Why does my printer suck.

Ok well I have been tweaking and upgrading my Makerfarm i3v 12" for almost 2 years..
ive become very familiar with calibrating and testing.

done lots of little things to improve the performance and quality of my printer.

But when I look at samples from other printers that dont look like really high quality machines.
it really blows the prints im getting out of the water.

Yeterday I saw a small delta machine with cheap looking parts.
print incredible detail.

and the layers were so smooth
I realize a thin layer height will get smooth edges but the incricate detail quality was soemthing ive never been able to achieve

This machine was likely running on a 32bit processor.. im running 8bit Rambo
it was likely 1.75mm filament, im 3.0mm
it was a bowden setup with some cheap extruder.. looked like a bulldog... im running direct using Bondtech extruder and e3dV6

it was running Reprap Firmware. im using Latest Marlin firmware
Im also using Simplify3d

what am i missing exactly?

I'm new to this forum and new to Makerfarm products (I have a Pegasus 10 that working beautifully) but I'm not new to 3D printing and certainly not new to CNC type machines. Two things I'd suggest for most of these units I've looked at is first and foremost is upgrade the Z axis to proper screws. Second get proper idler pulleys on the X and Y axises. (More important than most designers seem to realize IMO) Any structures made of wood (OK fake wood really) glue the pieces together if practical (Use good judgement because if you can't keep it squared up you are screwed once it's glued, hence "if practical") and go OCD on making sure everything is squared up and parallel. Attention to details is what separates the men from the boys when working with devices with fractional millimeter resolution. Frankly you are always going to be constantly tweaking any unit with a wood frame, even more true if it's not glued. I got the full kit when I bought a Pegasus 10 last month and the only piece of wood left on it is the spacer between the extruder and the X carriage. 8 bits is plenty to run a Cartesian format device, the math is pretty simple, the Delta format is an entirely different beast requiring much more complex math and then you can take advantage of the 32 bits. Now if you were going to include a touch screen controller and Octoprint all in a single controller package then 32 bits or even 64 bits starts making a lot more sense. A RaspPi 3 would be a good candidate for such a system although I'd modify it for an external WiFi antenna. The real Achilles Heel for the RAMPS is the lack of a proper RS-232 serial clock in the Arduino architecture 16 Mhz doesn't divide down to 115200 very well hence a jitter problem and the need to run at a out of spec baud rate of 250,000 (Which easily divides down from 16 Mhz and why it's more stable) The problem with 250000 is any driver or program that strictly enforces the RS-232 protocol isn't going to like it

Other than generalities without seeing the prints to analyze the actual problems it's hard to be specific. I printed a full sized one of these T-800 heads 0.2 resolution in about 15 hours and it looks as good as any of the examples made on an Ultimaker 2 or a Lulzbot Taz and better than most of the others. I even did a manual color change, it's an eSun PLA+ black base and the skull is eSun PLA white. Granted I've been a Maker about 3 decades before they came up with a term for us so your mileage may vary but you can't beat all metal construction. Expansion/contraction due to temperature is minimal and from humidity pretty much non-existent. Don't feel too bad my first printer was a QUBD Two-up .......

I ended up upgrading the Z screws to fine Acme screws and run geared nema 17s
the machine sits on a some ceramic plates to keep everything flat and level.
i upgraded he bed to a aluminum plus boro glass & PEI

Quote:

---

Originally Posted by longjohn119

I'm new to this forum and new to Makerfarm products (I have a Pegasus 10 that working beautifully) but I'm not new to 3D printing and certainly not new to CNC type machines. Two things I'd suggest for most of these units I've looked at is first and foremost is upgrade the Z axis to proper screws. Second get proper idler pulleys on the X and Y axises. (More important than most designers seem to realize IMO) Any structures made of wood (OK fake wood really) glue the pieces together if practical (Use good judgement because if you can't keep it squared up you are screwed once it's glued, hence "if practical") and go OCD on making sure everything is squared up and parallel. Attention to details is what separates the men from the boys when working with devices with fractional millimeter resolution. Frankly you are always going to be constantly tweaking any unit with a wood frame, even more true if it's not glued. I got the full kit when I bought a Pegasus 10 last month and the only piece of wood left on it is the spacer between the extruder and the X carriage. 8 bits is plenty to run a Cartesian format device, the math is pretty simple, the Delta format is an entirely different beast requiring much more complex math and then you can take advantage of the 32 bits. Now if you were going to include a touch screen controller and Octoprint all in a single controller package then 32 bits or even 64 bits starts making a lot more sense. A RaspPi 3 would be a good candidate for such a system although I'd modify it for an external WiFi antenna. The real Achilles Heel for the RAMPS is the lack of a proper RS-232 serial clock in the Arduino architecture 16 Mhz doesn't divide down to 115200 very well hence a jitter problem and the need to run at a out of spec baud rate of 250,000 (Which easily divides down from 16 Mhz and why it's more stable) The problem with 250000 is any driver or program that strictly enforces the RS-232 protocol isn't going to like it

