3144 5V Hall Sensors on duet?
-
@dc42 said in 3144 5V Hall Sensors on duet?:
SS441R
For USA, these are in stock for immediate shipment at Mouser:
@vlad, in what country are you located?
-
Thank you guys for the reply. I actually have already installed the A3144 model and as I said, very easily fed the voltage from 5V pin and it worked beautifully and easily, and this sensor is the most available on a market too, at least here in USA. I am also playing with stallguard, but so far it is not very reliable on my test machine, but still will try it on my upcoming delta. Very interesting concept if I could use that for nozzle leveling there.
-
@Vlad said in 3144 5V Hall Sensors on duet?:
I actually have already installed the A3144 model and as I said, very easily fed the voltage from 5V pin and it worked beautifully and easily, and this sensor is the most available on a market too, at least here in USA.
By "the market", I presume you mean hobby shops, many of whom sell through Amazon and eBay. Sadly, many of these shops haven't yet caught up with the general shift to 3.3V electronics, which happened many years ago in professional electronics. In the USA you have several electronics components distributors (Digikey, Mouser, Newark etc.) which will happily supply to individuals, although small orders typically invoke a carriage charge, or there mnay be a minimum order value.
I am also playing with stallguard, but so far it is not very reliable on my test machine, but still will try it on my upcoming delta. Very interesting concept if I could use that for nozzle leveling there.
Unfortunately, stall detection is at best only accurate to within +/1 full motor step. On most deltas this corresponds to a distance of between 80um and 200um, which is too high for bed probing.
However, I have found that it it possible to home a delta printer using stall detection, if a Z probe (preferably of the nozzle-contact variety) is used to do a single auto calibration cycle after homing, to remove the homing inaccuracy.
I hypothesize that stall detection works best when the motor is close to the physical endstop, so that the length of belt between motor and carriage is small when the carriage reaches the endstop. This ensures a hard motor stall. However, other configurations are possible if the belts are taut enough.
HTH David
-
@dc42 yeah, I meant basically amazon-ebay USA market. I shop on amazon 99% of the time, unless I have to shop elsewhere due to superfast free shipping, free returns and other goodies for customers. Very good observation about the belt length, will definitely keep in mind. I could go with 0.9 steppers for delta axis in order to somehow improve potential accuracy of stall detection, but I believe 0.9 steppers are not very good at very high speeds, so I stocked up on some 1.8 MOONS ones. My bed will be 4mm glass on top of copper plate, and I really don't have space on my custom effector even for BLtouch sensor. Need something very tiny. Already researched piezo types, but they seem to have a lot of issues at aggressive accelerations and speeds. Was trying to find your IR sensor for testing, but none of our distributors carry it anymore. Ideally I would want to level the bed with the nozzle itself, but have not yet found a solution to do that reliably. I am also not looking for any kind of mesh probing for my delta, it is factory produced and rigid as tank. I am only needing the probe to basically probe 3 points to calibrate a plane of surface. Also since I am using a glass surface, I am not worried about mesh much, glass is super flat always.
-
@Vlad said in 3144 5V Hall Sensors on duet?:
I could go with 0.9 steppers for delta axis in order to somehow improve potential accuracy of stall detection, but I believe 0.9 steppers are not very good at very high speeds...
I recommend 0.9deg motors for deltas because you get a finer more using them; but you need to use low inductance motors and 24V power to get good speeds from them. Even with 0.9deg motors and 16 tooth GT2 belts, you only get 80um full-step resolution, so no good for probing.
Ideally I would want to level the bed with the nozzle itself, but have not yet found a solution to do that reliably.
IMO the only appropriate Z probe for a delta is a nozzle-contact sensor of some sort. Take your pick from strain gauge (Smart Effector), piezo on the effector, or piezos or FSRs under the bed supports. The Smart Effector in particular is very good - but I would say that, wouldn't I?
I am only needing the probe to basically probe 3 points to calibrate a plane of surface. Also since I am using a glass surface, I am not worried about mesh much, glass is super flat always.
You needs to probe a lot more than 3 points to get good auto calibration. I agree that on a well-built delta, mesh bed compensation is not necessary.
-
@Vlad said in 3144 5V Hall Sensors on duet?:
I meant basically amazon-ebay USA market. I shop on amazon 99% of the time
Me too. Having said that, Mouser is quite painless.
-
@dc42 Thanks a lot for your replies David and for your advice about probe options. Really appreciate that! Smart effector is definitely a solid all around solution that is well priced, however, unfortunately it can not fit some of my requirements, so I can't go that route (COG and weight of magballs being among my reasons). And there is nothing wrong in liking your own product haha, after all this is why you made it!
You just gave me what seems like a great idea - piezos on a bed - definitely worth a try! My bed so far is hard-connected to a frame without a way to adjust it at all. I have done it to get rid of all possible vibrations. Going to buy some piezos now for playing for sure. I guess 3pcs will be sufficient.
In the last day of playing with stallguard, I did realize, while it is fun, it should not be a primary solution imo. It worked great on my old printer as a replacement of broken X homing mechanism tho. It also works great for where most of us don't have endstops. For example on the other side of X or Y axis etc. Stall detection for extruder is also very interesting, I am going to play with programming it later. However I soon realized that it is still much better to install my hall effect sensors for endstops and it is not that much harder to do, and will be a much better solution for fast homing, as they trigger about 8mm away from the sensor.
Just purchased this set: https://www.amazon.com/dp/B07Z8SXFPQ/ hopefully it will be compatible
As a side question, does my Duet2wifi have stallguard2 or the first iteration StallGuard?
-
@dc42 oh wait... strain gauge? for some reason I always thought you had piezo sensor in your effector. Actually strain gauge seems like a better solution out of the box. Is there anything special I will need to do firmware-wise to implement that? It is so tiny that I am sure I can find a way to install it on my effector where it would slightly bend from the presure. Soooo, I ordered this just now Too bad I won't be able to get it before this weekend already, but my effector is still in limbo anyways, due to unchosen ball joint type. Had a meeting with IGUS and they sent me their samples for those and for rails, both of which are unfortunately pretty bad in terms of tolerances and play. Hard to believe that it is so hard to find a decent tight lightweight ball joints on a market... oh well! I wonder what it would take to use small steel ball bearings and injection-mold a shell with PTFE or POM or something similar, and do it with proper tolerances. If you don't mind me asking, what made you choose strain gauge over the piezo type sensor?
Also what kind of electronics do I need along with strain gauge in order for it to work with DUET? Is there any wiki page to help me with that?
-
@Vlad said in 3144 5V Hall Sensors on duet?:
Actually strain gauge seems like a better solution out of the box. Is there anything special I will need to do firmware-wise to implement that?
Strain gauge amplifier and a comparator. Feel free too review the smart effector, it's schematic and embedded firmware is in GitHub.
-
Cool, I believe the amp is already included in that product that I have purchased. So just a comparator circuit then. Do you think a MOS module or an SSR could do a trick there (could be stupid question tho )? Ideally I will want to only do that procedure once a year or less, will see how sturdy my frame is once I get the machine up and running. I decided to first try to use hall effect sensors for homing and manual leveling procedure and try to store those leveling coordinates int he firmware. hall effect are very precize sensors and theoretically I shouldn't "need" to reprobe it at all. But that all is only a theory that requires testing still