Testing 5 different heated beds!
ฝัง
- เผยแพร่เมื่อ 30 เม.ย. 2019
- No two heated beds are the same - in this video, I test which beds have the best chance of getting your prints to stick!
I use a thermal camera to assess heatup times, temperature offsets and how even the 5 different heated beds heat up!
FLIR One Pro thermal camera go.toms3d.org/flir
Ultimaker 3 go.toms3d.org/UM3US/
Creality Ender-3 go.toms3d.org/Ender3Ali/
Original Prusa i3 MK3S go.toms3d.org/PROOSHAA/
BCN3D Sigma geni.us/OneSigma
Lulzbot Mini www.matterhackers.com/s/store...
Product links are affiliate links - I may earn a commission on qualifying purchases (at no extra cost to you)
🎥 All my video gear toms3d.org/my-gear
I use Epidemic Sound, sign up for a 30-day free trial here share.epidemicsound.com/MadeWi...
🎧 Check out the Meltzone Podcast (with CNC Kitchen)! / @themeltzone
👐 Enjoying the videos? Support my work on Patreon! / toms3dp - วิทยาศาสตร์และเทคโนโลยี
![[Full] 4 ต่อ 4 Celebrity EP.876 | 5 พ.ค. 67 | one31](http://i.ytimg.com/vi/m6eKqTfpx0g/mqdefault.jpg)
4th wall: exists
Thomas Sanladerer: it's free real estate
Being somewhat of a thermal imaging enthusiast myself, I got a few comments.
First of all emissivity. You said correctly, that glass is reflective, but so is virtually every other material. If you want a good comparsion you need to set the correct emissivity for each material you are measuring.
Next accuracy. The FLIR One series may say 70mK as NETD - but that is relative and far from real world use. The absolute temperature reading is within a margin of +/-2K or 3% whichever is greater. depending on which Lepton you have and gain mode used.
I hope the next part includes watt measurements of how much energy they take and deliver to the top, e.g. how much are hey heating the air at the bottom and edges.
The sensor needs to heat up itself to give the best possible readings. The FLIR One also uses a shutter to calibrate itself regularily. You can turn this function to manual and wait 5 minutes with the camera running to get the best readings. To get consistency, try to put a piece of dark tape as a measuring pad on all the beds. As emissivity will be consistent this way.
Presentation is god awful. Please try harder to center the bed and let it fill the frame resolution is low and it doesn't serve a good comparison if less than 30% of your pixels observe the bed on one shot you did. I know that the Iron palette looks most recognizable, and that's partly he reason I finally watched a video of yours again, but it's not great for what you want to show.
It looks like you used the timelapse or video function in the stock FLIR One app. Try to shoot images instead, they are radiometric and can be analyzed in either FLIR Tools or third party software like the great Joe-c Thermovision.
If you want to do this in a proper way, take radiometric images(the maximum framerate is 8.6hz anyways) and normalize them. Use a box measurement with high, low and average instead of a single spot. Use a fixed palette from 20 to 70 like the rainbow or medial on hand and turn off Histogramm equalization to really show a development over time.
It can look much better and I've meaningful data, even with your low end imager.
Do your own video
@@sickvic3909 he doesn't have all these printers
@@sickvic3909 Write your own feedback
I love how he keeps his tea warm on the ender bed... i have not thought of that XD
Your camera work is getting up there, Tom. Great video!
I really like that transition.
I'm not sure using the FLIR to measure the variance between reported and actual bed temperatures is a good idea. The thermal emissivity between the different materials would invalidate the measurements unless you explicitly calibrate your camera for each material.
Squares of tape applied the surface of each plate would give you a source of consistent material targets to measure against.
Agree, it's always good to add a piece of tape to surfaces you want to compare. At least Thomas knows not to measure reflective surfaces, I've seen that happen too...
I would’ve watched a 35 minute video!
But then we would have not gotten to seen is mad editing skills !
I hope it will have 100°C tests and some check after 10-15 minutes when the heat gets everywhere...
33 minutes?
A lot of time travelling in this video.
Do you have a time travell 3D printer - like This old Tony has his lathe? :D
Yep, looks just like a TARDIS... ;)
Thanks for investing your time to educate us Thomas!
