I had it up around 2" height for the other testing, for what you would roughly need to cut 1" mattering through. That shot was just to show quickly where I was applying force and measuring against. But yes very valid point. In fact for interest sake, I’d like to take measurements and different heights to see the effect it has.
One question. What is expected to be the predominant forces holding the gantry from sliding side to side on the table. The vertical tubes being against the edge of the table or unistrut in my case? The rubber wheels themselves?
I just gotta say it’s a ~$700 full sheet capable CNC router. I think you are expecting way too much from it and are trying to do things beyond its capabilities. To have about 95% of the accuracy of a (size capable) $116,000.00 Haas CNC router for less then $1000 is truest remarkable.
I suggest you turn your feed rate down (way down to start), get good cuts and then slowly up the feed rate to see how far you can push it. Since this is a home built machine, everyone is different and everyone has different tolerances. One thing we haven’t looked at, is your table build. How sturdy the table itself is directly effects the accuracy of the router. Another thing to consider (that I just thought of) is that it may not be the belt flexing but is more likely some flex/slop in the junction between the Z axis SS tubes and the 3D printed parts. I wouldn’t be surprised if the plastic isn’t flexing instead of the fiber reinforced belt.
Before you start pointing at other things lets deal with your 2.5mm slop that is absolutely not even remotely normal. This thread is already pretty detailed and a lot of people are spending a lot of time on this. This is the only issue that needs to be solved at this point, with that much mystery slop nothing else matters.
There is no reason to defend the machine to me. It’s a brilliant design that allows for it to work when people build imperfect tables as well as good ones. Ive never intended anything negative, it’s amazing.
I came from building a Maslow. Also an impressive machine for what it is capable of with the price. Online community is appearing to be quite different. Many of us are very active and design improvements and suggest ideas and tests, and then many of us would all go do them. Because we wanted to and enjoyed the process.
I’m getting the impression here that discussions like that here are taken as a slight against the machine. I have every intention to make changes and to increase its abilities. I can make good cuts with it, but I want to make good cuts faster. For a machine that is making cuts as big as full sheet I want to increase the rigidity to be able to cut faster. The process of doing so is the fun of the machine for me.
If anyone is getting the impression I need support or am expecting it, please let that go. Participate if it’s fun for you. If this community prefers not to operate that way I am happy to keep things to myself.
I want to get down and really isolate components to find out the areas of most flex that I can improve without going too crazy.
I am not asking for anything, just wanting to share for those likeminded. My involvement at the Maslow community forums first may have given me a different than is present here. I apologize if that has ruffled any feathers.
It’s a single piece torsion box made of MDF. Parts cut on a cabinet saw. It’s extremely rigid and heavy with adjustable unistrut rails the wheels ride on. Deflection across the entire table just over 8’ long and 4’ wide is less than .5mm
There are many potential sources of flex. I’d call it flex, not slop. After removing loads so far it springs back to almost exactly the same place. My definition of slop/play means it does not. Ie: backlash is play/slop.
If you are interested for sure. I need to finish reassembling the machine and I can test things properly. I’m happy to do any tests you want to your specifications. However, if this is work for you please don’t invest yours or anyone elses time. The machine is great. My intention is as I stated above. Isolate sources of flex, probably beyond loads most people here seem to be interested in. If you find it useful, happy to share.
I am here because I want to be, promise. There are a lot of other threads I/we try to help through the day so tackling one issue at a time really makes it easy for us to scroll back a few posts and keep up. If we are all discussing several things at once people will lose interest if they have to read through several pages to catch up when they get a chance to check in. Just a suggestion to keep as many eyes here as possible. We are all excited to improve and push the boundaries, but lets get you at least to were we are all at first.
In my opinion you have a major issue somewhere and I am most interested to get that handled. With Heffe’s quick and dirty test with “sloppy” cable ties, 6mm belts, pre-tiny spacers, and belt path optimizations, his numbers are pretty good. You should have tighter tolerance than that, but at least be on par. Get it put back together and let’s see where it stands now.
