepping without interpolation simply by using the jumpers; without soldering. Only if TMC are soldered to the board but used standalone (like on Creality's silent board) you might have a problem (as there are no jumpers to change the config).
So the main reason for not choosing TMC on E is the price. But as sets of 4 TMC normaly are the same price or even less expansive than 3 single bought TMC + 1 other driver, it seems mostly irrelevant.
Here is the bug report on TMC Stealthchop incompatibility with the Linear Advance in Marlin (and other firmwares as well): github.com/MarlinFirmware/Marlin/issues/11825
I did have skipped steps on the extruder with the 2208s in stealthchop2. I went 2208's in stealthchop2 for XYZ and 2208 with the One Time programable spreadcycle for E . Since I did this upgrade I feel deeper in love with my Ender 3. Thanks for the great videos!!!
Hi Michael! Thanks for all your great instructional videos! I am in the midst of doing a MKS Gen L with TMC2130s. Doing the loom for SPI was daunting (great difficulty with the Dupont connectors seating in the jackets, and wires coming dislodged when trying to force them) but your guides have made the process easier for sure. Thanks!
Watched your stepper drivers videos I upgrade my ender3 with the mks gen L board with tmc2100, tft 3.2 inch touch screen, and dual extruders I followed your vids to a tee and had flawless performance.
I'm a new bee to 3D printing and very much like these technical videos.Thank you!
For other noobs.. cnc kitchen is also a good technical source.
the loud sound you hear with LV8729 is just the firmware going into quadstep mode because the MCU can't keep up, thus going down to 1/16 and maybe skipping some steps
LV8729 are meant for 32 bit controllers. 32 bit + LV8729 is a much better solution than 8 bit + TMCs
Great video, I'm looking forward to the TMC2209's, with StallGuard, sensorless homing while in StealthChop2 and CoolStep apparently! 👌
Hi Michael, thanks for the series. I use the tmc2208 also for the extruder. Works like a charm.
Michael, thank you. You are a valued contributor to our community.
Regarding the Vref. It's good to check the actual resistor on the driver. The Formula for my LV8729 was assuming 100 ohm but there was a 220 ohm.
This, i came to write this 👍
clearly... I see all this stepper motor driver and I just die... with you, all is perfect now! more understandable ! thanks!
I just upgraded my Ender 3 this evening to an MKS GEN L board with TMC 2100 stepper drivers. It's scary quiet now.
This was an incredibly useful and meticulous series. Thanks a lot !
There are two versions of LV8729 drivers, each of which requires a different vref. That's probably why there were skipping problems with them in the video. If the bottom of the driver has R100 resistors, then, as in the video, the vref is (A/2). However, if there are R220 resistors then vref is (A*1.1). There's a very good chance you actually needed 0.83V for XYZ and 0.99V for E.
Excellent summary Michael, thanks for all your testing!!
