@BillKorn Good question, Bill. I wonder if this is overly complicated for our machines. If we can rely on the router to spin free enough to make it work, we could eliminate the bearings and their pockets from the design.
Yeh, shorten the frame and attach a 1/4" shaft directly to the top - threaded or epoxied - and cut away. I may have to cut another frame and try it. I donāt see any reason it wouldnāt work in acrylic or HDPE if the speeds were low enough.
Now itās time for a new diversion. I started thinking about a 4th axis. I know itās been done before, but I really enjoy designing, so here goes. Of course Iāll have to try making the pulleys, not buying them - Iām nothing if not cheap. Besides, F360 has great parametric modelling tools, and being able to design the pulleys around the system would be useful.
I think that Iām going to try the no-bearing drag knife, too. I have some corian that I can use. Itās 1/2" not 3/8", but I donāt see that as an issue. Once cut out, I can drill into the center of the top of the frame and either tap it for a 1/4" bolt or glue one into it. Cut the head off the bolt and go for it.
@BillKorn Iām way out of your league.
A 1-1/2" long 1/4" bolt would provide a lot of bonding surface in the hole, and thereās plenty of meat to drill in 1//2" stock. I think with Corian/steel you may want to go with epoxy.
Agreed. I think that Iāll go belt and suspenders. Tap and epoxy.
Fellows: I was trying to understand if the cutter turns freely at the spindle bearings or at that triangular housing, half way up?
I canāt figure out what that housing is for?
Sorry for my ignorance, but Iām really enthused about this thing. My girls would be thrilled if I would help them with their craft ideas.
1 Like
Hanis: The cutter turns freely in the bearings. The housings do not rotate. You can look at minute 8:48 here
to see it in action.
What Bill was saying is that the Makita router spindle spins so freely that the entire knife assembly is turning in the router. That being the case, Bill and I are going to try one without the bearings and the bearing cup assemblies. It will just be the blade holder and a shaft glued into it. If it works, it will be of a much simpler and lighter design.
If it does not work, it will be all Billās fault. 
Aha, I see it now. Thanks for the clarification.
And yes, straight shaft up to the collet, would highly simplify thingsā¦ Iām with you on that.
Keep it up 
Hansi
Grant, I simplified and shortened the frame, added a threaded hole in the top, and edited the thickness for your 1/2" Corian. Fusion did a miserable job creating the dxf but I attached it anyway. If V-Carve imports STLs, IGES, OBJ, or STEP files I can send you one that is much better.
When you are establishing the cut depths for each section note that the blade has to be dead center side to side under the pivot point or it will not track correctly. The frame positions the cutting edge of the blade behind the pivot so it follows the pivot. The original frame is 0.375" thick so I added 0.0625" on both top and bottom.
Let me know. Have fun.
Drag Knife Simple Frame 0.5 Thick.dxf (1.0 MB)
@BillKorn Thanks much for this, Bill. I did one this morning in corian, but I did not take into account adding to both the top and bottom. I just used 1/2" corian, but left all the cutting depths āstockā. So, Iām betting that mine with not work. Itās too bad. It really looks good. 
I can import both stl and obj files, so if you wouldnāt mind, post them or send them in a PM if you are concerned about copyright.
Wowza, how did I miss this thread taking off. Still reading through it, but this would be a great discovery if successful. Great work Bill! (Grant, Iām sure you did something toā¦ 
)
-Jeff
So just to clarify, and for the benefit of the slow folks like me that come after and read this thread, it sounds like you believe you may have a simplified drag knife design that doesnāt require bearings and can be milled fairly quickly on the Longmill? Is there much drilling or tapping required with this simplified design or can we basically cut it and go?
I finally got a few small pieces of 1/8" aluminum plate to try milling metal and an 8"x8" 1/2" (I think thatās what I ordered) block should be on the way in the next week (amazing what Amazon.ca carries). It sounds like perhaps I could cut one in that? You guys amaze meā¦
-Jeff
@jwoody18 Jeff: I just cut a new one this afternoon in 1/2" corian. I did one yesterday, but somehow broke it while threading for the 1/4" bolt/shaft. Stupid! If this works, it will simply be a scaffold for the knife. There will be no bearings. We are trying this because Bill found that, in testing his bearing model, the knife rotated on the router rather than on the bearings. So, I suggested that we just get rid of the bearings. Keep in mind that I have absolutely no idea what I am talking about, but I have corian scraps so I have nothing lose but time.
