The Sherpa Mini Is The Lightest 3D Printer Extruder I've ever seen.

[@DailyFrankPeter]

19:04 I don't understand this calculation - would you mind explaining?

[@DesignPrototypeTest]

Area of a circle is = πr² Therefore, 1.75mm filament has a cross section area of 2.4mm². 345mm/minute (measured in the video) multiplied by 2.4mm² means 828mm³/minute is the volume this extruder/hotend unit is capable of pushing. When printing, typical settings used by most people are .2mm layer height and a .4mm nozzle. so our extruded lines are .4mm wide .2mm thick, and extrusion length is the unknown thing we need to figure out. .2*.4=.08mm² area. Take our volume from before of 828mm³/minute and divide by .08mm² gives us 10,350mm/minute divide by 60 seconds in a minute and you get 172mm/second in extrusion length. In other words the max speed of this extruder/hotend unit is 172mm/s. in the X,Y travel directions. Exceeding this speed will cause the stepper on the extruder to start missing steps resulting in under extrusion and the poor print quality associated with that.

[@mikkellarsen4931]

@@DesignPrototypeTest Hello and good day to you. I would like to ask you if you could Make a list to me or send it to me on at message on what you have on your 3D printer. Ender-3 v2 That makes it go up to 300 degrees i really really much appreciate if you could do that for me so I can buy the same stuff That you did buy to get it to be so good. With all that extra stuff on it. Is it good enough that it can take it at that temperature And not destroy itself because of too much?? Please, it would help me very very much and I would be very glad if you could do that for me. Thank you on hand and have a nice day

[@berendlucasvanderweide]

@@DesignPrototypeTest Thanks for these tests, very handy in the search for parts. Did you ever tested the max flow rate using a larger nozzle? Or perhaps with the 3dsolex nozzles? It makes perfect sense that the torque on the stepper on the orbiter drops at the higher RPM's. You need.. if you feed that motor a higher voltage it most likely will get back some of its torque at higher RPM's. But there is a limit in terms of saturation.

[@mongini1]

@@DesignPrototypeTest sorry for digging this out again... i installed the new stealthburner on my printer which basically uses the same setup on the inside (same BMG style gearing, same nema 14 stepper etc...) paired with a dragon hotend (normal flow, plated copper nozzle, no funny CHT business yet) and im feeding just 0.2 amps into the extruder - yet im still able to print at 200mm/s (mostly for infill, or when i need a functional part fast) without any underextrusion. To make sure this is true, i repeated the same test you did. i started skipping at F415, with F410 being the last value without a measureable skip. This turns out to 205 mm/s by your math, which would be in line. The thing im wondering now: i just got a normal flow Dragon, and i just feed 200 mA into the stepper... how does that work out? is it because i got an LDO stepper? is it the filament? (used ASA for this) is it becaue of the construction of the stealthburner? Talking about that: the latch is so beautiful, once disengaged i can freely push filament through without noticable resistance anywhere. That reduced friction might have an impact aswell... just guessing at this point. All in all - the Sherpa mini is an awesome little beast, the Hextrudort for the HEVORT is very similar, and this printer gets up to speeds i'd never touch. Cheers

[@rouldiy]

@@mongini1 the bottleneck here is not the extruder but the hotend. On my ZeroG Mercury One.1 with a Sherpa Mini I print with speeds of 300mm/s without any under extrusion. The hotend is a Dragon HF hotend that is capable of delivering a volumetric flow rate of 24 mm³/s. Printing at 300mm/s, 0.4 line width and 0.2 mm layer height would add up to about 22 mm³/s, so there is actually some headroom just to be safe. What was tested here in this video is imho not the capability of the extruder but rather what the hotend with that nozzle and the set temperature can keep up with. 350mm/min with 1.75mm filament is about 14 mm³/s and that is exactly what a standard mosquito can handle.

[@trashpanda9433]

I can't wait for your design to come out.

[@rexinsightrobotics]

with 50% + weight reduction

[@nirodper]

I think you should use the units everyone uses for extruders or hotends: volumetric flow. 350mm/min of 1.75mm filament is 14mm³/s (2.4mm² × 350mm/min ÷ 60s) That's about the maximum the standard mosquito can manage on PLA You can also weigh the extruded filament with a miligram scale

[@kazolar]

I'm able to go to 600mm/min with the mosquito magnum and PETg -- Vez is using a slightly altered sherpa mini -- just to adjust for mounting with a mosquito magnum plus and he is hitting 700mm/sec actual print speeds with insane accelerations/jerk settings. One thing to do research in a vacuum and come to your conclusions, but it's another to just throw everything out the window where there are folks using the hardware you're testing and getting significantly different results

[@nirodper]

@@kazolar nobody said otherwise, I was just refering to a "standard" situation. Increasing the temp, using the magnum version, more filament driving pressure or a different polymer will change the volumetric flow you can achieve. But knowing the baseline performance -around 14mm³/s for a std mosquito- is useful, for example in this case where Mathew may not be running into the extruder limit, but the hotend's being a bottleneck

[@kazolar]

@@nirodper If you go back to when he started doing these "tests" he keeps insisting these lightweight extruders as he is testing them maxing out at 300-350mm/min are a limited by extruder. If he bothered to do mm^3/sec calculation at any point ...this is far from cnc kitchen level of testing

[@nirodper]

@@kazolar I agree

[@nirodper]

@Alexandru Ciobanu depends on the extrusion width (equal to nozzle diameter in standard slicer settings usually), and if your acceleration is high enough to reach 150mm/s. But yes, at 150mm/s with a 0.8mm wide and 0.5mm high extrusion the flow would be 60mm³/s. For that flow you either need a supervolcano (maybe a volcano could do it) or a CHT nozzle with a mosquito/supermosquito, unless you have those you are probably not reaching that printing speed. Not even mentioning the requirements to cool that amount of plastic