Always send M502 (load default values = the values from fw) followed by M500 (save values to eeprom) after flashing fw. Just to make sure to load the new values and save them to eeprom to make everything in sync. You will not need to do manual changes like you do in the video (sending M92...). The eeprom settings will be used before the settings in the fw (and flashing will not change eeprom) so making them the same after config changes could remove some unnecessary headache
I'm running Lerdge branded LV 8729's that have the 220 ohm resistors and not the 100 ohm ones on them, just like 4988's there is a different VRef calculation depending on the resistors these drivers have, and I have them in 2 of my printers, one on a mega/ramps 1.5 that i shoe horned in a cr10 and the other I just recently fitted to a RE ARM /ramps 1.5 in my 'franken kossel' which in the kossel are running @ 64 steps Vref is @ .5v and Marlin 2.0 bugfix branch, I just put these drivers in the to check the ramps & and my then untested ReArm combo was ok before risking the 2130's in it, but I was in for a surprise, they worked perfect right off the bat, I get no skipped steps or layer shifts or that noise you seem to be getting here, the unexpected bonus was the motors are running cool, even during an overnight print the motors were only just barely warm to the touch, even at 120mms, have also run them @ 160 mms and there is no pausing in the print to indicate a gcode processing issue but at that speed the kossel's arms hammer way too hard for my liking, I previously had 2130's on the trigorilla board in the kossel with hardware SPI and seemed to be not much quieter, but when the 2130's were fitted had to drop the acceleration down quite a bit as the 2130's would layer shift, I had suspected that a 24v conversion would fix this as I dont get this problem with my ender 3 with 2130's in it, these 8729 drivers work so well on my kossel that I'm not putting the 2130's back in it, as the heated bed heats up quick enough anyway it will stay at 12v, on the my cr10 these 8729's are definitely noisier but they are running at 32 steps and the sound a bit whiney, I reckon 32 steps @ sensible print speeds on a printer like the CR10 with an 8 bit board seem to be the limit, one thing I did notice is the motors are warmer on the cr10 @ 32 steps than on the kossel at 64, maybe its the load? I suspect if you try these @ 64 steps on a 32 bit board your going to get a much quieter result, the thing to think about is, with a board like the Re ARm when you have fewer pins to play with I reckon these drivers are a good choice, not as quiet as my ender 3 with its 2130's but plenty quiet enough
Great video, thanks for all the information you provided! It would be great if you could conpare other drivers too. St820, Toshiba tb67s109, and other not-so-well known alternatives
Congrats you have over 50,000 subscribers on your channel!
FWIW - in my experience, the TMC2130s are fine in extruders. I run them exclusively, the only issue is they use the first-gen Interpolation algorithm which means they 1) could be up to a single full step behind the other axes at any point with it enabled, and 2) can in some rare cases miss steps on reversing. Extruders reverse A LOT in normal operation between presssure/linear advance and extracts, so this is probably where some people run into issues. YMMV, but I run with it (Interpolation) disabled on extruders.
Marlin has a way of dealing with insufficient step routine interrupt rate, it can actually send multiple STEP pulses within a single invocation. So on the 8-bit hardware, for every move section, Marlin chooses an interrupt rate such, that it's below 10 KHz, but it'll multi blip the drivers as needed; but the driver might end up being confused and doing something not-so-nice. They actually have some timing tuning for the feature, so you should specify the type of driver you're using in the config!
On a typical printer with 16 microsteps and less than 200mm/s movement speed, you can assume that these double and quad stepping modes will not be activated.
Do you have any videos on flashing the latest version of marlin to a MKS Gen L board, I can't find any that show how to do it comprehensivly, and I know you go into great detail, and are very informative. Great video, and thanks!
I'm posting one to my channel this weekend. 3DPrintingTexan
adam french
th-cam.com/video/LNdMYgwez8Y/w-d-xo.html
For anyone else looking for 3d printing but finding gospel songs after clicking on Ben's username... th-cam.com/channels/0-8Fd9GL6En1i7hpzk93bQ.html
Hi, congratulations for the always very precise and effective explanation, I would like to make an upgrade on my sidewinder x1 it is possible to replace the original drivers with LV8729 both for the axis motors and for the extruder and what are the best settings to change the firmware . Maybe you can make a dedicated video. thanks
Would be keen to have your views on another popular combo. SKR 1.3 (32 bits with Marlin 2) and tmc2208 at least in uart mode. Considering this master?
I second that !
That's what I did with my Ender 3. Works well and you don't have to do any extra wiring.
Thirded!. Not that long ago I upgraded to an MKS-GEN-L and TMC2130 drivers after watching Michael's tutorials......and now we have what looks like an even better option of the SKR 1.3 and TMC2208! Doh!! I might just upgrade the mainboard as they look to be stupidly cheap, and keep the 2130 drivers. I'll wait and see if Michael does this first though - keen to see what benefits, if any in practice, there are to having the 32 bit board.
So a 32-bit board like the SKR V1.3 should be able to handle 1/128 stepping without any issues? How would that compare to stepping interpolation on TMC2208?
For statistics! Love that technical videos!