Tom ive said it before and i'll say it again. The production quality and editing skills have improved so insanely fast. That super clean edit where you "broke the fourth wall" was awesome! Having yourself still be in frame as the camera was removed from its location must have taken a while to figure out but man it looks great. Keep up the awesome work.
Another fantastic video. Thanks Tom, I was always wondering how my Prusa's Heated Bed is really behaving.
BTW love the humour with future Tom!
I use a Lulzbot Taz6 and changed to the modular heated bed. It has an aluminum plate with the glass/PEI sheet on it. Made a huge improvement, especially on large ABS prints. I think they realized the issues with their beds, but it was an extra expense. Nice to see the FLIR imaging.
Well done Tom, thanks for looking into this.
Dude your edits and cuts are on point.
Excellent comparison love this. Wham Bam system?
Waiting for part 2. Tom blows the fan on 💥💥💥
Very interesting comparison video with a lot of good detail. The only thing that I have to offer is that with my Flir (same as the one you are using) seems to be fairly inaccurate when it comes to giving actual temperature numbers. When compared to multiple IR heat guns, mine at least, seems to be about 15°-20° off.
Excellent info ,as always. I would like to know your thoughts about what is the best frame for a cartesian 3d printer. Regards
Very cool video. Can you do the next testing setup for flatness? i would love to see options for replacing a warped bed with something better because alot of issues come from warped beds. If we could fix the warped beds then we wouldnt really need the glass.
Great video! I'd also love to see a review of the uStepper S with and without the 3D Shield. Looks like a great solution for eliminating skipped steps for 3D printing.
Tom always playing up the Prusa steel sheet magnet finger pinch for dramatic effect.
@2:35 That was magnificent Tom!
But, I have to say the lighting is much nicer in your inserted "future" clip than in the main video.
Good video thanks. I've found the best combination of materials when making a custom machine is silicon heater bonded to thick (5-6mm) tooling plate aluminium to prevent warping, then a polymer sheet like PEI. The silicone heaters can handle much more power than pcbs (even mains voltage), the thick aluminium spreads the heat evenly, they don't warp.
Best Video Yet, Love the Bath scene lol, Very entertaining Tom. Good info as usual.
Interesting. I’d guess that the air gap between the heater and the printing surface would act as a “diffuser”, making the surface’s temperature more homogeneous. Of course, less efficient, but more homogeneous.
Interesting results.
Unfortunately, it just ends up being a gap for room-temperature air to enter and cool the edges. I wonder how it would perform with a foil gasket sealing the edges.
Nice! Thanks for the work, Thomas! Now we know which bed to choose :D
Very interesting indeed Tom. I noticed the airgap when I first got the Sigma R16, although it has always performed fine with the addition of a little Magigoo. My Prusa MK2 has never had any bed adhesion issues, even though it's still on it's original PEI sheet. The Borg (augmented CR-10 S4) has a 240v 1Kw bed heater from Keenovo adheered to the original aluminium bed, with cork underneath and another aluminium plate under that for security. It seems to perform great, and I have just replace the glass with the WhamBam FBS. All my measurements were only done with an IR temp sensor, but all work well so no complaints. Looking forward to part 2.
Did you ever have the idea to try to spread the heat distribution of the Ender 3 more evenly by insulating the heatbed from the bottom with a sheet of cork with the middle part cut out? That way the hottest area would not be insulated, only the "cold" edges.
Great indepth video. Must have taken you a lot of time and effort. Tks and keep up the good work.
Interesting video! But: Did you ever tested heatbed eveness unheated and heated on 60°C? I did on the Prusa MK3 and seeing your heatmap on the Prusa makes me thinking about warping and greater high differences between cold and hot. I messured this several times and the Prusa is really difficult to make the plate as even as possible. Even with the locknut-mod where you can really precise in adjusting the heatbed. I think the problem is the production variance of the PCB production in thickness.
I'd be interested if you have any of the Tevo printers or a similar printers that have a glass bed (some models/generations with PEI over it, but the newer are just glass) with the silicone AC main heater applied directly to the bottom of the glass and then insulation under that. I've been pretty happy with my Tevos, but it would be interesting to see how they compare to the other heater builds.
Could you look at one of the SeeMeCNC Rostock beds? It’s an interesting spiral PCB with glass on top. Just something that my school has and I have gotten to play with a little. The thermistor is in the center and I suspect that there might be a hot spot closer towards the middle
Excellent video and analysis!