Other than generalities without seeing the prints to analyze the actual problems it's hard to be specific. I printed a full sized one of these T-800 heads 0.2 resolution in about 15 hours and it looks as good as any of the examples made on an Ultimaker 2 or a Lulzbot Taz and better than most of the others. I even did a manual color change, it's an eSun PLA+ black base and the skull is eSun PLA white. Granted I've been a Maker about 3 decades before they came up with a term for us so your mileage may vary but you can't beat all metal construction. Expansion/contraction due to temperature is minimal and from humidity pretty much non-existent. Don't feel too bad my first printer was a QUBD Two-up .......

---

I'm following this. I've always been underwhelmed by the quality of the parts I get out of my i3v 10 inch.

longjohn119, are there pulleys you recommend?

Quote:

---

Originally Posted by mattanonymous

I'm following this. I've always been underwhelmed by the quality of the parts I get out of my i3v 10 inch.

longjohn119, are there pulleys you recommend?

---

I got a pair of these from Amazon but you can likely find them even cheaper. Some people like the smooth bore pulleys rather than with the teeth, something about backlash with the toothed idlers but I haven't noticed any problem in that regard. I just know my Y belt isn't riding in a different place and flopping back and forth all the time which couldn't have been good for accuracy. Frankly I can't see all that much savings using a pair of bearings instead of proper idler pulleys. I can understand using normal threaded rod where there is a pretty big cost difference and it can work decent **if** the rods are perfectly straight and stay that way but that actually rare. Especially with 5 mm where just using too much pressure to cut them can bend them a little and it doesn't take much of a bend to show up in your prints. Personally I think the Acme lead screw upgrade Makerfarm offers for 60 bucks is a very fair deal for what you get and that includes a better stepper driver

I like the DRV8825 Stepper Driver better because you can actually heatsink them properly. The type of IC package used is made to pull heat from the bottom of the chip to the back side of the PCB where there should be a fairly large ground plane and you put your heatsink on the back side. This type of IC does not sink heat properly from the top, it'll help a little but not a lot like the heatsink on the back side. The A4988 drivers aren't properly designed for heat sinking with most of the heat drawn through the bottom into an area that is boxed off and impossible to add a heatsink to. Putting a heatsink on top of the chip helps but not really all that much because of the way the IC is designed. The Pololu design will work but you need to derate the current by approximately 25% so instead of a max 2 amps it's more like 1.5 amps, anything more and you'll likely go into thermal runaway and burn them up. I wouldn't run out and replace all your A4988 drivers because even 1.5 amps max is fine for most designs but if one were to go bad I'd definitely consider the upgrade to a DRV8825

I'm interested in this. After having my i3v for a year and a half, I'm not at all impressed with the design. I finally switched to a solid-state relay after burning through enough of the relays Makerfarm provides, and I'm tired of constantly having to tighten screws on the frame and tighten my belts. I'm trying to take a bottom-up approach to tuning in my machine and increasing the precision. I recently ordered ACME lead screws to replace the threaded rods on the z-axis, aluminum extrusions to make an all-metal frame, and aluminum motor mounts, carriage, idler, and bed. That will be my weekend project if everything comes in by then.

Is it worth it to go from an I3v 10 to the Pegusus, and the ACME lead screws? I already have an E3dV6 Hotend? I need to switch to Simplify also.

Quote:

---

Originally Posted by usarmyaircav

Is it worth it to go from an I3v 10 to the Pegusus, and the ACME lead screws? I already have an E3dV6 Hotend? I need to switch to Simplify also.

---

I actually have both a 12" I3V and a 12" Pegasus, both with stock machine screws.

I have found them to print at the same quality except for the Z banding. I am having some banding issues with the Pegasus. But I figure those will be resolved with a little more tinkering.

After spending a lot of time on my I3V and getting it squared away, it is a solid machine.

I finally got some decent results out of my machine!

Here's what I think helped:
- I reassembled the machine paying close attention to things like the parallel aluminum extrusions being equidistant from one another at the top and bottom and making sure everything important was square and true
- I printed and installed the 2 nut spring version of these http://www.thingiverse.com/thing:795954
- I finally ordered and installed real couplers https://www.amazon.com/gp/product/B010MZ8SQU
- I installed the pulleys longjohn119 recommended https://www.amazon.com/gp/product/B0188IFK4Q
- I calibrated e-steps, set the filament diameter to an average of several measurements, and tuned the extrusion multiplier with this procedure https://craftunique.com/forums/view-thread/694/#7130

No need for expensive upgrades like acme lead screws.