I just put TMC2100’s on an MKS Gen L board on my Ender 3. I’m having some issues with homing, which are think are due to missing numbers for bed coordinates in Marlin, though I still haven’t tracked the issue. I was having lots of trouble with missed extruder steps on the original Melli board, which prompted me to make the change. Now I’m hoping to fix the homing issue, so I can continue to enjoy the Ender 3.
Great roundup Michael. As you know, I've been pondering which board and drivers to upgrade the Borg with. At the moment I think I'm going to go with the TMC 2208's and since I'd like to have a go with 32bit, it might be worth having a go with the SKR 1.3 board, do you have any plans to test the SKR 1.3? I'd love to hear your views on it first.
Great video Michael. So when you doing the Bigtreetech SKR 1.3 upgrade? lol. I've just upgraded my Anet E10 with SKR 1.3 board and Malin 2.0. Running TMC 2208s on the X and Y axis. It's so much quieter. I was hoping to run them on my dual Z axis also but 1 of the drivers is faulty so I'm running A4988s on the dual Z and extruder. Do you think I'd be ok running a TMNC2208 and a A4988 on my Z axis?
I just changed from TMC2208 to LV8729 for my extruder because i had constant issues with TMC2208 in stealthchop mode for the extruder. Changing the TMC2208 to another mode needs soldering and my soldering skills need training bevor I would try that.
No my ender 3 runs perfectly with an MKS Gen 1.4 and three TMC2208 for x/y/z and one LV8729 for the extruder.
Excelent! Have same problem extrude 2208 and linear advance!
Why did you disconnect the USB cable? On the Gen-L the logic voltage is 5v either way, and the vref is a current mirror connected to the logic voltage, not the motor voltage. And as someone else said, the calculation depends on the resistor value which will vary from vendor to vendor for the lv8729. And on a 32-bit board the logic voltage will be 3.3v so the vref value will be different, and if the value of the resistor is too high you won't be able to set enough current.
As others have said, the 8-bit controller can't really do the higher microstepping modes so the firmware goes into a mode where it sends several pulses at once, essentially reverting to 1/4 step mode.
Great video! Any advice for the jumper pins on the MKS Sgen L v1.0 board?
Fyi, issue with 2208 in stealthchop on an extruder is with linear advance on low step/mm extruders. A tmc bug will put the driver in shutdown. Geared extruders (over 280 steps/mm in my testing) avoid the bug as does spreadcyle.
AT2100 Stepstick Stepper Motor Driver Module Automatic Semi-Current Locking Motor Driver Board for 3D Printer?
As always a great video. Thank you.
Just setting up an Anet A8 with the 2208's in UART mode on a Skr 1.3. all vrefs can be done via gcode in Pronterface also no ugly extra wiring to do. just solder the jumpers on the 2208 board. The board will also run the MKS tft32 as well as the std ones like Reprap full screen ect. Board in UK cost around £20 and it's 32 bit
I have tested ST820, it's strong, smooth ultra silent in low speed, but in fast travel modes it's little noisy. S109 is really strong, smooth but noisy driver, good for extruder or CNC. I have felt LV8729 in 1/16 small jumps, It's not so smoth, I don't want it even on extruder. After all this experiments I'm back using TMC2208, but in UART mode in spreadcycle and I'm looking for TMC2209.
Please forgive if you have already covered this but I am replacing my Ender 3 V2 board with the BTT Octopus Pro 1.3 board. However, while I think I want the TMC2209 drivers, I am not sure how to know if that I can just choose the driver I want or if there is something I need to match on the printer. I hope that I suscenct stated my issue. If not, please forgive me. Thanx
Have you ever considered trying out the A4988 'fix' by shorting the resistor next to the cooling element? If you're interested I can give you some background information. What this does, it changing the decay setting on the A4988. I would really like to see your comparison since you use a Decibel meter.
I'm certainly interested in hearing more about this.
@@KayakingVince Here is more info:
reprap.org/forum/read.php?1,465761,851804
Sadly TeachingTech never replies to any comments. Would be worthwhile to try this out with a good decibel meter and vibration setup.
