[sploo]

Jon,

Many thanks.

That looks a bit less critical than I thought - i.e. it's working well, and the problem is not terminal.

Having seen a 3925 (though not in action) I was wondering if you could put some brush like material (similar to what people use for making dust collection hoods) along the sides of the machine.

The bristles wouldn't move - they'd be stuck to the sides of the machine, above the linear bearings.

They wouldn't be used for wiping anything, just preventing some dust falling onto the rails. Not sure how well it'd work to be honest.

Are you using any kind of dust extraction at the cutting end (i.e. a hood around the router over the cutting bit)?

PS Halfnutz, good call on the caps, thanks.

PPS Shame cbass doesn't seem to be around. It'd be nice to know the status of his 3925.

[Halfnutz]

Heres a link to THK's dust caps in thier catalog,

http://www.tos-world.com/TOSUS/HTML/pdf/exa1290.pdf

I dont know what size the 3925 uses, but probably 15mm or 20mm. Maybe the holes are the same.

[mll]

Hey K2 customers, if you need rail caps please call us and ask for Mike.

[AeroJDK]

I just found something out today as I decided to take off the ball nut on the Z-Axis. Upon removing, the balls came out of the nut and required careful placement back in before rescrewing the nut onto the screw. What I ended up doing was convering the outside of the ball nut in rare earth magnets (Neodymium). When I went to screw it back on, the balls didn't move a bit and the nut went back on like a charm. Once it was completely screwed on, the magents were removed and viola! As a side note, I also disassembled the Z-Axis enough to allow me 360° access to the ball screw for better visibility.

I would say that if you want to take off the ball nut, this might make it easier than Taus's experience, but putting the balls back in if any fall out will definitely still take time.

Jon

[AeroJDK]

On another note (this one being the reason I took my ball nut off in the first place), I finally figured out how to run machine higher than 20 in/min (F20) and have noticed that as I get up around 50-75 in/min, the machine becomes very "jittery" as it cuts. I am currently cutting a bowl out of foam and have noticed that during all of the parallel finishing work, the machine is seems like it's constantly stutering.

How fast have others gotten the KG-3925 up to? If you working with it as high as I am, are you noticing these symptoms? Any feedback would be great!

Jon

[Aksess]

I run my one around 200 ipm with great results on wood and around 20 ipm or slower on mirrors. Look at my photos they tell the tale. I was thinking of selling my machine and getting a laser. But cant quite talk myself into it

[AeroJDK]

Thanks, Aksess. After reading the previous posts in this thread regarding feeds and these machines, I guess a better question would be to ask who is cutting at speeds equal to or greater than 100 ipm with the new servo style motors? It appears that Aksess and Taus are using the steppers and they are working fine, but I am still having problems with the servos. I have been adjusting the speeds and acceleration of the motors, but that doesn't seem to change anything. Any input would be greatly appreciated...

AeroJDK

[CNCRob]

Hey Aero,
I have the servo setup with my machine, but I haven't really been about to test it at higher speeds because the z-axis on mine stalls out at aroud 70-75 ipm.

[Arch]

I have cut some letters for a large sign out at 200ipm and it worked flawlessly. I haven't tried it out any faster than that. On another note, my 3925 was cutting parts slightly smaller than drawn. Then I noticed on K2's website that they had a new profile for the metric ballscrews, now it is right on the money. I absolutley love mine and recommend them very highly to anyone interested. My machine is light years ahead of the Liberty I had. Fast, smooth and accurate. I did get the upgraded ballscrews and rails on all axis. That was probably the best money I ever spent.

[Aksess]

My machine has the servo motors, not stepper

[Arch]

I forgot to say I got the servo motors and controller from K2 with mine. They are very smooth, quiet and don't heat up like my old stepper system did.

[CNCRob]

I agree, K2 has been very good about working with me , with all of the problems that I have had with the machine. They have fixed all that I have called them about. I never even told them about my machine stalling out. I never really wanted to go very fast until I started trying my hand a photo carvings. I may call them and ask if it is covered by their 1year parts waranity since it has done it since I have owned the machine. My other two axis's work pretty much flawless and I get a ton of speed out of them.

I have cut some letters for a large sign out at 200ipm and it worked flawlessly. I haven't tried it out any faster than that. On another note, my 3925 was cutting parts slightly smaller than drawn. Then I noticed on K2's website that they had a new profile for the metric ballscrews, now it is right on the money. I absolutley love mine and recommend them very highly to anyone interested. My machine is light years ahead of the Liberty I had. Fast, smooth and accurate. I did get the upgraded ballscrews and rails on all axis. That was probably the best money I ever spent.

[carlnpa]

CNCROB my machine seems low on power also . Let us know if they have a fix for this. I haven't really tried to troubleshoot this as I am in production mode now. Maybe the power supply is a bit small for the motors? Gecko tuning issue?