[@kazolar]

A few comments on how you proceeded to test the sherpa. 1. You are testing a knockoff from China -- stop it. This is wasting everyone's time, everyone knows they're releasing an older version of the design, they're not sourcing authentic BMG gears. All we know that is any decent is they're forced to use a real LDO stepper (and I question the stepper they're using). 2. Your extruder is already not built right, it's binding -- i printed the body a few times out of a PC blend with adjustments to settings to get a clean print, you should be able to pull/push plastic through it very easily, the assembly itself should not binding -- again, more aliexpress related nonsense with what you bought. 3. Use the latest designs and stls and print it yourself. Get the latest 8t LDO stepper, it increases the gear ration 6.25 to 1. I have the latest version, and I actually have the modified version which is compatible with a PTFE lock at the top, I also did a few geometry tweaks to it, minor stuff, but i found it improves rigidity, and makes the geometry less complex. This is open source, STEP files are available. You don't like the mounting arrangement, CHANGE IT. 4. I have it running on my delta, yes delta, the fundamental difference between it and Orbiter is that the filament path is centered, where as on the orbiter the motor hangs out the back so makes making it adapt to a delta effector less ideal I tested it with angled filament path, worked, but sherpa is much better. I have it running with mosquito magnum, and I have not had any issues with 600mm/min feed extruder test, no skipped steps, didn't try higher, as the nema17 BMG can't go much higher either. I have done several prints with it and .8mm nozzle and 90mm/sec speeds with PETg -- PETg doesn't cool properly if you go faster. I have not had any extruder issues or heat issues. 5. My assembly with screws and PTFE locking coupler 113g. A lot of the speed racers are using the sherpa on their corexy machines, so it's definitely capable, but they're not sourcing LDO motors from the back of a truck as they do on aliexpress.

[@travistucker7317]

Woo sah kazo san

[@ficeto]

You looked familiar when I clicked on the video, then I remembered why I do not watch your videos anymore. Constant whining... this is an extruder, yet you did not print anything with it. All video talking what you do not like and how it could have been better... make it better if you know so much!

[@MirageC]

Great Video man :) you have a new sub in me! Once factoring in the bigger extruder gears from the Orbiter, we are very close in terms of Steps/mm to the Sherpa, HextrudORT and other BMG geared mini extruders. One stepper motor rev of the Orbiter will result into a 4.608mm extrusion. HextrudORT and Sherpa will achieve 4.524. Inertia of rotation mass is the main difference between the two systems. While the Orbiter benefits from more traction, the others are more reactive for fast retraction and pressure advance moves.

[@DesignPrototypeTest]

Thanks man! I love, love, love the work that you do!

[@avejst]

Great video, and interesting trophic. Love that you are making, more or less, the same test with different extruders👍😀 Thanks for sharing your experience with all of us👍😀

[@letitburn6]

I'm running the Sherpa with a Mosquito on my Hevort. Awesome extruder

[@JoshTownsend66]

You could have made a video about how Annex designed the lightest extruder you can PRINT YOURSELF. Then show how it was originally made for there K series printers, but they released a universal mount so others design there own. Also if you have trouble printing abs or want a kit, fysec makes one .... BUT NO WE GOT THIS VIDEO INSTEAD

[@user-yk1cw8im4h]

I know right, this guy’s mind is warped as fuck and sounds like he’s making sense but actually doesn’t.

[@Penofhell]

@@user-yk1cw8im4h I know right, I said exactly that a few months back on one of his videos (in fact going into some details regarding which points actually sound like they male sense but actually showed lack of understanding) and the guy responded saying I was a nobody on the internet and my opinion didn't matter... When I saw this post pop up in my feed for whatever reason, I thought well maybe things have changed around here.... Well.... Begining the video saying "if you're watching one of MY video and you don't have ADVANCED KNOWLEDGE of 3D printers" this is so fucking preposterous. That plus the overall "I could do it all 50% better and I'm treating an entire >1billion population country like they're all stupid" on top of the overall writing ... And what do you know, I look into the comments and people are seeing this too. I'm kinda reassured to see the ridiculous thumbs down/up ratio and will happily endavour to teach the KZbin algorithm to not direct me here any more :p

[@joseholguin436]

@@Penofhell haha same thing. This channel comment threads are becoming like the Jerry Springer of 3d printer channels hahaha

[@ADH-DIY]

Every time this dude reviews something from a company that doesn't pay him he just buries and trash talks the product then spouts offhanded comments about Chinese manufacturing. Then says "I'm not talking about people I'm talking about a place" as if a landmass can manufacture.

[@Nobody-Nowhere]

I would think that gear reduction does not always equal more torque in stepper motors, as stepper motors have the highest torque at slow speeds and loose torque a lot when the speed increases. There is probably a optimal reduction.

[@pr0xZen]

Looking at this thing, I got some thoughts. 1: You could probably use a N17 and not have to _drastically_ change the ratio. But you'll have to use a really tough material gear, as you'll need to drill it out leaving little meat on the bone around the axle. Compensated moderately by roughing the axle and gear ID, and epoxying the gear to the axle. Tooth profile slightly shallower. But - going with something like an LDO 36STH20 would probably be favorable. 2: If you intend to mod the design, then it looks like you could add another gear on the opposite end of the tension lever, for manual turning of the extruder gears. Adding another gear like this does add a few grams, but it grants manual operation access without need to alter the core mechanics.

[@ralphchurch3900]

Do it bud the community needs ur ideas

[@gangsterHOTLINE]

Gotta say man of all the information I took in about this extruder from your video... the sandpaper to grip ptfe tube will go forward with me forever. No longer will I look like an unprofessional fool with butterfingers.