I was trying to get eh LV8729 to work on teh E0 on my SKR 1.4 Turbo. I get zero movement out of the steppers. Anyone have an idea what could cause the issue? All Marlin 2.0 the TMC 2209's worked fine. I disabled stealthchop in the firmware for E and set the driver in the config file to LV8729. Still no movement.
Awesome work as usual!
I went with the 2208s and mks gen l. But I still need to install the card reader.
The difference is worth it, it makes much less noise.
Thank you
I'm running 2130 SPI on E0 at about 850 in stealthchop. No feed issues to date. Was giving me a headache in spreadcycle w/ a 12-14kHz whine.
Mike, as always, THANKS A LOT!!!
Hey. Great video.
Could you please tell me (if you know) wich is the best driver to replace in the Artillery sidewinder x1 without changing the firmware?
Plz make a video on the TMC5160 drivers (SPI INterface, with RAMP and features) :)
Thanks to your videos, I think I'm going with 2208, printer is in the room and 4498 are really loud, and on the other side, not really into sensorless homing. Already have MKS board and screen, but ordered wrong drivers...
I did the same! You won't believe how quiet 2208s are. Mine are still in legacy mode.
I am running X/Y with TMC2130, extruder and Z with A4988. I have spare TMC2130 and DRV8825 one piece each, should I change my extruder and Z with them and which go to which ?
M502 loads the firmware values and then save with M500 will solve the M501 reading so that you don't have to enter this manually
I've gone for the 2208s and MKS Gen L combo in my Ender 5. Going for silence, so the PSU is getting replaced along with moving it and the Gen L outside of the bottom enclosure. Big quiet fans and a new hotend cooling assembly with quieter fans there should do the job I hope.
great job, thanks
Why do not use the TCM2130 to detect jams on the extruder?
Are you looking forward to the TMC2209's? They're going to be sold from watterott in a few days.
What about the TMC5160's?
@@akselanthonisenribe9984
The TMC5160 and TMC2209 have some slight differences.
The TMC5160 has:
SPI as it's communication protocol which is slightly faster than the TMC2209's UART
dcStep - load dependent speed control
The TMC2209 has
stallGuard4 while the TMC5160 just has stallGuard2
Besides that, they're mostly the same.
After watch this perfect tutorial and changed my Lv8729 to 64 micro steps I hit a strange but.
When doing retraction the extruder pulls back the filament. The next thing in the gcode is to return the fillament back, but the extruder decides to pull back one more time (second retraction)
Using the same slicer as before and the safe firmware (just updated the esteps)
Marlin 1.1.9. Any ideas?
Hi Micheal. Another great video, as usual! I just noticed one thing that I missed in previous videos, you changed the hotend fan from Noctua? Any specific reason for this?
Will sensor less homing on the extruder detect a filament jam?
Run out is easy with a switch but jams are more difficult.
The extruder gear slips too easily, I doubt you'd be able to get consistent results.
@@deceitive3338 definitly right. Filament jam mostly leads to grinding down the filament, not to skipped steps (on a good extruder). Maybe a dust detector for the produced dust may detect it ;-)
But seriously: Prusas old mk3 sensor was a good solution (laser sensor). Sadly they changed it as it got 2 problems:
a) Reliability. Prusa uses black PET-G for printing the printhead. But PET-G is highly reflecting, even if you use black PET-G. I think they did the prototypes of the mk3 in ABS (like the mk2 was printed in ABS), which is less reflective. So they get it never reliable. I solved this issue by applying black non-glossy paint inside the detection chamber, never ever having false results.
b) transparent high reflective filament (PET-G transparent or PC). Ever moved a mouse on a glas table? Can not work. Simple solution: run the filament though some gears first, leaving marks on it. Like sanding the glasstable if you want to use the mouse directly: Not elegant but solves the problem.
Couldn't we use sensorless homing for G29 bed leveling?