My CR-10S is an odd ball, I have the standard heater and Aluminium bed but I then have the magnetic base sheet from the Creality cmagnet (I destroyed the surface part in a drive by printing) but I'm now using the TH3D EZFlex plate (sticks quite well to the mag sheet) with the EZMat Build sheet as the print surface. This is working really well heated with PLA and I really haven't notice any real time difference in heating (there maybe some but I haven't noticed). I am thinking of getting a silicone heat pad so I can reach higher temps, so that could be interesting.
As someone who works there, the Ultimaker 3 does not have a heated pcb. it's a silicone heating mat stuck to a thin aluminium sheet
Hi, Thomas!
What is Flash similar to in terms of the heated bed?
Kinda interested in the E3D super hot heat bed can do. Could you get them to give you a loaner and just use the electronics from a printer to heat it to see how well the bed works?
Oh yeah. This hit the spot. So good.
The Lulzbot Mini heated bed has been upgraded, this is the older design that was tested here. The new "modular" printbed has a spring steel sheet on top of the heater, and the glass sits on the steel. So the performance should be better than what was shown here. The glass can also be removed easily when needed for PEI replacement, etc without taking out the whole heater assembly. Older printers can be upgraded with a kit from Lulzbot.
How do we get started with dual extrusion video please. Trying to make a dual direct drive with 2 hot ends. Run on ramps 1.4, marlin, Slic3r. Seems like the cost/ weight of 2 print heads is better than a purge block.
Did you calibrate the emissivity of your thermal camera. The thermal camera gives a good distribution visualization but you have to use a contact thermocouple to calibrate the emissivity of your camera to know if the value given by the camera is accurate.
That's not how silicone heating mats are typically constructed. They are usually a kapton substrate PCB with silicone on ether side.
I am so glad that at least one TH-camr gets it that looooong video about anything is a boring video.
Hopefully BCN will rework the bed after seeing this. Love the Sigma, dislike the bed. So many issues with it. That air gap is the result of a flimsy frame attempting to level the bed and just doesn't work well. Making the frame more rigid should prevent the bending that produces that airgap, or, maybe abandon the long cantilever completely?
May the First be with you....
Actually, i'm curious about your IR camera now. We have an 8k$ Fluke Ti450 Pro where i work. Maybe i should borrow it for a weekend and see what my printers at home look like.
Usually, i always left the bed to preheat a bit longer once the sensor reports the target temperature is reached. I have glass beds everywhere, and i was always suspicious that the sensor might measure with some offset, because it's not sitting on top of the "isolating" glass sheet, where i actually want that target temperature..
You can also set first layer temperature higher in the gcode and drop it off a bit on subsequent layers
7:11 OW! I can't tell you how many time's I've pinched my fingers putting that spring steel sheet back on my MK3. I've learned to only do it by the very corners where the heated bed is cut out to keep from pinching your fingers.
Hey Tom, how about heating the beds further to 115°C or so for PC and ABS printing? PLA is just "daily business" and I expect this from even the cheapest chinese printer.
That was valuable and interesting. Thanks for sharing. Cheers! : )
slightly chaotic video. a back and forth and after about 10min the topic also starts. but in the end is quickly ticked off. the fun interludes in between always like. but talk about what will happen in a relatively short video robs time. apart from that, as usual informative video!
Every cool keep you the good work tom
Is there a relation between sticking to the bed and the temperature of the bed??? Or are other things more important like roughness or material , calibration ?
I feel the same way as Future Tom when I'm editing my videos. Past me, I know you can geek out for 30 minutes about how neat the science of SLA resin curing is, but we don't have time for that! Stick with the script!
Where did you learn what material prusas pcb was made of? I thought it was some sort of ceramic material when I worked with it. Do you know what the thin black covering layer is? Is it just paint or a something different?
Glass, having a good bit of thermal impedance while having about the same heat capacity as aluminium, might serve well to even out temperature over time, and maybe even over the area if the bed underneath is conductive enough. Considering bed logic usually doesn't have a PID loop and uses simple hysteresis control, even though bed PID is available in major firmware families.
A 3D print timelapse with that FLIR might be an interesting watch.
A 3D printing person with good knowledge and fast, honest responses. He & I come up with the same answers almost always, but I'm long winded and a bit to verbose (far more that he has ever shown to be, it's why I don't make such videos!).