[@johnh3ss]

From what I understand, for higher gear ratios where the motor has to spin faster, you actually want the current to be lower, not higher. The reason is because the higher current creates a stronger magnetic field that takes longer to collapse, so it can't cycle as fast with more milliamps. This is especially true for remote extruders like the Nimble/Flex3D. I ran my Flex3D at 400ma with a Nema 17 pancake motor and that worked quite well.

[@DesignPrototypeTest]

Thank you for the good information.

[@enginerdy]

I don’t think you can control that just by varying current. At high speeds, the back emf of the motor windings becomes more significant, and you are not able to push the current into the winding, which causes a torque drop. The gear ratio and motor need to be matched for the application.

[@danielfernandes2813]

i got my 1st 3d printer from china. I'm happy they just push out products, its upto the consumer to choose wisely.

[@im_ricebowl]

I design frames and I'm not gonna lie, that mindset you have would be very suited for frame design for fpv drones....I would love to see how youd tackle weight savings vs durability

[@Lobbytime1988]

"Mellow Lightweight NF-sunrise V6" Have you seen this one on aliexpress it has the small pancake direct hotend mount lightweight and all metal except for drive gear. It looks like a contender in your competition for lightest smallest.

[@CraftedChannel]

The reason the slower running motor (less gear reduction) made less heat is because of magnetic reluctance. Running at a lower RPM required a fewer number of magnetic pole reversals forced in by the stepper controller.

[@FinepixF30]

I would add than is not only about pushing plastic faster but being able to melt this plastic faster, a lot of the skip steps problems are due to pushing plastic faster than the ability to the hot end to melt it since you can't push solid plastic through the nozzle. A volcano hot end with and a higher power heat cartridge might be required

[@enginerdy]

A couple of things about motors: the increase in winding counts increases the magnetic flux and lowers the current to torque relationship. It’s hard to compare motors of different construction because it depends on other things too, but for motors of the same build, more windings means higher voltage and less current for the same torque. Second, when your stepper gets really hot, the windings increase resistance and depending on the operating conditions you can cause the maximum torque you’re producing to be reduced.

[@osmanpasha_diy]

Amazing analysis, thank you!

[@enosunim]

Is it OK to call feeder an extruder? While in the beginning, when you showed full assembly it was an extruder. But lately it was only feeder, as it only feeds, it does not extrude anything on its own.

[@livedeliciously]

It's been 3 years, where is your design that's so much better?

[@GrimmChristopher]

I honestly think that this video is the best you have published so far. You picked up the comments of the community from your last videos and integrated the conclusions in this video. Therefore, starting a discord is a great idea, as your research will be community supported if done right. I also really appreciate, that there is only minimum rant.

[@citronski]

Loading and unloading Filament could be a menu option in the firmware. Loading would turn the motor one way, unloading the other. By removing the option to manually load or unload you could just remove most of the lever. Great Video, thanks mate.👍

[@AnnaVannieuwenhuyse]

The lever is also there for pretensioning. That's what the spring is for on the screw :)

[@radicalxedward8047]

13:00 tip for gripping PTFE without marring it, just use a little gaffer tape. plenty grippy especially while being held, removes cleanly, doesn't mar surface, etc.

[@TheP00N]

Damn... Simple, but ingenious solution. Thank you!!!

[@foggyfrogman1]

I love my Sherpa mini, I printed it in PCCF. Going to try an ASA on another machine.

[@henricoderre]

Creality also has an affordable modular direct drive extruder. I'm trying it out now. It is the first mod on my new E3V2 printer, and the printer has not been used yet. This is my second E3V2 printer. My older printer still has the Bowden setup. I am not impressed with the Bowden tube extruder system. Creality's direct drive comes with all the necessary installation hardware parts. The package simply includes a photo identifying the ports on the v4.2.2 motherboard, and no instructions. However, installation is rather straightforward, and could be done in about an hour. After removing the extruder stepper motor, stock Bowden tube, and hot end from the X-axis gantry, I continued to disconnect the prior from the motherboard. Then, I installed the direct drive connections to the motherboard. The connections are identical to the stock equipment. Once finished, I only had to readjust the E-steps which were only slightly off. I need to tweak the printer a bit, i.e. bed-levelling, but everything is working fine. I did not weigh the unit, but I found it light. The only thing I don't like about this direct drive is that the extruder assembly is made of plastic just like the stock extruder. I'd replaced the plastic one for an aluminum one on my older printer after the lever cracked. The lever is notorious for doing this. The plastic extruder on the direct drive is installed behind the stepper motor, and should be replaceable with an aluminum one.

[@Luka1180]

You can completely mitigate those vibrations by using an accelerometer.

[@coolcat312]

A Discord would be great!

[@MyklCarlton]

Nice work, as usual! I think the boxes came out of the same factory, rather than indicating the parts inside came from the same factory. Just a thought.

[@enlightendbel]

These round steppers are in fact NEMA 17 steppers. NEMA standard is a size standard. NEMA 17 means NEMA standard size 17, which tells you the holes and square inch size of the face. The 11/17/23/24/... means x/10 square inches for the face. A NEMA 17 motor has a square inch footprint of 1.7x1.7 inch. As you noticed, the round one has the same holes and face diameter, making it a NEMA 17 motor. NEMA size motors come in all sorts of force ratings for the same size and the small round one may be lower power, but that's ok because of the gearing. That's the difference between a common 42-40 NEMA 17 sized motor and the pancake which is a 42-25 NEMA 17, it's only 25 deep. NEMA size motors can come with vastly different internals too, the ones we use are commonly 2 poles, but 4 pole ones are common too. The gearing mechanisms in the Sherpa, Orbiter, etc are there so you can fit a smaller stepper. The smaller stepper is the point, not a disadvantage. The gearing is designed to make use of a weaker motor possible.