Did the 2208 upgrade little over a month ago. Much quieter now but didn't really help with some interesting vertical patterns or zebra striping Issues I have. Wondering if my table is too wobbly or is I should turn up the v ref on one or two of my drivers. What do you think? Thanks as always for your effort and excellent info.
What vref values did you use? Are the stripes mostly on one axis?
@@deceitive3338 So I used the formula from his previous video and the stat sheet from the drivers I bought. Id have to check again to know exactly. I can see the layers, assume that's normal, but I seem to have vertical stripes on the walls of some prints. Not as noticeable on others and I know they aren't seems.
i just put my extruder driver in hybrid threshold personally. it works fine
How do you change the steps if your board does not have the jumper sections? Like the gorilla board on the Anyubic?
I've got an 8 bit main board with removable drivers, but no jumpers underneath! Am I doomed to stay at 1:16 microstepping?
I have all LV8729s at 1/32 and it is quieter but fast X travel makes the same ringing noise as seen in your 1/64 example. Might just get 1 TMC driver for X
check out squash ball feet i put a set on my cr-10 with stepper motor dampers(had them of for a while now) and my printer is almost silent with the a4988 drivers.
Hi, just curious, I am using the Bigtreetech skr v1.3 board on my CR-10S, along with the LV8729 drivers. So my question is this, where do I locate the vref values for the 10S.
Have you checked out the tmc2225? I see they sell on Amazon for less than the tmc2209.
I'm getting an BTT SKR 1.4 Turbo for my Tarantula Pro, I want to put in updated drivers, but I need it to be somewhat easy, I'd like to just drop them in. Is 2209s easier than 2208s? How about the higher series with no pot adj, SPI? Any suggestions? Thx
TMC-s starting to have problems on extruder when you use linear advance.
Same problem hear, tmc2208 with linear advance stop extrude, mega 2560
Hi, I would like to make an upgrade on my sidewinder x1 it is possible to replace the original drivers with LV8729 for both the axis motors and the extruder and what are the best settings to change the firmware. Maybe you can make a dedicated video. thanks
this is marlin v1.1 ? if yes use v1.1.9 good ?
i see marlin 1.0 not show #define X_DRIVER_TYPE in configuration.h
LV8729 use with Marlin 1.0 ?
i use marlin 1.0
Just looked last night for a review of these. Mind reader!
Excelent video, exist problem with linear advance and tmc2208, stop extrude. Lv8729 possible solution...
Hi, could You by chance do a Video on howto do wiring of TMC2208 (UART Mode) with an MKS SGen 1.0 Board? Seems there are some Videos with the SBase but the SGen is somewhat different.
Should I buy 8729 or 2209 v3 (spreadcycle) for E?
2208..and 2209.. Have problems with linear advance.
Mine Vref for this lv8729 (exactli same with 220ohm) i set to 0.96v and no skipping or hot motor. Got it on E on xyz i got TMC2208
I woudl agree that 2100 / 2130's shouldn't be used on extruders. 2208's shouldn't really be used in bowden or geared, however if you have done the direct drive mod and your retracts are relatively short (1mm) and the speeds are relatively slow (30mm/sec) it seems to work fine and be silent. Switching to a smaller hobb can help as well.
I have a question could you do the same stepper review but with the smoothers as i have found that adding the smoothers does make things considerable quiter, i have the pmaxilion 320 printer and using the a4988 steppers and after adding the smoothers its is about 80% quiter
I just got my skr 1.3 and debating on the drivers.
Will micro stepping give me better resolution to make a noticable difference?
what about tmc 5160
About the Vref on the LV8729, I think you got the R220 resistor drivers so the right formula would be:
Vref = I x 1,1
as far as I know this is how it should be measured:
R100: Vref = I x 0,5
R220: Vref = I x 1,1
Can you confirm this?
Datasheet says I = (Vref / 5) / Rsense.
My skr v1.3 will be coming soon. I ordered 2208's and to be sure a lv8729 for the extruder. Why do people advice 8729? Do they have more torque? And can I run them 1/128 with the 32 bit board?