Hi

Never underestimate the influence of filament and temperature on the whole process. In some cases very small changes can have significant impact on the results.

That said, I'm *very* impressed by how the 12" Pegasus prints.

=====

That said, why did my earlier printers have issues from time to time:

1) I didn't get everything level / parallel / orthogonal and keep it there. On some printers that was a lot of work and a lot of repeated work.

2) I didn't keep everything tight (in terns of belts) and loose (in terms of friction and binding).

3) I switched to the latest version of XXXXX and didn't get it straight before I tossed the latest version of YYYYY into the mix. Firmware / software bugs (or settings issues) are a pain when they start to pile up on top of each other.

4) I tried to run things fast, convinced myself that "it's fine" and moved on. I spent a lot of time chasing things that actually were speed related without dropping the speed.

5) This filament was fine with this speed, that temperature, and when it was new. Now two years later it seems to not work as well after I forgot all those settings and it's soaked up a ton of moisture and beer fumes....

Simply put, there are blind spots in the process. Everybody has them. If you look at my printer, you will see things a bit differently than I do. You can't give me a list, because it's just the stuff you see. Even if you gave me a list, the list is based on your background and tools. Your list doesn't work for me. Equally, my list isn't really of a lot of use to you.

Bob

Quote:

---

Originally Posted by adamfilip

Ok well I have been tweaking and upgrading my Makerfarm i3v 12" for almost 2 years..

[...]

what am i missing exactly?

---

I've been tweaking my Makerfarm printers for a while, and I'm getting great results. I have several coworkers, also with MakerFarm printers, who are not at all happy. I don't really think there's a fundamental problem with the printers, but there are lots of details that matter. Here are a few tips:

• Put your printer on a flat surface. I'm using Travertine tiles.
• Get real E3Dv6 hot ends. They really do make a difference.
• Spend the $150 and buy Simplify3D. (You already did.)
• Do everything you can to keep stray air off the bed. If you're using the standard MakerFarm X axis parts, the hot end fan blows right on the bed. Switch to a different extruder that vents out the top or tape foil, under the X carriage to redirect all spill air upwards. If you can feel air from the fans on the back of your hand held under the carriage, it will affect the print, causing warping. Even spill off the blade tips of the fan intakes will affect the print.
• Make sure the bottom ends of your Z screw rods are flat. Flatten them with a file, and then drop a 4mm steel BB in each coupler to give the rods a solid single-point bearing on the end of the motor shafts. This eliminates a very common problem with these printers that causes odd Z-banding.
• Use auto bed leveling and do everything you can to eliminate bowing in the glass. You can get Marlin to output a topology matrix after bed leveling, so you can see how much warping you have. Shim the corners to correct.
• Use the first layer height in Simplify3D to adjust bed adhesion.
• Use Garnier Fructis Extreme Hold #5 hairspray on glass for ABS. There are other solutions, but if you get the airflow off the bed, this works great.
• Try underfilling your parts slightly (turn down the extrusion or e-steps). I find that slightly underfilled parts, while they have a little bit of open mesh you can see under magnification, look better and fit together better. A lot of fit problems are caused by extra filament squishing out.
• Use a geared or belted extruder. A 4:1 belt drive extruder (cough, Itty Bitty, cough) has something like 8X the resolution of a direct drive one, both because of the drive reduction and the smaller diameter bolt.
• Make sure your drive bolt is sharp and clean. When I started milling my own bolts, I was shocked at the print improvement. I didn't think it would matter.
• Use 1.75mm filament. This also significantly increases the volumetric resolution of the extruder.
• Temperatures are everything. I'm running 255C with ESun 1.75mm ABS.

All of this advice is good (probably), but without seeing the results you're getting, it's hard to guess what your problems might be.

Can you upload photos showing the issues?

I dont have a good example to show. the printer prints fine. just not great.. maybe my expectations are too high

Quote:

---

Originally Posted by adamfilip

I dont have a good example to show. the printer prints fine. just not great.. maybe my expectations are too high

---

Hi

There is such a thing as an unreasonable expectation. I can show you some pictures of prints from this direction or that. They look superb. If I grab similar parts and shoot pictures from other directions with different lighting, they look a bit worse. If I'm posting a picture of "look what I did", guess which picture I'm going to post.

I see a lot of this at work. We have a couple different printers running around. This one gets compared to that one on a regular basis. The "bragging rights" parts are *never* representative of the normal output of the printer. Give me some "clean up time" with a bit of sandpaper and some acetone and they will look even better ....

Bob

Quote:

---

Originally Posted by BLKKROW

I actually have both a 12" I3V and a 12" Pegasus, both with stock machine screws.

