Tevo Seriously Need to take a hard look at themselves

General topics regarding the Little Monster
Post Reply
Posts: 18
Joined: Sat Dec 23, 2017 7:25 am
Location: Australia
Been thanked: 3 times

Tevo Seriously Need to take a hard look at themselves

Post by Redemptioner » Wed Jan 03, 2018 3:01 am

I wanted to start a thread about the major quality issues and the poor choices Tevo has made in putting together this Delta system. Hopefully Tevo will take a look at this thread and correct these issues, I am certainly more than happy to lend them some of my time as an engineer to work on resolving them. for the most part these issue are about taking a little more "care" and not using really poor quality counterfeit parts.

The overall design of this product is great, so much so that i have bought several of them knowing that there are some things that needed to be fixed in order to get a decent printer, which is not surprising seeing as it is a direct rip-off a great design put together by one of the main developers of the Duet.

So let's get into the bowls of this machine and look at what needs to be corrected;

Main Control Board
My main bugbear with this printer is the main board, to put it simply is it complete rubbish and I am pretty sure a hotcake from McDonald’s would do a better job. This clone MKS board, well let’s be honest, it’s not even a clone it is a poor counterfeit as it does not meet the any of the license requirements for a clone, just does not work well, if at all. The board suffers from poor quality components, at lack of heat regulation and really bad shielding from electrical interference. The Ethernet port is basically useless and their own literature states not to use it, the USB port uses the main processor to run so anything coming in over affects the ability to print and so does the LCD monitor. The WiFi kind of works some of the time but there is no instructions on setting it up and again is utilising processor overhead in order to operate and has little range so don’t expect it to be working in a house with concrete or brick walls without the router sitting right near it. The SD card slot slowly corrupts the SD cards so you need to ensure you have backups or suffer the loss of the config and override files. One of things I truly hate about the board, apart from inconsistent results and the random reboots or freezing, is the fact the main processor is interrupted anytime you touch the LCD panel. You want to decrease/increase movement speed, change the temp, adjust the extrusion rate etc, well every time you touch the panel the printing stops for a split second and this creates artefacts and misalignment in the prints, just another indication of a board not really up to the task.

The heatsinks for the stepper motor drivers I don’t think have actual thermal tape holding them on and are on the wrong side of the chip to be doing much good anyways. This is compounded by the fact all the heatsinks are all tied together ensuring any heat that is picked up by them is shared over all the chips heating other drivers unnecessarily. To add to the heat woes is the fan being mounted the wrong way round in most cases, sucking air out of the case along all the other components before it can get anywhere near the heatsinks (no slots for air on that side of the case) instead of blowing cool air in (a super simple fix) and the wiring is so poorly routed that they block the majority of the airflow to begin with.

Surely with this being the most important part of the entire printer you could spend a few more dollars and replace this with something reliable like the Duet WiFi. Changing over to Duet WiFi board is a night and day difference in the quality of the prints and the repeatability of these prints. This change alone fixes nearly all of the major printing problems (aside from other quality check issues to come later), it really is a breath of fresh air for this printer and makes the difference between hours of frustration and it simply working. Now don’t get me wrong, the Duet board is worth 3 times the price of the Tevo counterfeit MKS board, but considering they have been giving a couple hundred dollar discount on this printer lately there is surely room for Tevo to buy the Duet’s at wholesale and be putting them in just for reduction in customer service & support requests, let alone the improvement to their reputation or the improvement in the printing.

Like the main control board this counterfeit hotend is a complete waste of everyone’s time. You know the quality is going to be pretty bad when spare parts are provided in the box for this item. My biggest issue with the hotend is the fact it is sold as “all Metal” when clearly it is not even close to being all metal, this is particularly bad when they advertise the printer for use with high temperature filaments like Nylon . What is made worse is the PTFE lining in the hotend isn’t even round, I am not sure if this is because of the super low quality of the product or if it is the method they have used to get the lining into the hotend, add to this the fact that many filaments adhere to extremely well when they cool and you have a recipe for poor prints. Now take into account that this lining also starts to become unstable at 200*C and seems to leach “substances” at around 230*C you have a recipe for disaster. I would hazard a guess the inconsistency of prints and the majority of the clogging people are having is a direct result of the poor quality hotend. Again we are only talking about $70USD to use a decent hotend, less if you are buying them wholesale in bulk so surely this is another area you should be investing a few more dollars on.