[@kyleo1236]

They definitely could have opened the hole on the other side of the hobbed gear from the filament path to avoid needing that front cap to be removable. It could still have tight tolerances in the filament path that way.

[@dhirajbokde4670]

Great video. Maybe it would be better to test maximum extruder rate by: 1. Reducing minimum extrusion temp to below room temp 2. Removing nozzle/heater block to isolate the test to extruder stepper motor and its gears 3. Extruding cold filament to calculate the maximum achievable volumetric flow using pi*r^2

[@fumo1000]

great view from an engineering perspective!

[@livedeliciously]

The Sherpa Mini is an amazing extruder. All the criticisms you hear is just nitpicking.

[@zola5584]

Good content 😁......I would be curious about maximum extruder speed ,what doesnt cause warming stepper motor....and there is someoption for cooling: heat sink and air channel to stepper from hotend fan to stepper......and the current/voltage compare regular extruder motor

[@mikeyearwood]

I totally agree. Integration in this case makes for a better product overall. This idea does not apply in software. Software is not physical.

[@toast47624]

Thank you for these in depth reviews. I'm about too scratch build a 3d printer and now I have also decided to scratch build the print head/Extruder. I'll design it in Fusion and manufacture it on my CNC. If you don't mind I will use a great deal of your thoughts here.

[@beaker2k]

Did you use the same stepper your disassembled? You take apart a stepper and it will have less torque after reassembly

[@DoRC]

70c isn't going to hurt the magnets. Magnets don't start losing magnetism until around 80c. Even then it's temporary until the magnet cools down again. Permanent magnetism loss occurs at much higher temperatures.

[@mikem6549]

After an age of nozzle clogs and chopping of Bowden tubes....I installed a direct drive on my ender3. I don't know what was wrong but this is wayyyy better.

[@RinksRides]

Monologue is a bit tangential at times like when you're comparing the weights made some of the video hard to follow. But extrusion testing made up for it in spades. Great video with great production quality, maths, testing methods, and explanations. You are underrated sir. IMO, you're right there with Thomas Sanladerer, Teaching Tech, Chris's Basement, and Joel (3DP Nerd). Subd!

[@qwewer6]

Please try klipper with its resonance compensation*, that should give you better ringing results even with heavy DD extruder setups. Edit: *with an accelerometer

[@DesignPrototypeTest]

Reprap firmware has resonance compensation as well. I don't know why everyone thinks you need Klipper for this functionality.

[@JoshTownsend66]

You can't use an accelerometer with reprap only klipper. You obviously haven't bothered looking into it or you wouldn't even be worried about weight so much.

[@JoshTownsend66]

He doesn't listen anyway

[@qwewer6]

Well, looking how easy and convenient a duet board is, if I had a duet board, then I might not have tried klipper at all, so I can't blame him for not trying it.

[@rothlive]

@@JoshTownsend66 Read Please : duet3d.dozuki.com/Wiki/Accelerometers

[@Skwisgar2322]

This might be a good fit for my 3ku Delta printer.

[@rothlive]

Great video !!!!! You are a great industrial designer. I can learn a lot from you. I'm excited

[@enlightendbel]

The stepper motors can run up to nearly 170-200C without issue, they are designed for that. The only thing you need to keep in mind is the glass transition temperature of the material your extruder extruder is made from. So for PLA 60C, PETG 80C, etc. And with the silly cooling setups people tend to run, it's trivial to route some of that wind force to blow over the motor. It'll cool down really easily (they are aluminium bodied for a reason).

[@sonderklasse9680]

Matt, regarding your negativity towards Bowden-style 3D printers makes me think you’ve never used an Ultimaker. Go ADA!

[@userskorea]

fyi circular small stepper motors are upto over 100-120C rated for operating temperature. so 50 or 70C may not affect torque or any performance. great video tho!

[@macjoel2009]

Have you looked at the AnkerMake m5 extruder? It's seems to be a hybrid bmg/sherpa combined with a really weird stepper. It seems to be a 36mm but it's square. I don't get it but it works great. The only part of the damn thing they got right

[@dreamcat4]