Because someone had a problem and didn't know how to diagnose what was happening, ignorance basically. They then told a friend who then told a friend who then told a friend, and so a big bright new baby bs fact was born. And all those who didn't know any better swallowed it as "good advice" and passed it on knowingly. In other words, the standard method.
I use 2208's on XYZ and E on my 10s Pro on a SKR v1.3 via UARR. I have no problem with the extruder.
You will if you use linear advance. It simply won't move the extruder....or worse, it will cut out after a bit.
What is the direct extrusion mod on that Ender 3? Didn't manage to find it mentioned in the videos. Thanks!
going with tmc2209
The reason none of your settings take effect after uploading marlin is because you still have settings stored in EEPROM. You need to do an M502 to reinitialize eeprom with values from firmware, then all your settings you set in firmware will be in place. Then an M500 to store in EEPROM and you're good to go.
So, with a 32 bit board.
The LV8729's are a really good option? (the last ones)
on a skr 1.3 i tried both tmc2209 and lv8729. put the lv8729 @ 1/64 and they crush the tmcs. they are both quieter AND have more torque AND don't need fiddling with stealthchop/spreadcycle/hybrid modes and the like
So LV8729 relies strictly on the micro controller for creating steps and risks skipped steps at really high accelerations, but how do TMC stepper drivers work with 8-bit because they take 1/16 micro stepping pulses and interpolate internally 1/256 micro steps and take less computing resources from the 8-bit controller
I hope I am understanding this stuff right because there is a lot of here say and snake oil on the internet
No tmc 5160s?
Hey! I think the website that you used as reference got the Vref formula wrong. In the official LV8729 datasheet (www.mouser.com/datasheet/2/308/LV8729V-D-278452.pdf) the formula is Iout = (Vref/5)/R. The formula that you used was propably for DRV8825 driver. That is propably the reason why you have experienced some skipped steps before you upped the Vrefs. Just wanted to let everyone know to save some troubleshooting time :)
What stepperdriver is best for lin_adv. , i have a skr 1.4 turbo , an can i mix stepperdriver ?
Maybe lv8729, i have problems 2208 and linear advance
@@Marcushin 2208 you absolutely need to run in SpreadCycle mode for the extruder.
5th. Finally I made it in the top 10. 😁
Another reason for me to upgrade my Ender3 main board I suppose.
th-cam.com/video/duNHOPlh2Pg/w-d-xo.html
why didnt you try it with the 128 micro stepping ????
Per his own words in the video, The 8 bit board had a hard enough time processing the 1/64. He encountered skipping on the fast X movement. It would be presumed that 1/128 would be even worse for skipping steps.
@@ylojkt5174 oh my bad i literally missed that since i was watching while cooking..
thanks for the clarification :)
OK, so why are the TMC drivers not suitable for extruders?
Oh i have to see all the video to conclude that they are good for extuder.
And... The Anet E10 it's necesary put un spread cycle or stealchop it's enough
If u have skr v1.3 no need dupont cable,
Your have something wrong with your mic, you sound a bit robotic. Maybe low bittrate? Apart from that great video!
you had idle whine in spreadcycle mode on the tmc2130's - i had this same issue - that noise can be removed th-cam.com/video/0OFpiJAMYZQ/w-d-xo.html
creality is becoming apple they wont let you upgrade their hardware
13:20 "Challenge: Build a printer that uses four different drivers"
Wouldn't surprise me if a Chinese manufacturer has already completed your challenge through sheer incompetence.
Are TMC 2208s drop-in compatible with a CR-10S logic board, if replacing the driver chips directly?
Usually those came with a Melzi-like board with A4988 soldered in. The only pinout compatible chip is maybe Hero HR4988, which supports higher microstepping modes (configurable). At 16 microsteps, it's a little less annoying sounding than A4988, but i don't know about its behaviour quirks. Straps for 16 microsteps are the same between A4988 and HR4988, but for all other modes they differ.