I have found them to print at the same quality except for the Z banding. I am having some banding issues with the Pegasus. But I figure those will be resolved with a little more tinkering.

After spending a lot of time on my I3V and getting it squared away, it is a solid machine.

---

Check that the bottoms of the Z screws are flat and file them flat if they're not. The axes are always a little misaligned, and when the non-flat cut surface rotates with the motor shaft, it rises and falls, causing Z banding. If you file the screw ends flat and drop a BB in each tube or coupler, this goes away completely.

Quote:

---

Originally Posted by clough42

Check that the bottoms of the Z screws are flat and file them flat if they're not. The axes are always a little misaligned, and when the non-flat cut surface rotates with the motor shaft, it rises and falls, causing Z banding. If you file the screw ends flat and drop a BB in each tube or coupler, this goes away completely.

---

My machine and threaded rods are square. I found the issue to be slicing as I was trying to find the best retraction. I recently printed a fan setup and have no banding.

To me both machines have the same quality, that is because I have tweaked my Prusa until it printed well. I used this knowledge immediately with the Pegasus. This means they both print the same quality.

However, the setup on the Pegasus is a lot less, you don't have to build motor mounts and multiple layers of plywood to create a frame.

Quote:

---

Originally Posted by BLKKROW

.......

However, the setup on the Pegasus is a lot less, you don't have to build motor mounts and multiple layers of plywood to create a frame.

---

Hi

At least in my experience, the "plywood" based machines had a hard time holding adjustments over time spans in months or years. That seems (so far) to be less of an issue with the Pegasus versions.

No matter which printer you dig into, you will find people that have slicing issues. We all are using basically the same set of programs and seeing the same issues with them. Since most of us like *free* software (I certainly do), that's all a little bit silly. This is indeed a world where free does not encourage people to keep working on the code base. Bugs hang around for a long time. Getting an un-paid team organized and working together over a period of years is a very tough proposition. Much of the free software (Apache, Linux etc) that is successful is actually paid software that is given away. We don't have the kind of money in the 3d printer space to fund at those sort of levels.

Bob

Quote:

---

Originally Posted by uncle_bob

Hi

At least in my experience, the "plywood" based machines had a hard time holding adjustments over time spans in months or years. That seems (so far) to be less of an issue with the Pegasus versions.

---

That very well could be. The one really critical adjustment is the Z height, and if that drifts, printing will suffer considerably. Using auto bed leveling mostly solves that problem.

I say "mostly" because if your bed is warping with temperature changes, the Z height can change after the leveling process finishes, and can even create banding in the parts as the bed cycles.

This was a big problem with the i3 because the third bearing was near the center of the wooden bed. I put a dial indicator on mine and it moved down to about -0.5mm during the initial heatup, and then moved up to about +0.2mm as the heat soaked through the wooden part. Colin told me this only happens if you paint the wood. All of my printers are lacquered, so I can't confirm or deny.

The i3v has the wheels out closer to the corners, so it's not as bad.

Hi

My early efforts on ABL were driven by the wood parts moving around. It was pretty much the only way I could keep things running. My experience was that the humidity mattered a lot. Let the gizmo sit for a month and fire it up .... it's all over the place. Let it run for a day, it settles down.

Up until about 30 minutes ago I was doing fine with manual bed leveling on the Pegasus. That's when the glass broke in the middle of printing a fairly large ABS part. I guess you *can* get the part to tight down to the glass :)

Actually a *far* more likely culprit is getting a bit over aggressive on the pad in the center of the bed.

Bob

I have split a piece of glass with ABS before. I had eight parts lined up so as they warped, they all pulled up in the center of the bed, and they yanked out a chunk partway through the print. I ultimately modified the part slightly to allow a little warping in a non-critical area so I could continue printing full beds of parts.

Quote:

---

Originally Posted by uncle_bob

Hi

My early efforts on ABL were driven by the wood parts moving around. It was pretty much the only way I could keep things running. My experience was that the humidity mattered a lot. Let the gizmo sit for a month and fire it up .... it's all over the place. Let it run for a day, it settles down.

Up until about 30 minutes ago I was doing fine with manual bed leveling on the Pegasus. That's when the glass broke in the middle of printing a fairly large ABS part. I guess you *can* get the part to tight down to the glass :)

Actually a *far* more likely culprit is getting a bit over aggressive on the pad in the center of the bed.

Bob

---

Hi

This was the fourth or fifth time I've printed basically the same part. My guess is that the part did not help things at all. It was certainly not the only reason the glass broke.

I'm currently printing a much larger version of the same thing in PETG with far fewer issues .....

Now, what to do with that inventory of ABS spools :)

Bob