just watched your video today and it was nice to see here a realization of some of the finer points about why the 5:1 ratio and ditching the planetary gears of the orbiter does not result in a loss of performance, hence the orbiter design is not preferred over sherpa mini. Now in terms of taking apart the round nema14 it was very good to show that in the video. however bear in mind also there has been a strong recommendation not to do this on a real production use units because if you are not very careful, hitting the stator against the sides may cause a demagnetization event and then after reassembly it may not perform so well. however since your unit was an unbranded knock off that was a good reason to disregard that. if you are ordering a properly branded official ldo later on, then just replace it. for peace of mind. but maybe think twice about opening a real ldo motor. and check with other on the discord, before dismantling the real thing. but i do understand how very desirable it would be to replace that back cover of the motor. i just wanted to let know about the risks because aparrently if the motor becomes demagnetized in this way then it cannot be remagnetized back to the original factory condition. so maybe order 2 of them to have a spare one which you can keep unmodified for reference / comparison. would be a good idea i think now about the hextrudort... clearly having come out well after the sherpa mini the hextrudort does make some further and worthwhile advancement in extra weight savings and higher level of integration. and is also a pretty good choice. if you dont like the sherpa mini so much for whatever reasons. you can really use either one as your starting point. then extend and modify their respective open source cad files. and (re) print to become your own version. as is your liking. however to first become initially familiar with this family of extruders the stock sherpa mini is really the best starting point + any of its existing stock adapters or other pre existing user mods. so what you get there is most of the hard work done for you + thousands of hours of prior testing by annex engineering members. which is incidentally why i did not mention previously to you to start with the hextrudort before that. because it is significantly newer on the scene and a more reduced design, and has had less total amount of testing hours to date (by comparison to the sherpa). however the hextrudort is also an excellent extruder. and of course it is a logical and excellent progression path to move forwards after starting out on the sherpa mini i dont think there is too much to add because you are well on your way now. combined with your own pre-existing knowledge too. except more tips: it is worthwhile know that driving the ldo above 0.35 amp RMS (not peak, which would be 1.41x that)... yeah may cause the motor to die prematurely. but also of course you will loose torque if that also is meaning the motor is driven too fast. because of this i cannot recommend you a specific Amps on RRF because RRF is different and a bit weird. just should test and try somewhere in the range between 0.4 and 0.5 and see if you can find an optimum setting. and dont worry about the case temp so long as it does not go above what was it? 74c or something? lets say 70c to err on the safe side. but the better or more correct way to overdrive these and avoid the torque drop off curve at higher RPMs speed is to use a HV 5160 driven at 48v. instead of 24v. some people will be testing this soon. as for your preference of wanting to manually feed in filament sorry it is not in the design. they are designed for power feeding / unfeeding from the web interface. if you decide to try to modify the sherpa mini then maybe it would be difficult to add that feature compared to other modifications. then perhaps look at the hextrudort instead. and see if its any easier on that one. also it might be the case that hextrudort gives a slightly different type of OSS license to permit you to distribute your extruder modifications in your preferred way (to patreons only as a perk of that). however of course i would also encourage you to eventually release your own additions to the wider community. because it then lets others take your work even further along. for which everybody can benefit (including yourself too!) so it is by no means a waste of time to give away such work. and the original creators open source projects also can likewise benefit in a similar fashion finally i also want to mention that of course these extruders are not metal. so that is another area for future progression. a metal version would be a great further improvement for higher temp situations. its just the sherpa mini only came out during last 1-2 years and the hextrudort is even more recent. but the creator of the nova hotend seems to be working on a new extruder too. that one might be metal. or somebody else might figure something else out. currently we found the best out there was an all metal BMG clone from china. machined in aluminium. in fact there are 2 different ones. and the newer one now has the 4 mounting holes for the mosquito hotend. however it additionally needs some more material removed away from the housing of it. and replace the 1 large plastic POM bmg style gear inside with a metal one. and add some water cooling jacket or something to also cool the stepper too. so this extra missing work is why to instead maybe wait for the one being made by vlad of 3dpassion. who would do all of that harder work for you. and then charge a higher price but still not too high and probably be quite a reasonable value. YMMV

[@l3d-3dmaker58]

i said it once I'll say it again, hextrudort on nema14 + new lighter Nova 1.2 and klipper running input shaping, can't beat that for speed on a Delta or coreXY

[@Olof123]

Nova is pretty much a volcano with a polishing job and a smaller heatsink (plus I could never support that guy), and I don’t the the hevort Extruder is that different (I think miragec even said it was inspired by the Sherpa mini).

[@lucastonoli3256]

@@Olof123 Every single extruder out there in the market right now was inspired by some other extruder and many/most are just incremental improvements. Just saying.

[@Olof123]

@@lucastonoli3256 I’m pretty confident that the Sherpa mini was the first to go with bmg gears and the nema 14.

[@lucastonoli3256]

@@Olof123 I'm pretty sure someone came up with geared feeding before BMG did. I'm not talking down on the sherpa, great little extruder, just saying it's all built upon something. The Hevort extruder has merits to it too, all valid.

[@rattslayer]

If you think the Sherpa Mini is light.. have you heard of the Sherpa Micro, which also has an exposed gear for manual advance!

[@CraftedChannel]

It occurs to me when we look at the gear ratios it really means nothing because it is not including the diameter of the hob on the feed gear. Perhaps we should be giving feed rates in rotations per volume. Anything else is only part of the story. For the reader who isn't keeping up, 8, 12 and 18mm hobs all have different circumference's and thus dramatically different amounts of filament fed per rotation.

[@marcduchesne906]

Sherpas are great. Stop your free rant.

[@natanelbenzaki3071]

You've really gotten into my head about china and the "china teams", I now see it everywhere I look...

[@MichaelJHathaway]

Great video! I think my E3d hotend is actually lighter than my mosquito. Yes I have a copper block and aftermarket titanium heat break that is more expensive than the E3d's. I do love the mosquito except when it clogs and the E3d is superior there. I can machine one of those Nema shafts down for you on a lathe. And if you redesign the end caps and possibly the body, I can print them in my super pa6. I have thermal conductive filament now, so heat sink back plates are an option.

[@bcarroll03]

Running a Zesty Nimble V2 here. 20:1 gear ratio ( I had a v1 previously with a 30:1 gear ratio and It will fall on its face with higher flow rates) The Nimble 2 with is 20:1 ratio is fully capable of pushing 24mm/s of filament through volcano hotend. The motor never skips, the hobb always slips first. The bottleneck on the nimble is its single hobb design. I will be able to test its true limits when they release a dual hobb version. in a nutshell, stepper motors have a long way to go gear ratio wise before we loose substantial torque. 20:1 is great 30:1 is at the upper end. Forgot to mention I am running a nema 17 pancake.

[@MechanizedMedic]

A bit of input regarding your extruder testing... Extruding at a particular rate doesn't reflect so much on the extruder, more of how fast the hotend can melt the given filament. When trying to compare how powerful different extruders are a better test would be to measure its pulling force on the filament. This would give an apples-to-apples comparison for different extruders.

