New build in Clapham, North Yorkshire, UK


Started printing my bits for the MPCNC ‘Burley’. I thought I’d use this thread as a blog to document the process so that other people can learn from my inevitable mistakes.

I brought an Ender 3 Pro just as the Covid-19 lockdown started. I’m fortunate that I can fully work from home so apart from the lockdown, it’s great working from home rather than commuting to London to work.

The Ender 3 Pro is a great little printer. I can’t really comment on the quality as my knowledge of 3D printers is limited to just this one. I can say that trying to get a BLTouch V3.1 to work with it was beyond my capabilities so after wasting a week, learning how to reflash the little board it comes with though having only a 128KB of memory seems rather small, I reverted back to stock and pulled the BLTouch off. After saying that I used to develop UNIX on a 64KB multi-user AT&T 3B2 some many years ago :slight_smile:

My thoughts so far is that the Ender 3 Pro is nice, but the printing estimates in the bill of materials are about half the time I’m getting. As an example. I’m printing TopCornerF2 x 2 at the moment and Cura states that it’s around 8 hours 11 mins, which is twice as long as the BoM. My best estimate is approx 180 hours of non-stop printing which is unlikely as I’ll need to get up in the middle of the night to change things. What’s the easiest way to speed things up? New printer?

I’m very familiar with software development and OKish with electronics and utterly useless with tools, handwork or cleaning up after myself. I’ve ordered a 32bit replacement motherboard for the Ender 3 Pro, an SKR with TMC2008’s but may decide to try it on the MPCNC first. I can happily get a toolchain set up for MPCNC firmware and recompiling the code looks pretty simple.

Outstanding work as of 13/4/20:

  1. print remaining bits - best estimate next Sunday / Monday.
  2. Source a sheet of 18mm plywood to use as base. Tricky without any hardware shops being open. Anybody local to Clapham/Settle North Yorkshire with 18mm ply going spare?
  3. Source a spindle. The DeWalts are expensive here :frowning:
  4. Build a wiring harness.
  5. Buy wires for a wiring harness.
  6. Buy heat-shrinkable rubber tubing for wiring harness.
  7. Buy crimping tool for all the plugs I suspect I’ll need to make.
  8. Work out right firmware to flash SKR V1.3 MB to.
  9. Get PlatformIO working with Emacs. Whilst the Arduino IDE is OK for “Hello World” , it’s not a serious dev environment (INHO).
  10. Sort out PSU requirements.
  11. Calculate size of tubes needed with OD of 25mm.

Think that’s about it. If anybody lives nearby, say “Hi”.




Hi Rob, and welcome!

What’s the easiest way to speed things up?

You can change the print speed in Cura. Lower speeds give better finish but you can compromise by only increasing the speed for the infill and inner walls. You can also try printing slightly thicker layers. If you have a 0.4mm nozzle, try 0.3mm layers. Just change one thing at a time, by about 30% and see if it makes a difference.

Buy crimping tool for all the plugs I suspect I’ll need to make.

It’s really not that many. The way I did it was to buy 3 ready-made stepper cables from ebay (they’re not expensive), cut them and solder them together with the short wires that came with the motors to make the series harness. I’ve seen several people have problems that were traced to poor connections in unnecessary connectors, so I tried to avoid them. If you’re happy with soldering, I think it’s a sensible way to go.

I recommend adding some nylon braid sleeve to your shopping list. A 10m roll of 12mm wide braid is plenty, even if you double it up in places for the tape measure trick. Also some 12mm heat shrink to tidy up the ends.

The £110 Makita trim router is popular, or there’s a Katsu copy at about $40 that some people are using.


Iv’e had the pleasure of using both the Makita and the Katsu routers. While it is easy to differentiate based on quality the Katsu will work and the same mounts will work for either router. While working with the Katsu I did have one of the brushes ping across the room once so you may need to check that they are tight. Good luck on your adventure and trying to find plywood during the lockdown.

1 Like


I only have the nozzles that came with the Ender 3 Pro. I lost a week trying to get the BLTouch V3.1 to work. I gave up in the end as there seemed to be issue after issue, weird bugs. I ended up digging into the firmware code for five different versions of Marlin and just gave up. I haven’t changed the nozzle as I wasn’t sure if it would make a big difference or not. Would I need to change filament and all the stupid questions I would need to ask as I’m very new at this.

I’ve tried to follow the infill values provided by V1, but I’m taking baby steps as I learn. Though I do now have a slowly growing collection of bits. Just started the TopM_Corner x 2 and Gantry Spacer x 2. Cura estimates this at 11 or so hours, so I’ll kick this off overnight. Once these are done I’m 42% of the way through (by time). I have another roll of Red and Black filament along with Fuscia :frowning:

Good advice on the cabling. I’m happy to solder, I’m also teaching my daughters how to solder as well, not wholly sure they’ll ever need it, but it’s a life skill. Yesterday we did socket wrenches and how to put your chain back on your bike.

The nylon braid sleev is what I really meant with heat shrinked tubing.