Stepper Motors
The stepper motors used in the Tevo Little Monster are also on the scale of “complete rubbish”. There simply is no reason for not using decent 0.9 Degree stepper motors as standard on delta considering they are basically the same price as 1.8 Degree motors and produce roughly the same torque. But this is only the tip of the iceberg for the problems with them motors, they are also unbalanced (or very poorly balance, although I can’t see any balancing “goop” on them) with many of them having the shafts off-centre or not straight in line with the stator and axis of rotation. All these problems add up to vibration, differences in torque depending on the step point and way more electrical noise than they should be making. When you combine this with stepper drivers that are low quality and not setup for driving these motors (no tuning done for current and voltage) you get a lot of noise and inaccuracy in your prints. For anyone that has done a large print with infill (not the ridiculous 2 print reviews done on the Utubes in vase mode) you know what I am talking about, not only do you get the “salmon skin” effect, you also get misalignment of layers and various other layer artefacts. Considering the difference in cost between rubbish motors and decent motors is $2-$3 a motor there is simply no excuse for putting such low quality motors into this printer, NO EXCUSE!

As for the extruder motor, it is WAY oversized and adds 3 times the weight it needs to it considering it has a massive gear reduction so there is simply no reason for it being so big. Dropping the size and therefor the weight of this motor will mean far less mass swinging around above the effecter resulting in higher quality prints and easier to setup (and more tolerance to the setup) the tension on the belts holding the motor and extruder in place.

Print Bed
The print bed is said to have a 350mm print area which is complete rubbish, the bed isn’t even 350mm wide let alone the fact the print area is reduced down to about 270mm due to the heating element only running out to about 25mm away from the white line, then even out it’s perimeter the heating element is around 15-25 degrees Celsius cooler than the middle. This is just false advertising at best and complete fraud in general as much of the decision people make around 3D printers is the build volume and this is clearly not able to do 350x500mm. I also don’t think there is any excuse for exposing 240V/110V out of the power supply to the heated bed. I understand in some countries it is all about “safety Third” but it simply is not good enough for a couple of thin wires which some basic, cheap and nasty cloth covering to be carting this kind of dangerous voltage down the side of the a metal tower down to the build plate, just unsafe BS, what were you thinking. Also the temperature sensor is just taped onto the bottom, centre of the heating element which makes it super inaccurate, this combined with the heating element not going all the way to the edge of the build area means you really need to give the machine a good 30mins or more to heat up the build plate to not only get an semi accurate level calibration but also to be able to print anything larger than 100 mm in diameter.

The heat bed is glass which then sits atop an aluminium plate that holds it in place and uses some cheap silicon to hold the glass plate to the aluminium support. The silicon is not heat safe or heat resistant which means a few prints in and your bed is coming up off its retainer if it didn’t already come out during shipping. The aluminium plate has a chamfer for the plate to sit in which at least it has that but it is aluminium which means it draws the heat out of the edge of the plate which compounds the fact the heater element does not come all the way to the edge or heat evenly. It also means the aluminium plate warps when it heats up (a well-known issue with aluminium) resulting in the uneven bed height and shape being further compounded by the already extensive uneven heating. Add to that aluminium stand offs holding the bed off the base that are not adjustable or even in height and chuck in a completely inconsistent BL Touch sensor that can’t be used to calibrate before each print and you will have no end of trouble with inconsistent adhesion of prints to the bed. This would all be fine if bed was adjustable in height and there was a goof insulator between the aluminium retainer and the glass bed and a far better heating element that actually went all the way to the edge, even a reliable height sensor would make a big difference here on the consistency of print adhesion. Personally I find the counterfeit “Buildtac” to work really well when you account for the issues above but I clearly understand why people add and additional glass bed on top, not only does this give a much greater build volume they can get a more reliably flat/even build surface and move/rely on adhesives on the glass to attach their prints.

Auto Calibration
This is a feature that, in my opinion, is a must for delta printers even though it doesn’t get around all the manual calibration it does drastically reduce any further ongoing manual calibration requirements. Unfortunately they implemented the BL Touch sensor as the solution for providing the bed level sensing, which is a big mistake on a fast moving delta printer. BL Touch is a product designed by hobbyists and made by hobbyist and also that steals its design from an open source project developed in the community. What do I mean by “designed by hobbyists and made by hobbyist”, well the sensor has a cheap crappy Hall Effect sensor which has low accuracy and is compounded by relying on a small piece of heatshrink to set the length of the probe pin to the Hall Effect sensor trigger. When you place this item in front of a heat source (i.e. in front of the hot air coming off the heat break) the length of the trigger probe changes due to the heatshrink changing size dependant on the temp and this change is neither predictable nor is it consistent so you can’t even compensate for this. You then add to this a super poor coil setup for holding the probe up which is also massively effected by heat (reduces it’s effective hold strength) and you have a sensor probe that is not only inaccurate, it will also drop the pin out randomly on fast movements as the sensor heats up.