[@robmawston4986]

Do you have a video / article on what you've used on the mosquito hotblock, I assume for thermal stability, and strain relief ? This video is exceptionally well timed for me, as I'm wanting to upgrade my voron v0.1 extruder - thanks for the great content :)👍

[@BikerCaf]

The answer to the weight problem for 'direct drive' extruders I'm sure is to simply remove the motor from the equation. ie, move the motor to the main frame and couple the motor to the extruder via a solid sliding drive (square tube or spline etc) or flexible drive. Doing that also means you can use any motor you like, no power limitation, no step skipping and no more over heating motors.

[@mijaiscares8580]

Now are faster printers hahaha I wanna see new light extrusers

[@espanian52]

I`ve setted Vref to 0.3-0.32 V and it made stepper slightly warm and no issues with steps skipping. Retraction distances 1.6-2.0 mm. Retraction speed 45-60 mm/s. P.S. All Vref above 0.45 V makes Nema 14 pancake LDO very hot with no effectivness. Manufacturer board Vref for TMC2208 was setted on 0.78 V, so it have to be changed. P.S.S. And extruder steps (M92 E[...] command) setted to 700 (..// M92 E700 ).

[@DesignPrototypeTest]

Eventually it will skip. You haven't pushed it fast enough. Try following my procedure I showed in the video. I would be interested to know at what extrusion speed yours starts to lose steps.

[@espanian52]

@@DesignPrototypeTest Tested. You right. It realy skip at higher speed. Due to poor chinese quality of LDO my motor (marked FYSETC) is skipping at lower speed than yours. Had to up Vref to 0.7 and it still skipping at 300 mm/s. Motor is hot. Downvolted to 0.55-0.6 V. At this Vref 200 mm/s is no skipping and motors are not so hot. Think the limit of this motors is as you tested, but this limit is reachable on lower currents. Higher currents brings nothing to torque, just temperature loss. Vref at 0.3V = no skip at 100 mm/s, but no sure its no errors in huge amounts of retraction, because tested in linear extrusion of 100 mm of filament.

[@biggityboggityboo8775]

That's not how motors work. For equivalent current it is but more turns = greater magnetic flux being generated at the expense of requiring more voltage overhead (because of higher back EMF). If you've got the voltage overhead you end up with the most efficient design where the motor has just the right amount (higher) of turns to reach the desired rpm at the desired torque. That's with all things being equal with two otherwise identical motors and the only difference being the wire guage and number of turns. What's important is the amount of copper within the winding slots as a higher percentage of copper = a lower resistance for the required number of turns to reach your desired rpm. For example if either of these steppers were designed for 200 volt operation they'd have hair thin wires in them. They'd operate on very low currents for low I^2*R losses too. They'd just need all that voltage to work. If these two motors have the same number of turns then the NEMA 17 will be using a longer wire to allow for the same number of turns of thicker wire. One will increase the resistance and other will decrease it. One thing is for sure though, if you heatsink the NEMA14 so that it doesn't overheat you most certainly can crank up the amps. For that matter the circular design looks a lot easier to cool as the windings are just under the surface of the motor.

[@justinwalker8364]

After all these years you still create problems that don’t exist.

[@DesignPrototypeTest]

???

[@justinwalker8364]

@@DesignPrototypeTest all the talk about changing the design, gear ratio this and that - talking about how hard it is to hand feed filament through - these issues are not real issues for normal people.

[@DesignPrototypeTest]

Troll.

[@guidodumont589]

Thank you for the video, super interesting! I follow your channel for quit some time and personally working also on a heated encloser with an DIY IDEX 3D printer. I hope to print PC one day just like you do, but wonder where you buy PC?

[@RKisBae]

The youtube gods seem to think I need more of these videos and I agree, I've really been enjoying the detailed analysis of these extruders. I found you trying to figure out the best extruder setup for printing 60A. My search has led me to Bondtech LGX, Biqu H2, Hemera and more, the only thing that works so far is a single gear Diabase Flexion setup but it has terrible extrusion ratio (40%), slips unpredictably and has an inconsistent tensioning mechanism. I'd like to know if you have figured out any ways to fully constrain and successfully push a wet noodle like 60A.

[@DesignPrototypeTest]

This extruder has the tightest filament path I've ever encountered. It might be a winner given your needs. Also, perhaps you should switch to 2.85mm filament. It will be stiffer because of the increased diameter.

[@notsam498]

my only issue here is weighing the extruders with the wiring on the scale? a good deal of the wiring isn't involved in movement on the hotend carriage. I could easily see discrepancies of 5-10 grams from the wiring alone being realistic. especially since one manufacture might include a little more wire than the other.

[@siwiskate]

Rants about china in every single video... Then proceeds to get more products from China...are we sure this is not just publicity? There's no such thing as bad publicity they say...

[@xSNIPRIDERx]

Maybe flip the hobbed gear 180°, move the posiotion of the filamentintake to the front then you are able to remove the frontplate and assamble and deassamble it from one direction. Also would remove the thumbscrew and take take a M3 rod and glue a 3D Printed thumbnut on it😉

[@kenlipper874]

A rubber glove is great for gripping PTFE as well.

[@dduncane]

To be honnest, I am interested in a fair test between those extruders, but I'm not sure you need to do a video about all of those tests. An Excel sheet to summarize the results that you can refer too (like the one you have for printers) should be good too (for those you have already tested).

[@TsukiShimizu]

If you are interested in (some) results based on calculations - Annex did some really nice force graphs regarding the pushing force of the extruders. github.com/Annex-Engineering/Annex-Engineering.github.io/blob/master/extruder_force.md

[@JoshTownsend66]

You mean annex who actually designed the sherpa. That one?

[@cil7ea]

KZbin is like its own bubble, 1 year behind what's actually happening in the community :)) Stefan and Thomas discovered Voron not too long ago, now this guy discovers the sherpa mini... TeachingTech made a video about input shaper with an accelerometer (probably thousands of dollars lost because of that). You're all late to the party guys... youtube creators need to step up their game, join some frikin discord servers where actually smart people actually push the community forward.