The Makita is probably the one I’ll end up going for. My first tool though will be a Sharpie and a pad of A3 paper. I suspect that will all be I need for a month or so.


Pinging router brushes is not my cup of Yorkshire tea, ba’gum. My brother in law is a proper carpenter in York so may see if I can tap him up with a trade purchase of a Makita. I learnt my lesson along time ago that when you pay for quality, you only regret it once when you buy it, when you pay cheap stuff, you always regret the purchase.

18mm indoor flat plywood is going to be an issue during the lockdown. Not sure that whipping down to a DIY store in Leeds (around two hours) is actually classed as an essential journey. Kendal is 40 mins away, Lancaster is 40 mins away, nothing closer unless you count sheep and heather. The police are really cracking down on people driving long distances around here.

Many thanks


1 Like

Most printers come with a 0.4mm nozzle, and 0.2mm is the ‘normal’ layer thickness for most people but it’s fairly common to either use 0.1 layers for fine work (like D&D miniatures) or 0.3mm for large things with no fine detail (like the MPCNC). The same filament is fine. If you have a larger nozzle (say a 0.6mm) you could try that, with a layer height of 0.4mm, but I was trying to keep things simple.

The easiest option would be to post your speed settings so we can see if there’s anything unusual. IIRC, V1’s estimated print time is based on 30mm/s print speed, which is quite conservative, so there’s no reason yours should take twice as long.

Not sure that whipping down to a DIY store in Leeds (around two hours) is actually classed as an essential journey.

Even if you got there, I think they’re only selling stuff they consider to be essential, and you have to order it in advance (and when I went to check, the web site put me in a virtual queue). The builder’s merchants and timber stores are also only selling to essential projects. You might be reduced to looking for scraps in skips…

Actually, if you’re desperate, I noticed you can buy plywood on ebay, especially if you don’t need a full sheet.


Thanks for the information. I have just had a quick look on Amazon and can see that the range of prices go from 30p / nozzle up to around £38 / nozzle. I don’t think the £38 fits my printer anyway (thankfully).

Is there any major difference between them? Seems like a massive price range to me. I wonder if nozzles are the same as the loudspeaker cable marketing with directional copper flow and nonsense like that. I have seen fire optic cable advertised as directional and better quality. I have built a number of data centres in my work and I suspect some of my larger ones move Petabytes per day, never yet used a directional fibre optic cable.

The print speeds are based on Cura (AFAIK). Here’s a screen grab of the Cura settings

The Cura estimates and my Ender 3 timings pretty much line up.

Before I started this, I hadn’t realised that there were so many possible options for printing out an object. Now I think about it, I can see why but it’s all learning.

I’ll look around and see if there are any wood merchants doing business, we have a lot of farmers here so they have a constant need for wood for repairs. I might see what they can suggest. I have just had a look on eBay and clearly I can get it cut and delivered from them. Might work as I suspect they will cut it a lot better than I ever will :slight_smile:

Just finished TopM Corners and Gantry Spacers, actual 11 hours, 6 hours based on V1 timings.

Now to 4 x Spacer_Corner_F_Burley. Cura estimates 2 hours 35 mins at 65% infill. V1 Estimate is two hours.

Many thanks for the help.


The alloys and mechanical precision on the nozzle varies, people are saying - usually reflected in the price. And it makes sense to me. The nozzle does heat and cool down alot, and is subjectet to much wear, pushing all that filament out. I guess the tiny variances can influence the print quality a lot. But what do I know??? Perhaps they are all the same? They say UK-based E3D is the preferred brand, but it costs more than the chinese brands. I do recommend trying out different nozzle sizes none the less, as it’s a relatively cheap upgrade that opens up many possibilities. (but aren’t there too many of this “small upgrades”, in the end? Suddenly, without realizing, we have thousands of small upgrades…)

1 Like


Thanks for the update. I have no idea but am happy to buy a cheaper 0.4mm nozzle and fit it. Not sure I’ll pay for an expensive one though.

I’m pretty happy with my Ender 3 Pro. It’s big enough for me but I think it has three major drawbacks:

  1. The motherboard only has 128KB of memory. This is not enough now.
  2. It’s not as fast as I would like.
  3. The support from the factory is utterly, utterly useless. The instructions for installing a BLTouch are not good enough, software is delivered poorly (if at all), it’s out of date, it’s wrong in some cases. I can easily see my Ender 3 Pro becoming frame and motors only. I’m already using Octoprint, I’ll change the motherboard for a 32bit version. To be fair the rest of the stuff seems OK to me.

I’ll add a sample from Amazon and see what it prints like.


I would be surprised if your current nozzle isn’t 0.4mm, but it’s possible. It should be written on the side. It’s also a setting in Cura.

I think the expensive nozzles are made from stainless steel, or have ruby tips, and are for printing with hotter, more abrasive filaments. You don’t need them for PLA.

Looking at those printer speeds, several of them are quite slow. Putting the inner wall up to 50 will make a difference, but you could increase everything a bit. I’m using 30 for the outer wall and 60 for the inner, and that’s fairly conservative.