There is nothing like the feeling of ruining a 15 hour print after 14 hours or having a print run overnight only to wake up in the morning to a smouldering mess of goop to clean up. Even a blind engineer would tell you using a bit of heatshrink in the manner they do is not going to give good results and the difference I have measured can be well over 0.1mm which can be the difference in multiple layer heights. Just about every time you run the auto bed level calibration with this sensor you have to readjust the Z height offset to correct for this inconsistency, so not running this check before every print like you should be to compensate for the small changes depending on what temp you are running the bed at for the particular material you are going to use. Now BL Touch could resolve some of this issue by simply making the probe pin ever so slightly larger, putting a thread on the end and screwing the Hall Effect sensor trigger onto the end of it, considering they are selling a $4 product for $40 there is clearly plenty of headroom for this small improvement that would drastically improve the quality of the product, another $2 on the Hall Effect sensor and another $0.50 on some more copper windings on the coil and you might have something worth using.

The Tevo Little Monster has a black bed, it has a black “buildtac” sticker as well so why not use an IR sensor instead of the BL Touch. They are not only cheaper, they are more accurate and they do not rely on something actually touching the bed so no issues with heat expansion causing you woes or pins falling down into your prints, they are even slim enough to be mounted on the other side of the hotend away from the direct heat. If you upgrade to a Duet, which I strongly recommend you do, then make sure you add the IR sensor into the shopping cart while you are there.

There is a major component to any 3D printer and that is cooling but Tevo seems to have given this “meh, it’s good enough” approach to this requirement in their design. It starts with the cooling on the main board where they mount the fan the wrong way up sucking air out of the case instead of blowing it in, that and there is no dust filtering which means everything is going to get coated in plastic particles and plastic cobwebs in no time. They also don’t put any vents on the side of the case where the main heatsinks, over the stepper motor drivers, sit nor do they do anything about containing the wiring which pretty much blocks the airflow on the side of the case with vents. A 120mm fan filter is less than $1, so flip the fan the right way around and add a basic fan filer (I suggest one that has a removable foam pad that can be cleaned) just to start with, add some additional vents to the case or change the side the vents are on and add some cable management to keep everything out of the airflow so it can do its job.

On the hotend there is also come issues, the 30mm fan is not only noisy as hell, it also lacks a fan filter/guard. This fan is the most critical fan on the whole printer and is most susceptible to getting particles and more importantly filament strands in it reducing the ability for it to cool the heat break or stopping it working altogether. A small mesh guard over the intake would go a long way to towards stopping any issues with this fan. Now on the other side of the 30mm fan there is nothing forcing the airflow through the fins of the heat break, this means 80% of the cooling fans ability to cool the heat break (the reason it is there) is not being used. Now personally I think heat breaks should have their own temperature sensor and have the fan speed controlled by the main board in order to keep the noise down and improve the lifespan of the fan (as in most cases if the airflow is directly appropriately “through” the heat break the fan only needs to run at half speed). The air, after it has passed over the heat break, also needs to be directed up and away from the print area so as not to effect the printing or control of the cooling at the nozzle, admittedly nearly all heat break cooling solution on basically every 3D printer fails in this area but it is an easy fix (and if they insist on using the BL Touch maybe a shield around the BL Touch to reduce the effect the hot air is having on its accuracy). The side 40mm fans are great little fans, lots of air, pretty quiet and have a decent static pressure rating so are good for ducts. Unfortunately Tevo decided not to put any actual ducts onto the fans themselves leaving them pump air on the area you don’t need it. When ducted correctly the fans really never need to go above 60%, this means they are near silent and allows for some decent ducting to be used. I am personally a big “fan” of ducts that use a combination of positive and negative air pressure cooling as this not only reduces the effect of the air pressure pushing around the filament as it is extruded, it also keeps the temp of the air up a bit higher preventing it from cooling the extrusion too quickly or too much causing poor adhesion of warping. The last issue with cooling on the hotend is the fact they mount all these fans on an aluminium mount which is in direct contact with the bottom most layer of the heat break. This means a lot of heat is absorbed into the fan mount (especially if the cooling fans are not on) causing any fan ducts you have printed to warp or melt unless they have been printed out of a high temp plastic. The irony is you need additional printed fan ducts in order to print the fan ducts in these higher temp plastics so it can be an iteration of ducts before you get a permanent solution if you don’t have another 3D printer to be printing ABS or Polycarb on.