[@DesignPrototypeTest]

Your comment operates under the mistaken assumption that the bleeding edge of technology adoption is all that matters. The thin tail of the bell curve where early adopters live does not have much to do with what will eventually be successful in the wider context. In the case of Tom and Stefan, they are only now making Voron videos because the groundswell of popularity for those printers is such that they have determined it will be a profitable use of their time to make and film such content. They are shrewd tacticians choosing carefully which projects to align their channels with based on what will help them be more successful. I do no such calculus. I talk about the things I'm interested in. Full stop. This can partially explain the differences in "success" between my channel and CNC Kitchen despite being started around the same time. Sorry if I didn't get here first. I was busy making videos and figuring things out on my own so I didn't have time to read your forums. If you would like to help out the channel, perhaps you could tell me about the hot new things when you first see them. Have a good day, and please don't be so insulting. It makes it hard to find the value in your comment.

[@johnkim3858]

Although existing, to be completely honest, bleeding edge technology development move so quick that if a KZbinr makes a video about it, it seems to get irrelevant not really valuable within few weeks. Also, making 20 minute video takes days of work. Due to KZbin rewarding sensational videos rather then videos that opens design discussions it is hard enough for the KZbinrs making such content.^^

[@ale6242]

@@DesignPrototypeTest I think a large part of why I prefer Stefan's videos to yours (no offence, yours are still good!) Is the "scientific" approach he takes towards his testing's. Not only that, but he actually seems to do his research on the current version of whatever product he is using. You've now made 2 videos on outdated designs that actually provide no real help on the current version as the criticisms you made are no longer present on the latest versions.

[@DesignPrototypeTest]

@@ale6242 The Original Ender 3 is still for sale. The Prusa Mk. 3 is still for sale. These printers dominate the online conversation still. I bought what any reasonable hobbyist would buy. Most of my criticisms still apply to the "updated" versions of the extruders. You can't update a physical product the way you can a software program. You need to release a whole new thing that looks different and has a different name. You need to get it right before showing it to the world. This is what China does not and will never do; test their own products. R&D is laborious and expensive, but without it you get the plethora of crap products like we have in the hobby of 3D printing. When the original and the "updated" are both available on the internet from all kinds of sellers, and brand names don't mean anything coming from China, you can't expect people to buy the "right" one. Also, is there even an updated Sherpa Mini? How is it different from the one I tested?

[@ale6242]

@@DesignPrototypeTest well you make a fair point in that this is what a hobbyist would buy. Youve made a great review of this specific version from China. But as a few people in this comment section have pointed out the model of extruder that you claim to be reviewing is still being developed and there is no "official" kit, you are supposed to print the latest version yourself. You've put a bunch of negative marks against a project because of a manufacturer that is making a flawed version of the current design. Calling this a review of that projects end product is like taking a kit car and comparing it against a competitor to the car your kit is modeled from.

[@81Capman]

hello thanks for video . I have a problem with mine . TPU slide when extruding . And even when the wheel is tight, the TPU slips. I can even get it out by hand by pulling lightly. On the other hand, no money with pla petg and other hard filament . do you know what could be the problem? Because I disassembled several times and there are not 50 ways. I put a piece of filament to center the gears but it the same problem . Thanks

[@radicalxedward8047]

just out of my own confusion, why is the orbiter the only one weighed while not holding the wires?

[@johnsmith-000]

Why not just turn the shaft of the square motor to smaller diameter? I mean if you think there's a real probability of considerable improvement, this should be a relatively simple operation, whether you do it or have someone else does. Adding 50g seems a pretty bold strategy to me, though. But I'm by no means an expert, and maybe there's still something to be gained...

[@williamrobertingram]

Hey Yo...Ive been messing around w CD laser diodes hanging offa my xruder. Might be on to something w a layer adhesion problem super solution.

[@DesignPrototypeTest]

Nice, Pre-heating the under layer could lead to better layer adhesion. If you get it working, you should definitely send some samples to Stefan at CNC kitchen and see if he will them on his testing machine to determine if it adds strength tot he parts.

[@johnkim3858]

Is the idea basically a lightened up 5:1 BMG?

[@johnkim3858]

I think the idea is coming true. They are calling it the Sharpa Heavy ---> cdn.discordapp.com/attachments/697049948222390282/853592908213256212/image0.jpg

[@mikoaj5850]

Is there any way to contact you, other then patreon/youtube comments? I wanted to talk about prusa drama - I found your channel trough TTs recomendations (yup saw that you ranted on him too, most arguments were pretty true, but I think its better now) and saw that you really know what you are talking about. I can see that you are smart and know very much about 3d printing, but prusa drama videos are just weird. I dont know if they are a sick joke, satire or what you truly think. I just want to get and understand your point of view and maybe change my point. Also: 1) sorry if I type something wrong, english is not my first language, 2) Im not from some czech comment farm, Im pretty close in poland but that does not mean that I work for prusa. Im not even legaly able to work so... Edit: checked and tt doesnt recomend any channel so it was other channel that I found you trough. I think thats not important but maybe you find it useful for spreading your channels reach

[@DesignPrototypeTest]

Click on my channel, then click on the "about" tab. Click "I'm not a robot" and you should see an email address. The short version of the Prusa story is that Prusa copied all the innovations on his printer from other hackers on the internet without giving attribution/credit. Starting with the Mendel frame design which was created by Ed Sells. I believe that he got the idea of putting a temp. sensor in his P.I.N.D.A. probe from one of my videos because when I met him at the Bay Area Maker fair he insisted that he knew me and had seen my videos. When I emailed him at the address he provided to me he didn't answer instead Shane respond to me (I interviewed Shane instead of him at the BAMF on camera). I asked for a printer to review. They refused. I made a video where I essentially took credit for giving Prusa the idea of a temp sensor inside the P.I.N.D.A. probe. Prusa's response to this was to sick his entire cult and sock puppet farm upon my channel. I've been harassed by them ever since. All I wanted was a printer to review to help my channel grow. Instead I got my channel made into the enemy/bad guy of the 3D printing "community."