[AeroJDK]

EDIT: I found a solution. After some more researching on the internet and looking for the right key words, I found that switching to Constant Velocity made all the difference. I effectively change the G61 to a G64 at the top of the G-Code and viola...amazing. Thanks again for all the help and feedback on this matter...
_______
Might power be the issue that causes the "jerking" motion? I haven't had a lot of experience with servo motors of this kind, but I could see how power could be an issue. Does anyone know if there are adjustable potentionmeters inside the K2 box to increase the power to the motors? Again, any input or suggestions for increasing the speed of the motors while removing their "jerkiness" would be greatly appreciated!

AeroJDK

[tauscnc]

Hey guys,

I am using steppers on my machine,,was using the 180 oz steppers and they did well but after a few stalls I decided to go to 640 oz NEMA 34 steppers. These puppies are huge. I am also wondering about my power supply now as it is 52 volts at 15 or 20 amps. No sure if you connect to transformers together in series 24V +12V if you get the full amps from both or full from one and half from other, etc.

Anyway, are you guys going off of what Mach2 or such is saying for IPM movement or have you timed it?

The IPM and the gittery movements might also be a computer thing. Your computer may not beable to handle when all motors are moving, etc and at that speed. I don't have mine setup yet but hope to get it running this weekend. I will post some pics then.

later guys,

tauseef

[ger21]

Anyway, are you guys going off of what Mach2 or such is saying for IPM movement or have you timed it?

The IPM and the gittery movements might also be a computer thing. Your computer may not beable to handle when all motors are moving, etc and at that speed.

I don't think that can't happen with Mach2 (or Mach3). When in 25 Khz mode, it will always be able to output 25,000 steps per second for all axis. And it won't let you set up the motors to go any faster than that. Same with 35Khz and 45Khz mode. You shouldn't have to worry about multiple axis slowing anything down.

[cbass]

Hi everyone,

I’m sorry for delaying in my response. I was waiting for my refund from K2 and I just received it the other day.

I returned my KG-3925 for several reasons…


-The machine frame was not flat. Specifically the extrusions that form the X-axis sagged visibly in the middle (at least a 1/16” over a 24” span. If I were to mount a rigid table top to this, I risked pulling the x rails out of true. I don’t see much point in advertising that a machine can hold .001 tolerances if its not in a (flat) plane. To K2’s defense, I recently visited Techno-Isel to see an LC router, a much more refined machine (and at $13,000 it should be). But guess what? It too had sag in the middle of its table! And they were using THICK aluminum extrusions throughout the whole table. It seems that the only way around sag on a machine of this design is to have a welded, stress relieved and precision ground frame. BTW, Techno will do this for an extra $4,000 (!). Ron at K2 has mentioned that they are considering this approach for the future.

-The cross brace that reinforces the frame in the middle of the machine protruded from the rest of the frame. This meant that when mounting a flat plate to use as a working surface, the plate would rock. Realigning this cross brace by loosening and re-bolting it was not an option as everything is drilled and tapped. I started sanding the extrusion down but quickly realized the other side was lower than the rest of the frame!

-When I got the machine running, I kept getting backlash in the x-axis.

http://www.cnczone.com/forums/showth...light=plotting

I trouble shooted with Kevin at K2 on this for more than a week. We both ran out of ideas. Incidentally, this issue was due to loose bearings on the x ball-screw. I only found this out after I had the machine disassembled and in the crate ready to ship back.


I was especially upset at this last issue, as had they tested the machine before it left their production, they would have easily picked up on it.

This is not meant as an attack on K2 nor do I hope to turn away potential customers. I was very excited to meet them when they were still in San Diego, and I sincerely hope they will continue to improve. They have been very good to deal with (especially Kevin) as others here will attest to. I think that many of the problems I experienced could have been avoided with some QC.

If you are on a limited budget, and are not too concerned with holding .001 tolerances over the entire length of the machine, then $5,000 is still a bargain.

I’ve come to the realization that what I’m looking for in terms of precision is better left to a steel framed machine.


Carlo

[zealouse]

I feel the need to keep things updated and fresh about K2 products. I know allot of people are going to be mislead about K2 by the post on this forum. When I've read about the problems that some where having I almost didn't buy my K2-39x25, which would have been a Hugh mistake. I'll post picture of the delivery and an in-depth look at my machine. I guess that allot has change since K2 first attempt in to the CNC market and its all good. Jason

[carlnpa]

I think the comments in this thread have been very tolerant (maybe too much so), and are very current. They have been stating factual observations of the K2 machines, primarily the 3925. I really don't think they have been at all misleading.
IMHO
Carl

[Arch]

I would have mounted a MDF board to the top of the machine and then used a fly cutter and flattened that top. You may not get it to a Fadal's specs but you could have removed 99.99% of that 1/16th sag you were talking about. My K2 - 3925 is the best bang for 5k I've had in a long time. A 6" caliper is all I have so I've done that size rectangles and circles and they have come out within the range advertised. I do have the upgraded ballscrews and linear rails which I think is very important for overall accuracy. That being said I couldn't be happier with my machine and I still think they are the best value in the under $1o,ooo. market.