In reply to your direct question, you need to tap two holes 10-24 for the screws that hold the blade clamp. If you donāt have a tap, you can simply drill the body of the knife to the same size as the clamp and use nuts and bolts. For the shaft that connects the knife to the router, I am using a 1/4" bolt with the head cut off. I chose to drill and tap the body of the knife with a 1/4 x 20 tap and then coat the threads of a 1/4 x 20 bolt with epoxy and screw it in. If you donāt have a 1/4 x 20 tap, you could just drill a 1/4" hole and glue the bolt in. Then, just cut off the head of the bolt so that the shaft fits into the 1/4" collet of your router.
Bill has been my mentor on this. The plans that we bought were for 3/8" material. Since I am using 1/2" material, the cutting depths need to change. Bill told me that. It would have been better if he had told me BEFORE I cut the first one, but you get the advice you pay for. 
I hasten to add that I have not made one that works yet. Bill has. I hope, tomorrow to test the one that I cut today. There are a couple of critical dimensions. The first is that the cutting edge of the knife must be offset from centre. This is built into the plans. The second, is the blade must be precisely on centre front to back. Thatās where I made my mistake, not compensating for the thicker material. (In hindsight, I may have been smarter if I had sanded/planed the corian to 3/8". )
All that said, I hope to know tomorrow is this bearing-less idea works.
Jeff, the original was not a hard project, and machining all the aluminum was a good exercise. But as Grant said, it turns out that the chuck in the router spins a lot easier than the bearings in the drag knife, so after a couple back-and-forths Grant tried the simplified design. It has a couple things in its favor - it requires less machining, fewer parts, no bearings, and puts the router and the blade closer together, which reduces the moment arm and should make the whole thing a little stiffer and therefore a little sturdier. The router is designed to handle substantial side loads so no issue there.
The frame, now the only piece besides the blade clamp, can be make from any reasonably tough material - acrylic, aluminum, maybe HDPE, and as Grant can attest, Corian. I donāt think MDF or a soft wood would work long term, but a hard wood might.
If Grant agrees, I think we may have changed the design significantly enough to make the drawings and pics available.
1 Like
Bill, brilliant iterative engineering here - and a great discovery. Making things simpler is always more difficult than making them complex.
So am I correct in understanding that the CAM software uses the pivot point at the tip of the knife, combined with downward pressure and a sweeping arc on the toolpath in order to cause the blade to rotate to the new desired position? So to take an outside 90 degree cut it might stop with the point on the corner and then sweep through arc until the blade would be expected to be at 90 to the original line?
Iām quite looking forward to trying this out, if you decide youāre able to share, when my 8x8 6061 block arrives, hopefully late this week.
I guess I should order a pack of knife blades as well? Any suggestions there?
-Jeff
Jeff, youāre almost exactly right. Some software, including the add-in for Fusion, stops the cut at the corner, raises the blade, does the pivot, and lowers the blade. Hereās the video for the add-in - https://www.pbmilltec.de/Produkte/DragKnife/ Skip to about 10:00 to see the toolpath.
Itās kind of a waste of time with thin stock, but radically improves the corner appearance in thicker stuff like foam. One thing that would be easy to miss is the need to select where the knife starts, and what direction itās going to go, and then to align the knife in that direction. And start all of the succeeding cuts in the same direction so the knife ends one cut facing the direction itās going to start the next one. Fun to watch.
I donāt think everybody needs a drag knife, but it does open up the ability to do a whole range of materials you wouldnāt have thought of doing on a CNC mill. And itās very quiet!
@jwoody18 An update, Jeff, the bearing-less version did not work for me. I tried a couple of things last evening and the common problem was that the knife did not rotate as required. I guess that itās safe to say that there is a reason that every design of a drag knife that I have seen included bearings.
That said, Bill may have better results.
The experiment continues.
Jeff, Grant and I almost have a simpler modified bearingless version to try. Preliminary tests on previous versions werenāt too promising, but fingers crossed this one will work better.