[@RentableSocks]

give a caveman a lighter and it may take years for him to figure out it'll make fire.

[@toast47624]

I'd like to understand why you say the motor was too small when it was able to exceed the typical extrusion speeds? My intention is to build the ultimate Poly Carbonate printer 500mm on the X with 300mm Y and 300m Z. It will be double glazed and fully sealed thanks to you. In fact you have pretty much listed what the ultimate printer will look like. I can't justify buying one but I certainly can build one. When I saw this video I had hoped it would be the answer to my print head but you destroyed that in a positive way. I am in no way as experienced as you using 3D Printers so you have saved me a considerable amount of time and I'd like to help you in some way with one your projects at some point in the future as a way of saying thanks.

[@DesignPrototypeTest]

>I'd like to understand why you say the motor was too small when it was able to exceed the typical extrusion speeds? Because the whole point of a lightweight extruder is to print quickly. 300mm/s is a decent goal. currently this extruder/hotend combo can only get about half way there.

[@toast47624]

@@DesignPrototypeTest Wow 300m/s. Ok, I did not know that was even possible. Just homing in on my targets at this stage. Can the Duet run multiple X,Y,Z steppers? I don't really care how much this thing costs I just want it to be the best I can build. The hot end will require possibly 2 elements.

[@victor_glekler]

How about using this extruder on a delta 3d printer? Does it make sense?

[@rfpeace]

What do you think about using a larger/heavy-duty stepper extruder at the filament roll as you would with a typical Boden tube set up, to assist a lighter duty steeper on a direct extruder?

[@georgekatsinis5224]

My guess: Unless those two steppers were EXACTLY in sync, you would be struggling with one pushing, and one retracting. Even for milliseconds, that's an issue. AND you'd be adding the disadvantages of a bowden tube between spool stepper/extruder and hot end stepper/extruder. AND you're adding the cost, and wattage needs of the additional stepper/extruder . Interesting idea, but I do not think it will work.

[@rfpeace]

@@georgekatsinis5224 good points dude!

[@4DR14N0]

What is the name of the software you use on the tablet to execute commands?

[@DesignPrototypeTest]

Google chrome. The Duet3D printer control board running reprap firmware essentially acts like a webpage server. You log into it through your local area network using your web browser. This gives you full control over the printer, firmware and all.

[@v1Broadcaster]

Lgx lite is out now

[@CCCfeinman55]

Yes, please. I note here that if you use the pancake motor with its larger shaft and resize the second gear to maintain 5:1, you may end up with your manual (exposed gear) advance. Great video and superb analysis of the device and its advantages and shortcomings. Truly, this kid begs for a redesign to superior operational capability. So, would you be using sos to print your redesign?

[@darkmask123]

mm / min or mm /second ????

[@Microplastic_Therapy]

This extruder makes my hemera look “obese”

[@librasd8087]

Hi DPT, may I ask what happened to the e3D Tool Changer? you only made two videos about it, I thought after spending so much you would make more

[@DesignPrototypeTest]

Yeah, you would think. It just sits on my shelf. I do intend to get back to it some day.

[@librasd8087]

@@DesignPrototypeTest I hope you will be able to make some videos about it soon, I know it needs work but it can give satisfaction

[@doron910]

please review sprite extruder

[@Exstaz]

A coreXY does not need the bed to move in z. A printer like voron 2.4 are also a coreXY but bed a rigid Mounted.

[@nirodper]

yes, but the point remains the same: the heaviest components on the slowest direction

[@NicksStuff]

9:33 You should not open a stepper motor, it will never go back 100%

[@DesignPrototypeTest]

What? Source? Why would this be true?

[@mariuszczerwonka5249]

where to get mosquito to wherpa adapter?

[@DesignPrototypeTest]

I made it. Go become a $1 Patreon supporter of mine and I'll give that to you. Normally I share all files with $5 supporters but since all you want is the adapter.

[@baljazz]

Do you have any videos of it printing? Does it work on CR10s Pro

[@saritsotangkur2438]

If China changes the design to improve it for SLS, then can they still call it a Sherpa extruder? If they then call it something else would you then complain that they “stole” the Sherpa design? I think there’s nothing they can do to avoid your wrath other than to do nothing.

[@DesignPrototypeTest]

Yeah, I'm the problem. It's all me. The tiny KZbinr who calls out other people's bad behavior. I'm totally to blame.

[@proto_hexagon5649]

what NEMA 17 get? manufacutrer company?

[@spreaddeezhazelnutz7511]

I could picture throwing this on a fast delta like the FLSUN SR.

[@Domi1988]

Thats my plan as soon as my SR arrives.

[@spreaddeezhazelnutz7511]

@@Domi1988 I’m excited for when those printers will have a lot of aftermarket support. Once I get mine, I want to set it up with a 300c DyzEnd with their 0.3mm tungsten carbide nozzle. I still can’t get over how awesome that printer is. Can’t beat the price on it.

[@subuser9627]

Turn the axis down on a lathe ?

[@sjoervanderploeg4340]

When I try to swap the Numa17 with a pancake stepper it always skips :(

[@doron910]

I get rough layers with the sherpa and before it I had perfect prints. Anyone knows what could be the exact source?

[@livedeliciously]

E steps or retraction settings?