The motors are another area that suffers badly from heating as the main board is rubbish, with the stepper drivers being old and not setup correctly, combined with poor quality motors means the motors heat up way more than they should. This creates a reduction in torque, change in shape and most importantly a reduction in accuracy over time. This means any prints that run longer than an hour start to suffer in quality and the longer they go for the more they are impacted, but not only that, it drastically reduces the lifespan of the motors. Changing the motors for decent ones along with the board is the sure-fire way of correcting most of this, but adding $1-$2 a motor in heatsinks has a massive impact on this as well.

The extruder is another counterfeit product done really poorly, the main cog for feeding the filament doesn’t even have a groove in it to align and better grip the filament. Not a hard or expensive item to source/create so there is not much excuse for this and it would go a long way to improving the consistency of the feed rates. The bearing holding this feed cog in doesn’t have a spacer to keep it from touching the side of the clear cover (at least it has a clear cover), this means you have to play with getting the right amount of tension to stop movement in feed cog and preventing it rubbing on the casing. This is also not a sealed bearing so it needs some lube, but there is also nothing on the other side of the bearing preventing any lube you might add migrating up the feed cog to the filament and getting into your print layers. Metal on plastic gears is simply engineering 101 “no no”, the metal gears start to eat into the plastic allowing “wobble” in the gear surface and inconsistent filament feed. Anyone that has used their Tevo Little Monster for a while will knows the sound the extruder starts to make when it is cold, it is hard to describe but you all know what I am talking about, well this is that “wobble” in the gear teeth from the metal on plastic, lube will help but only delays the problem.

The Bowden tube holder on the bottom has no threads in the extruder, it is simply jammed in with some thread tape and then a prayer is said to hold it in place. The extruder is held down to the effector by the Bowden tube with pressure added to this tube by the belts and springs to stop the extruder moving about. A “who gives a toss” press fitting is not good enough for this, it needs a proper fitting and this is only $1 so why not add it, you have clearly had this moulded so why not get it moulded with a thread.

The mount holding the extruder and motor is not aligned correctly, you put a clear front on it then have it twisted to one side so you can see in it anyways. Seriously WTF guys, I mean really….. who dropped the ball here? The mount also needs to be made much wider to reduce the angle the belts feed back to the carriages on and the recommended 150mm Bowden tube length is not long enough. If you up the quality of the Bowden tubing, so the diameter inside is a tighter tolerance to 1.75mm, and then raise it up to 200-250mm long it makes all the difference and won’t affect the filaments you can print with while making it much easier to setup the mounting and remove the effect it has on the quality of the finish on prints.

As mentioned earlier, the motor on the extruder is way too big and heavy, reduce the size and weight, up the quality and you will probably still come out with a cheaper motor to purchase.

Pulleys & Belts
Another WTF moment here and serious sign of the poor quality checks being done as well as how BS the main stream reviewers are on the Utubes (Sorry Tom, you lost my respect due to these kind of things not being noticed). The rather glaringly obvious issue is the motor pulleys and idler pulleys are too larger for the rails, this means the belts rub on the carriage wheels. As you can imagine this is going to quickly wear out the belts and wheels and put flat spots and groves in these parts. This adds additional friction and noise to the motor movements which changes over time, is dependent on the speed of the arm movements and the height of the build (as more wear will be in the lower half of print volume creating deeper groves in the belts), all this leads to poor print quality that will vary and increase over time, reduced lifespan of parts and generally add frustration. In most cases the belts are put on with a twist in them (multiple unit s of mine have had twists in at least 2 of the belts), the belts are of low quality rubber which stretch with heat/ time and they have used the most cheap and nasty zip ties to attach the belts to the carriages leaving them to break once a little tension is added. Here is a tip TEVO, zip ties are not a permanent building product and should NEVER be used to attach something like a belt that has to be placed under tension.

The idler pulleys are also mounted on a bolt instead of a bearing cross member (should be smooth without threads on the part going through the bearing and fit snug), this means there is play in the bearing on top of the bolt allowing wobble in the idler pulley adding more noise and wear as the bearing is not spinning on the ball bearings inside all the time, they are also rotating on the shaft of the bolt…..not good. There is also some spaces required that are not supplied to keep the nut and plate away from the bearing, this is very noticeable as you start to tighten the nut down holding the bearing on as causes the bearing to rub against the plate and nut.

