Wait till you print a few things with it! You can crank up the acceleration and speed and still get great prints. Oh and now your circles will be circles too!
Enjoy!
Thanks!
Steven
Thanks again Steven. It really is a nice upgrade. I was so happy I went ahead and did the y axis. That was more difficult, because the bearing holders are just a hair too small for the replacements. I ended up only putting three bearings in and leaving one of them with the stock plastic bushing to prevent binding. But it is rock solid. I actually think leaving one plastic one in on the y axis is a good idea anyway.
Interesting approach. I spent an hour or so tightening the Y holders just perfectly so it would have no binding. Using 1 plastic bearing might have saved me an hour.
Let us know how it prints once you get it all dialed back in.
Possibly a stupid question but when you guys do the bearing/rod swap do you have to change anything software wise to accommodate afterwards or does the machine just print better quality all around with the upgrade?
This is a Post I made showing off some of the speed and print quality that I can get with the bearing upgrade.
https://forum.lulzbot.com/t/high-speed-taz4-5/1372/9
I just put up a high speed video of how some artifacts are made due to nozzle deflection at high accelerations–I think it’s cool at least. For that video I had my TAZ 4 acceleration set to 2000mm/s^2 and the first layer velocity at 100mm/s.
https://forum.lulzbot.com/t/strategies-for-resolving-print-artifacts/1400/2
@chasis
This is one of the earliest posts that I found when I had this issue created by 1013. It’s about 2 years old now but some of the info is still relevant–I think someone already put this link up on this thread, sorry for re-iterating.
https://forum.lulzbot.com/t/taz-very-loose-solid-rod-bearings/112/1
Here is erv’s post on the issue. The thread is full of some really good info on the shaft/bearing upgrade–I got what I needed to make it happen from here.
https://forum.lulzbot.com/t/linear-bearing-backlash-big-x-axis-issues/966/1
Yes did a print with the default print speed set to 9000 and it came out great.
Now can we go into the firmware and tweak the acceleration numbers? And how much can we bump them up?
You can do it through the LCD menu. Control => Motion => Accel. I wouldn’t go higher than 2000mm/s^2. I’m not so much worried about the nozzle (x axis) but the bed (y axis) is already heavier and only gets heavier the more filament that gets laid down. This can make for some serious nozzle deviation when you are getting higher up the a axis.
Yes you can do it via the LCD screen. Make sure to choose Save Config otherwise the changes will be lost when you turn off the printer.
I run 1500mm/s2 acceleration and am very happy. I could go faster, but 1500 gives me silky smooth prints, 2000 started to show small harmonics in the print. I think @Brew4You posted a cool slow motion view of the harmonics when accel gets super high.
Thanks!
Steven
@wantmys2000
Thanks for getting the details I missed in there and that you watched that video. I’m planning to make a v-rail setup for the x axis. This should help the harmonics out as well as fix a fair amount of the issues that arise from using a dually setup. I’m going to throw a half sized extruder stepper on the carriage and pull the fan and shroud off to drop the weight. I’m also looking at turning the extruder around on the carriage so that the heavy part–stepper motor–is closer to the x axis plane. I think this stuff should help.
Brew4You,
Have you thought about maybe using a Bowden setup and getting the extruder completely removed from the X axis?
It’s something I’ve considered, but does a Bowden setup work well with 3mm filament?
@muggs
I love the bowden conversation! 
It is an idea that I’ve pondered a while on. While I like the bowden idea and I think how the Ultimaker line of printers are setup is the best way to implement a bowden setup.
Why I won’t put it on a TAZ:
- The bed still moves and already has a greater mass than a single extruder so it can provide a lot of force when using high acceleration. Granted the mechanics of the bed allow for far less flexing of the frame but with a big build volume there is the potential that the whole moving bed + plastic on it could weigh quite a bit.
- The TPE world is pretty much un-functional in a bowden setup as far as I know. Squishy plastics gotta be hard to run through almost a meter of PTFE tubing without having oozing or non extrusion points. I already have to retract 3mm and have a 0.5mm Z lift to keep some of these nylons from stringing all over, I would think this issue would become a nightmare with a bowden. This leads to #3.
- With the extra long retractions and high Z lifts needed for some of the “fancier” filaments I would think that the speed you’d gain from higher accelerations would be washed out.