I'd love to know how people are communicating with you in your live streams. I've not been able to find as much (not here and most not through Twitter so I'm a bit lost). I'd love to join in so we can see our identical but differently phrased answers!
Hey Tom, Could you do a video on the Klipper firmware? I've been using on my delta for a few weeks now, and I think it is far superior to Marlin. There aren't many Klipper videos on TH-cam yet.
If you have a heated bed with an enclosure does the increase in temp cause the belts to lengthen so then creating slack?
The traces do go all the way to the edges of the bed, but since they have air on the other side and not traces they're not being heated by adjacent traces, which is why the edges tend to be cooler on the Ender 3
YAY so glad your back to regular content was getting very very bored with all the 3d printing channels only talking about MRRF
9:18 I wasn't even questioning it until you gave us this look xD
I use a RepRap Mk2 heater pcb on my A8, with a glass sheet on top. I found the original aluminium heated bed distorted way to much. The corner supports are printed form PETG, and are flush with the glass surface, so there is nothing for the nozzle to hit.
My other printer is a home-built Hypercube style, I couldn't find a heater pcb to fit (300x300mm) so I'm forced to use an aluminium heater, but I have mine upside down - with the heater traces topmost, and glass on top.
Both heat quickly, and evenly, and the glass stays very flat. I print mainly PETG and TPU, occasionally ABS, all stick to the glass well, and release easily when the bed cools.
And what's about heat isolators( foam or cork sheets for bottom side of heatbed)?
For my experience, it helps to reduce heating time and corner cooling, but i don't have a flir or something like that(
Sorry for bad English ;)
very nice vid again thanks mate
Great video last for a list!
You may have forgot to compare another type of heater bed which is composed of the silicon compound stuck under an aluminum plate and then a glass plate on top that most of DIY Chinese printer use.
@Levisgil47 ouch ... I was thinking to do that as got aluminium with 6 adjustments and getting glass to replace scratch board on top and thinking to put silicone heated bed under as my printer has a conductive sensor thing
I feel weird with all that time travelling . Nice job Tom
How about the Ender 3 using glass bed?
Why not use a bit of painters tape on the surface where you want to get the actual temp? I think it would not change anything sufficiently but it would make all surface temps equal.
not all the bed are usable area like ender 3 it's 235x235 but usable area are 220 even some are 215 cause of end stops and other stuff, so the corners and edges doesn't matter for some printers
What about the Ender with a glass bed?
I kept wanting to borrow the thermal camera from work for this.
Just something to consider , I have to admit I am a little bios when it comes to Ultimaker! Sadly we dont see enough of them being used in TH-cam channels! But it is worth mentioning,The borosilicate glass on the UM's Do retain heat longer and better. I print a lot with ABS lol yes i know it's 2019 but ABS is still my go to filament of choice, along with 3Dlac as bed adhesive. and i found it to warp less on the UM2 them on my I3's , and all my printers are in a heated enclosure.
Borosilicate also has an exceptionally low thermal expansion coefficient. So it is going to hold it's dimensions and flatness under heat much better than other materials.
Try the e3d heated bed
Short version: Physics works, and the results are what we would expect. Slightly longer, good to see empirical verification.
Glass bed: Cheap flat removable surface. Poor conductor of heat. Without help leads to uneven heat distribution, and adds minutes to reach temperature.
Aluminum bed: Superbe conductor of heat. Fastest to reach temperature. Most even heat distribution. Cheap version is not as flat. Flat version is not cheap.
If you do not care about time to temperature, use cheap glass beds for a flat surface - but use a sheet of aluminum underneath to even out the heat.
If you care about time to temperature, do not use glass. Cheap / not-flat aluminum beds need software hacks as an imperfect work-around. You can get very flat aluminum (pretty much the ideal), but it is not cheap.
A glass bed with an air gap to the heater is about the worst.
I'm not supprised the one with air gap was heating badly. Air is a huge thermal insulator. That's why power components have thermal compounds even if they seem flat on flat metal.
You should've made shots from different angles to make sure that surface reflections are not affecting the thermal imaging.
When comparing thermal conductivity, what about the material thickness? If you had thin sheet of aluminium, it would also suck at spreading the heat.
True, but for perspective, even a 0.1mm film of aluminum is still heaps better than a 5mm sheet of glass for spreading heat.
I have to wonder what a thin sheet of copper would do in spreading the heat evenily?