Suffice to say there needs to be a serious rethink about how the belts are attached, spend another $5 on the quality of the belts and another $2 on changing the pulleys to something that is small enough to ensure the belts clear the rails with the correct cross pin for the bearings and matching spacers to keep the bearing off the plate and nut.

The frame is rather tall and is only held in shape by some bolts going through aluminium plates into aluminium extruded uprights. As mention earlier, aluminium suffers uneven shape distortion as it heats up as well as being a rather soft metal, this results in the whole frame warping and moving everything out of alignment. A delta printer relies on pretty tight tolerances on alignment of the towers and lengths to ensure accuracy, changing the top and bottom plates out for steel should help here (or even a set of steel cappers) and ideally some tension cables to be able to pull the entire frame into alignment and held keep everything super stiff. A set of crosses on the back 2 sides and a couple of vertical ones on the front with a small top and bottom cross cable set would make it a breeze to pull the entire tower into perfect alignment and prevent the heat distortion and any lean created from not being perfectly level. Ideally you would include a jig that would sit inside the towers you could use to measure “squareness” and distance between towers to help get this perfect without the need to measure, probably another $10-$15 worth of parts but it would improve the higher print quality and layer alignment drastically.
You do not have the required permissions to view the files attached to this post.
If it ain't broke then modify it

Posts: 1
Joined: Wed Jan 03, 2018 4:40 am

Re: Tevo Seriously Need to take a hard look at themselves

Post by bobog112 » Sat Jan 20, 2018 3:31 am

This is a good start for an upgrade thread, if only it had recommendations for the upgrade parts.

I know I am probably taking the long way around on upgrades.

To start the train of thought off:

E3D VOLCANO - https://www.filastruder.com/products/e3d-volcano

E3D TITAN EXTRUDER - https://www.filastruder.com/collections ... n-extruder

Stepper Drivers
SILENTSTEPSTICK - https://www.filastruder.com/collections ... tor-driver

Driver Control Extension Shield -
Just search on Amazon for: Printer Stepper Control Extension

I am currently researching what and where to get Stepper Motors but this is an in-progress upgrade. Eventually I would like to go down the Duet path, but the cost and time to learn is something I cannot take on right now.

User avatar
Posts: 87
Joined: Tue Mar 28, 2017 10:56 pm
Location: Australia
Has thanked: 35 times
Been thanked: 24 times

Re: Tevo Seriously Need to take a hard look at themselves

Post by royg » Wed Jan 24, 2018 12:29 am

As an alternative supplier for E3D components, I'd recommend buying directly from https://e3d-online.com/ who are the people who invented and who continue to develop the E3D range of 3D Printing products. I don't know about "Filastruder.com" - they may sell genuine E3D products, but you need to be careful as many suppliers, especially in Asia and in other pleaces, sell clones or even worse, fake products, labelled with E3D's brand.

I have no association either E3D UK - just a happy customer of the genuine E3D products, after having been caught out badly buying cheap clone and fake E3D products from Asia and Europe outlets.

Tevo Tarantula - Large Bed - E3dV6 - BLTouch
OctoPi (on a Raspberry Pi 3B)

Posts: 18
Joined: Sat Dec 23, 2017 7:25 am
Location: Australia
Been thanked: 3 times

Re: Tevo Seriously Need to take a hard look at themselves

Post by Redemptioner » Sun Feb 04, 2018 11:03 pm

Well Filastruder is a genuine product supplier of E3D and is stated and linked too on E3D sites so there is no concern to be raised about going with them. Let's no cause any misinformation there. I have personally found Filastruder to be far higher in customer service than E3D directly, they are much faster to respond and do a lot more 3D printing as well as filament production and as a result have a lot more knowledge about both E3D products as well as other products. They also personally check each item before shipping and most orders come with a hand written note thanking you for your purchase and providing some proof that your order is being looked after at a personal level.

The thread is aimed at letting Tevo know what they need to do in order to correct their product, I have started writing up the modification/upgrade thread for changing out all the parts to bring the printer up to a pro standard. Keep an eye out in the coming days for this as it will list all the genuine items to purchase, how to go about it and the all the additional parts you will need to complete it (wire types, tools etc). It will detail rewiring the whole unit, changing motors, controller, hot end, fans, pulleys and belts, z-probe and modifications to things such as power unit and controller housing, heated bed as well as adding a heated enclosure. viewtopic.php?f=27&t=393
If it ain't broke then modify it

Post Reply

Who is online

Users browsing this forum: No registered users and 1 guest