Speaking only for myself:
I like and appreciate the tests, the speculation and the problem solving, but what I really worry about is people not knowing the difference and thinking they need XYZ to make it work and end up missing the fun, while they chase the number on the indicator.
I hope you can understand that. I’m defensive of the people who are trying to absorb all this information and not seeing the whole picture.
Be really careful with anything like “more rigid” or “improved”. Conversations like this need to be qualified because 2kg is not the normal load on any wood milling, let alone precise work. Someone might not know that, or they might not know about the balance of a machine and think that they too need to do XYZ or they won’t get good results.
That’s what you’re seeing in the responses. It’s not against you testing or changing things.
You’re obviously enjoying this, and I want to enjoy it too. I went into my garage on my own free will. I haven’t left any of my V1 machines alone. They all have differences from the originals.
I’m sorry if I made you feel like this isn’t welcome. Don’t be surprised if I disagree with your results either. It’s more about my curiosity and not a judgement.
Gotcha, I was misinterpreting where you were coming from. Sorry about that (that’s the problem with non-verbal communication, no tone of voice or body language involved).
With the torsion box, we can probably rule out the table as being a major source of the flex you are seeing.
One thing you could also do to test, is put your dial indicator on the top of the XZ Main and measure its movement when you push/pull on the X carriage, then do the same at the bottom of the Z Tubes and see how much lateral movement you are seeing over the Tubes themselves. It may not be the belt, but something in the XZ assembly that is causing the slack to appear.
This is gonna be a stupid question, but do you have the spacers in the XZ Main between it and the X-Z Tubes? I’m just wondering if we are seeing s symptom and not the cause of the problem in the belt flex. I could be barking up the wrong tree but it’s a thought.
It’s supposedly a day off for me, it it’s been rather full. I’m close to getting the machine back together but I’m rewiring it now to clean it up otherwise I won’t. Not a fan of wiring sadly.
Thanks for clarifying where you are coming from, helps me understand how it works here. Much appreciated. I should probably add to my topics of this nature. ''highly experimental, there be dragons here" I understand where that was coming from now, makes sense!
I have to make some wiring extensions to be able to wire cleanly but I can start playing / testing the machine soon again.
When looking up manufacturers recommended feeds and speeds for even 1/8 bit for softwood or MDF, the loads reach up to around 11kg. I would not agree with the statement about loads of 1kg being worst case. I have no doubt that statement is true for the cuts as you are making them, but I would disagree from my understanding that it’s generally true in wood milling.
I’ll check the rest of the replies again but don’t think the rest probably needs to wait until I have my Lowrider back together… (So close! Darn wiring. If I let I had a friend who loves wiring… I hear there are some crazies who really enjoy it!)
You are missing the largest factor, chipload, and most every chart is 1xdiameter DOC. And you are using a dual flute for your calcs as well all that will do is increase surface speed needed, nothing else. Best to compare the same type of tool, the recommended single flute since out router is such high speed.
The 3x on this means the overall flute diameter, the DOC is 1x or even .35 for some. All tools should be listed this way so you can do your own math for the DOC you want. This 1/8" is listed as a chipload of 0.0049, $22
Opps. Doing this onmy phone. First half of my post disappeared. I’ll have to retype. I’ll edit.
I was waiting on some details on cutting cabinets panels out at a friend’s cabinet shop. He said they cut 3/4 material in one pass with a 1/4" bit 2 flute at up to 200ipm.
My calculator shows that at 57kg. Yes a totally different machine. What I am addressing was the statement that testing beyond 1kg force is pointless because in wood milling the worst case scenario forces don’t pass 1kg. That implies that it’s a limit of what a bit in wood can do. Now if the statement said it was the maximum people are doing on the Lowrider that would make sense to me. So I am wanting to clarify that. If load was pointless to go beyond that then I would be totally wasting my time. Dont want to do that!
Oh, and Ryan. Your numbers are very reasonable for this machine.
