So I picked up a few 6mm bits and I’m wondering: am I supposed to slow down the feed rate for larger bits? Or speed it up? Should I cut deeper? Right now I’m looking at 40 hours to cut my bookshelf parts with 1/8" bit, and I’d really like to cut that time down.
Hmmmmm. Let me answer the question first. I look at it like this, larger bit is removing more material given the same perimeters. Also a larger bit has less flex and should allow you to go a bit faster (or deeper). So add those two together and I think over all you get a slower cut but it is removing more material, so if it is a large pocketing job it is probably faster with a larger bit on this machine, but if it is a cutout job probably a bit slower because you don’t benefit from a wider cut. Just my best guess really.
For big cuts like you are talking about I would do a bunch of test cuts first, with the same bit and material and maybe some verion of your cut scaled down to only take a few minutes. I use a rectangle with a hole in it, the hole will either be drilled or a pocket depending on the job size.
From there you can see if a full depth cut at a little slower feedrate is feasible. If it is you can always leave a little bit and do a faster finishing pass to tighten up the dimensions and make a nice cut surface. Or just dial in maximum feedrates and depths with a given rpm. Some cuts on a table saw , the fancy cuts with the cnc, ect.
I do all my jobs very conservative speed wise but If I had bigger things I would spend some serious time dialing it in.
Larger bits can take off more material than smaller bits each time the flute comes around. Typically from what I’ve seen, manufacturers have tables to indicate how fast to set the feed rate for a given bit. This is normally for a depth equal to the diameter of the bit (1D). Then if you want to go twice the depth of the diameter of the bit (2D) you half the speed. For three times the depth (3D), you half it again.
The issue is the amount of force it takes to move the bit through the material and keep it nice and straight.
Upcut, downcut, straight flutes and the number of flutes all play a part in feed rate.
Traditional and climbing directions also play a part in the results of your cut. Climbing is only really used in really solid and sturdy machines.
I’d say don’t go too slow on the feed rate. You’re bit won’t be happy and will produce heat and dust which are not good. You wan’t chips not dust. Go deeper first, then slow down only as much as you need. Once you are happy with the results listen to the sound of the bit and try a fresh bit and see if it still sounds happy.
This process has worked well for me as I’ve just done a bunch of testing to sort out feed rates and depth. I too used small squares 2"x4" for testing.
Good luck.
From what I tried so far, it seems like you can go faster with a bigger bit. In my opinion, it is because a bigger bit will do a more efficient job at quickly removing the debris out of its path, while a smaller one will clog up more quickly.
The only thing you really need to take care with a 1/4 mm bit in wood would be missing steps, I don’t think the MPCNC has enough power to actually break a bit of this size.
So it will mainly depend of your spindle motor. From my (little) experience, the CNC won’t miss steps as long as the spindle can keep up. If the spindle slows down too much, due to different wood hardnesses in different spots, then it is likely to miss steps and all your work will be screwed up.
So, first thing, do some test in a similar wood at various speeds and see at what speeds and deepness the CNC starts to suffer. Take this value and remove 10-20% for safety, then you should be good to go.
Also, check the current of your drivers to adjust them well before doing these tests. You don’t want to run at low power, or don’t want to run too hot either otherwise you may have troubles in the middle of a job.
The advice of vicious to actually do your easy cuts on the table saw is something you might consider too. Sometimes the good old techniques are far more efficient 
@P3DCNC : very interesting post.
Do you have some pictures of what the chips should look like if the feedrate is correct? would help me a lot to know when I’m close to the right setup
Just keep in mind the rigidity of the machine, the dewalt has enough power as do the steppers, our week link is overall rigidity, mostly determined by the users Z axis length. Flex and chatter are our enemy. Small short machine can rip with a 1/4" bit very easily, but the size most people are building this thing 1/8" is recommended. As P3DCNC says proper feedrates, depth and RPM can make any bit a monster. Just because it works doesn’t mean it is working to it’s potential.
You notice I tend to stay away from actual numbers…because 1" difference in z axis length can have a drastic effect on rigidity.
I’ll take some shots tonight and show you the difference. But they really are chips and don’t look like dust. Very noticeable difference.
Thanks for the advice all. The bit is a cheap 6mm HSS 4 flute bit (https://www.amazon.com/uxcell-Straight-Shank-Flute-Mill/dp/B00NWFU0ZU/ref=cm_cd_al_qh_dp_i). No info on chipload that I could find 
I appreciate the advice on doing cuts manually, but it runs counter to my purpose. I’m trying to explore the limits of CNC joinery. I have coded up a parametric bookcase that uses double-blind dogbone joints on all sides with no visible holes (and alternative versions that use dovetails or rounded joints, but those require 3D milling or specialized bits).
This is more practical than it sounds – I am literally incapable of making accurate measurements manually. As far as I can tell, it’s something akin to a learning disorder – if I see more than 3 or 4 line markers in a row, I lose the ability to count them. Back in school I aced the theoretical portions for chemistry, but failed every lab because I can’t measure worth crap. Coming from a long line of woodworkers, this is a particular source of shame (half-joking).
I’m also investigating custom flat pack furniture as a side hustle, since my wife could stay home and run the mill while I’m at work. She has no interest in doing anything with traditional tools, but she could clamp a few pieces and push a couple buttons and (most importantly) listen to make sure nothing goes wrong for a bit of extra family income.
@vicious1
Is there any chance of an “XL” version for either the MPCNC or the Lowrider? I’m thinking something with maybe 1.5" or 2" conduit… I get that it would be problematic for high speeds, but you can offset that by cutting deeper (which is preferable for tool life, from what I’ve been reading).
4 flute is a bad choice, we use a very high rpm spindle, we can move fast enough for a 4 flute you will end up burning your material.
No other plans for the LowRider until I see someone other than me use it. Why bigger tubes? Don’t go trying to fix a problem that isn’t there…or starting people thinking there is. As far as I know I am the only one to try it and it worked great as is.
@vicious1
I’m using a speed controlled Chinese spindle that goes up to about 10000 RPM – the DW660 was just too loud (anyone want to buy a spare DW660?).
As far as creating a problem that isn’t there… I can’t speak to the Lowrider since I don’t have one, but I get noticable flex in both X and Y on my 25.4mm stainless MPCNC (4’ x 7’). These have the same 0.065" walls you recommend for the Lowrider. For sure it’s a different beast, and eliminating the 7’ segment would go a long way in terms of rigidity. Have you done much in the way of 3D work with it yet? You’ve mentioned before that you’re primarily concerned with through-cuts, and I have no doubt it’s great for that, but I would like to know how well it handles precision 2.5D and 3D cuts.
I’m not knocking your design, I’m just stating that I think my purpose might be different that what you’ve designed for. I’m looking for a $600-700 alternative to a ShopBot. I know there are plenty of conventional designs out there already, but I’m impressed by the ingenuity and simplicity of your designs. If your Lowrider truly fits that niche, then I will gladly build one. Can you give some more info on how well the Lowrider performs in this regard?
10k is still fast you will have a better time with single or dual flute.
I can’t say much about the LowRider yet. I rarely get to touch it. Luckily business has been great so far this year and is keeping me really busy. Once I get the rest of the instructions and ramps case up I will go try a 3d cut. Judging by how much better it cut in 2D it will do just as good in 3d. It is more rigid that it looks.
Glad business is booming! I may end up building a Lowrider anyway, but that would mean cannibalizing my MPCNC, which I’m not quite inclined to do yet. Please let me know how it goes, I’m certainly happy to reuse my existing hardware if at all possible.