I think you are right that the nozzle is 0.4mm. Currently on a print so not 100% certain. The setting in Cura is now called “Line Width” and is also set to 0.4mm. Not sure I need some super exotic Unobtainium material as I only have PLA.

I’ll wait til this print is finished and bump up some values (one at a time and see what happens). Mind you I could model all of this in Cura as that seems to be accurate enough for me.


At some point, there was a strength test and I remember 0.25mm for a 0.4mm nozzle being the sweet spot. Any of them will work though. I don’t remember where I saw that though. I know Ryan was printing at 0.25mm until he got larger nozzles.

Just finishing the largest piece I’ve printed yet, the XY_F_Burley. Printed on 0.2 line height, 0.4mm line width, 55% infill.

I have absolutely no idea if this is appropriate or not. I suppose the advantage is that if it breaks, I simply print another one.

So now ordered plywood from eBay, lots of wiring to make the wiring harness, the right number of stepper motor drivers, I can’t seem to count to five correctly.

Still to buy, PSU, spindle (probably last thing), braided hose, heat shrinkable tubing.

My spreadsheet shows another 100+ hours of printing, which is a real drag.


1 Like

I’m pretty sure you could use 0.26 line height, this will be a bit quicker. Stick with the recommended infill percentage, the XY_F_Burley parts should use the same line height however.


Just finishing the XY_F_Burley and Z_Motor Burley. Current estimate is 20.5 hours, it’s sloooow. I’ve done everything on 0.2mm Layer height so have stayed with that.

The next big things are the Roller’s. Cura has them as 8 hours 57 mins at 0.2mm Line Height and 7 hours and 5 mins at 0.26mm Line Height. All four are timed at 36 hours and 19 min with 0.2mm Line Height.

I’ll start that run and it should finish by the time I finish a bottle of wine around 9:30pm tomorrow night.

Still a long time to go on printing.

Most things now ordered, need to source a decent power supply. I’m ignoring the spindle as that’s the least of my issues.


Did you try increasing the speeds? They’re all slow, but increasing the inner wall height is the safest place to start.

Yes I did and moved them to the speed you suggested. I use Ultimaker Cura and it’s reasonably accurate. It may say 11 hours 20 mins and it’s 10 hours 57 for example.


Work on printing the MPCNC now stopped as I managed to set the motherboard on fire.

I was upgrading the firmware on the motherboard to try and get the BLTouch to work and managed to drop a metal ruler onto a live motherboard. Sparks literally flew and the board had a couple of mini fires on it.

Thankfully nobody hurt but the board is wrecked. New one ordered and should be here today.

Lesson 1 to be learnt, “Don’t short out connections on a live running motherboard with 24v”.

Lesson 2. Investigate longer cables to the Ender 3 so the enclosure can be mounted away from it. The main reason the motherboard was so close was the Z axis end stop cable is as small as it can be, meaning that to get access to the JTAG means you have to put the motherboard very tight to the chassis.

Lesson 3. Don’t use metal rulers. Nice for cutting with.

Lesson 4. Don’t be stupid.

I failed on all counts, particularly Lesson 4. In what little defence I can muster, I haven’t blown anything up in 25 years, the last time was a Sun workstation PSU.

Oh well, forces me to focus on real work.



Printing now started again. Purchased Creality V1.1.5 “The Silent One” Motherboard.

Slightly worried as when I rewired it in, nothing happened, no noise, absolutely zilch, null points, apart from a fan on starting and stopping.

Pulled the PSU off, started checking connections, fuse OK, 240V in, 24v out to the terminals on the MB. However nothing on the LCD. Assumed I’d fried everything. Rats.

Spent an hour tracing and testing (carefully, see Rule 4 above) and couldn’t find an obvious problem. No LCD, dead.

Then decided as a last throw of the dice to plug the Pi in with Octoprint and see what happened, lo and behold, made a connection, homed X and Y, moved things around, heated the bed and extruder. Master, the monster lives !

So it appears that I fried the motherboard AND the LCD. The move to V1.1.5 motherboard was one I was thinking about to reduce noise, I’ve got a new LCD on order and that should be here Saturday. However Octoprint does do everything I need. I have no idea what firmware I have on the new motherboard, but it definitely works with Octoprint. Back to printing F-Roller.STL.

If anybody would like the old 1.1.4 motherboard, let me know and you can have it for spares or research.


I’m not sure which lcd you have, but it may just not be configured in the firmware, or just plugged in wrong.


The new motherboard is the Creality 1.1.5 MB. The LCD is the one that came with the Ender 3. The board is advertised as a plug and play replacement for the older one which has incredibly noisy stepper driver motors. I’d be surprised if the MB isn’t configured for the LCD that comes with the Ender 3.

The LCD plugins are orientation coded, i.e. a small protruding tab, you can’t put it in wrong. However I think you can blow it up by shorting it out :slight_smile: Another one is due in a few hours, so I’ll wait til the next print job finishes and see what happens when plugged in. Since it’s a new MB, I’d expect the LCD to work. I’ll find out in a few hours :slight_smile:

Still more than 90 hours of printing to go …