- The whole machine is designed for a direct drive system so things like the z drive system would be way slow and overkill for just a nozzle on a carriage. I’m sure there are more things.
When I would put it on a TAZ - A bowden would be great front nozzle on a flexidually. This is where the bowden would become a reality for me and the TAZ. Keep the nozzle center of gravity way closer to the x plane if the bowden is out front. Then the “fancy” filaments can always be run on the back nozzle and the bowden could be saved for ABS, HIPS, PLA and other bowden friendly plastics.
Got any other ideas? I’m but only moderately versed in physics and mechanics so I’m sure I’ve overlooked some points
Bowden does work with 3mm. My other printers have bowden setups, and I am torn. They don’t work well with flexible filaments, and the retractions have to be very aggressive on normal filaments. I have fought ooze on all of my bowden’s and I normally don’t win the battle.
The direct drive minimizes most of these issues with the trade off of lower accelerations.
I think @mhackney is putting a dual bowden on his RazzMaTAZZ KITTAZ Build that he is documenting in the KITTAZ forum. Hopefully he can add some details on how it works out!
@Brew4You I love the idea of a bowden on the dually! Thats a great way to fight the weight induced sag I see on it.
Thanks!
Steven
I like the idea of a hybrid Bowden/direct toolhead. Solid stuff through the Bowden and flexi through the direct drive. Maybe it’s as simple as replacing the aluminum plate with a new redesigned plate… Printrbot has a nice compact direct drive that looks like an extension of the stepper motor.
I wanted to tell you about some bearings I am testing to get rid of slop. They are drop in replacements for the LM10UU bearings and the inside is teflon. The outside is aluminum.
Now these won’t be for everybody due to the cost of around $13 a bearing but I am pretty confidant that this is not only a fix but will keep the qualities of the plastic ones that come on the TAZ printers. That is that they will be quiet and need no maintenance (lube). They should last the life of the printer and then some.
I am not affiliated with the company that makes or sells them in any way
So, I got 3 in today to test on my TAZ 4, I got 3 so I could put them on the X axis. Thanks to Lulzbot it’s pretty easy to change out the bearings. I put one on the shaft and tried to wiggle it and although it slides easy I could feel zero slop (perfect fit on the 10mm shaft). So I went ahead and installed them. Now, I am in the middle of my first print and so far it looks very good. I will come back and update this after I get some real work put on them but I wanted to just come in here to start.
The bearings are FJC10 made by PCB Linear (Pacific bearing). I will post a few photos.
Another -1 for the Bowden setup. Frankly I hate the design. When you think about impulse and momentum, in action a Bowden setup is like a giant slinky. Every time the stepper starts and stops you get an impulse in the filament. This results in uneven flow out of the nozzle, and poor print quality, especially in walls.
Frankly I believe the Bowden extruder design is fundamentally flawed.
@brew I have been motivated to do some analysis based on your slow-motion video’s. I have noticed that the stock hex extruder mount has a lot of flex in it. Now that the rest of my setup is rigid with the bearing upgrade, this seems to be the prime culprit for the harmonic ripples I still have and that you video shows so well.
I have attached a slow-motion of my own that shows the relationship of the extruder pinion and the rest of the X carriage. If you look close you can see there is flex during the direction change. I wonder if that gets stiffened up, or extra mount points are installed if it will help. If you look at the current design all of the extruder mass including the stepper is only attached to the carriage with the 2 bolts that the hot end uses at the bottom of the carriage. Thats a lot of mass to attach to a sliver of plastic with quite a large lever being that the stepper is a good 1.5 inches above the mount.
Thanks!
Steven
I am thinking something like this to support on the top and bottom. I might print this up and do a comparison.
Thanks!
Steven
This looks good I bet it would make a difference for sure. Is there a spot for a bolt at the top? The hexagon nozzle sticks up into the extruder body so I might be hard to get into this extruder design. Can you post an stl? I’ll print it and give it a try.
This is not printable yet, I had to base it off the non Hex Extruder parts because the TAZ 5 directory does not contain the Hex Extruder parts in a NON STL format for me to modify.
I have another thread going to see if anyone knows where they are. I will let you know once I get one drawn up with the correct parts as the base.
Thanks!
Steven