Guess you had to go back in time to find a Lulzbot mini... Now with the Mini 2 out, they use a different bed config that is much better. Why wouldn't you test that printer as that is what has been out for almost a year now?
for your ulitimaker, Im guessing their is no clips on the back of the bed.... if you put two on the back, it might heat up more evenly. :)
It's not that everyone has short attention span. I appreciate your effort of doing shorter and more concise videos. In my case, I'm 48 and I don't have the time for watching 2 or 3 hour videos; life itself will interrupt my attention one way or another, so I do not commit to long videos I know I can't possibly finish.
Hey Past Tom...Future Tom...Present Tom...?? There's an excellent medium for conveying comparable data, which you began to utilize with the Flir imaging, and that is with visual aids. Your buddy Stefan in the kitchen could whip-up some nice Excel graphs to complement the exhaustive research and compress the data into delicious bite-sized chunks. Thanks for all you do and please be careful when time travelling; you know, paradoxes and stuff.
As usual, the Ender 3 remains king.
Don't see many reviews for the wanhao d9 MK2
I wonder if heat pipes could be helpful.
The e3d motion system please
you proberbly shuld have included the tevo tornado because it has a 230v ac heatbed
you mean t3vo tornado. else it doesn't count.
@@SianaGearz yes the t3vo tornado
That transition at 3:00 :o
Looking at fire hazard. Would you trust the ender 3 (pla 200C 60C)?
I trust it at 240C/100C day in and day out. Learn how to mitigate risks.
@@AERuffy Yeah but im not shure. If i have flashed the Firmware and enabled thermal run away protection and changed the psu to mainboard connector i would trust it too.
Fire hazards are when you have a 12v bed running on 24v using an automotive relay which you had to stretch the spring on so it would activate at a lower current.
Never would I have guessed just how fast that bed would reach 200c. Suprisingly the printer was perfectly fine afterwards.
@@williamforbes6919 lol isnt the ender 3 24v?
@@chris_0725 This was a 3DSystems Cube 3 which I had gutted of everything except the cast aluminum frame and motion platform.
Actually held better tollerances than any machine I've ever seen despite the jurryrigged 24v ramps board and hotend. 3003 aluminum bed managed to keep a flatness of under 2 microns. I know this because of the time I tried a 0.035mm first layer and somehow it worked without any height related issues.
Would someone please help me with an issue I have as Reddit's 3d group never even replied and FB neither. I have Gino pads and my glass is stuck to them and I can't get the Boro glass to release and trust me I even went to prying. How can I get the pads to release the glass?
How to make a nasty, nasty heated bed review ;)
Thanks for accommodating my short attention span
Great comparison. Good to see how the various technology performs. I would expect the 'generic' aluminum 200 x 200 beds found on RepRap machines would perform similar to the Ender. I use glass (mirror) on top of mine. I have a superstitious belief that the mirror coating helps distribute the heat. I will now have to get a FLIR camera, a bottle of wine, and some candles so I can give my mirrored glass a nice sponge bath.
All that said, how important is it? I think you should do a quick review to see how the heat distributes once there are parts on the platen as they will absorb heat and possible distribute it back into the build surface. I'm going to guess that heat is more uniform under the print simply due to convection. Or something. I may have just made that word up or used it in the wrong context.
Prusa is shaking his head 😂
Glass is not transparent to ir light. You are seeing the surface temp of the glass not the heater underneath it.
Tom is the goat
It always feels like i'm one of the few people that really values heatup times. The difference between a 2 minute heatup and a 7 minute heatup time is me getting board and walking off before the first layer goes down. I would not want to wait for many of these to get to 110c.
Hi you missed the MK42 :D But a good overview.
MK42 is quite exotic tbh - and MK52 probably isn't far off from it, either 😉
My apologies on being pedantic, Tom, but borosilicate-glass has a conductivity of 1.2 [W.m-1.K-1] (about twice as much as 'regular' glass).
7:31 Such an ambiguous term, W/mK is Watt per milliKelvin or Watt per meter Kelvin... Enjoyed the video tho!
no super heat bed from E3D??? T_T
Try adding an aluminum plate between the glass and pub
For the second part, why not just print something with part cooling instead of simulating?
Because the cooling fans are different on each printer. I don't want to test the printers specifically, but want to see what approach is most valid.