Machsupport Forum
Mach Discussion => General Mach Discussion => Topic started by: Dom on June 17, 2009, 11:00:37 AM
-
Just wondering what the realistic limitations are of pulse speed when using a pc parallel port. Will it run fine up to 100khz with a modern computer or will I have to limit Mach3 to something much lower?
-
Just wondering what the realistic limitations are of pulse speed when using a pc parallel port. Will it run fine up to 100khz with a modern computer or will I have to limit Mach3 to something much lower?
*Some* PCs will work at 100kHz. Most seem happier at 45kHz. If you want high step rates, you should consider a SmoothStepper, which will step up to 4MHz, and with MUCH more stable, jitter-free timing than the PP.
Regards,
Ray L.
-
I'm trying to troubleshoot a problem which initially I thought was intereference but now having earthed everything and put in loads of 0.1uf caps everywhere that matters and still haveing problems I'm starting to think it is either to do with the computer pulse frequency or needing thicker cable running to the machine.
What I do know is previously I was running at 100khz (on parallel port) and had problems with juddery axis movement which usually manifested itself when i was moving two or three axes at once. After trying lots of other things I dropped down to 75khz and things improved a lot. However I'm still getting intermittent glitching (though much less than before) so perhaps dropping down to 45khz is the answer. Problem with that of course is it will reduce my max speed by half. However I might be able to get away with it by using 'electronic gearing' in the rutex firmware so 1 pulse moves the axis twice as much. That way i should be able to still get 5 metres a minute just by dividing my current mach3 steps per unit by two to give me a figure of nearer 400 rather than 800 and then I should still be able to get 5m/min (or close) at 45khz.
Either that or maybe this smooth stepper thing you speak of is another solution ... but is this not really expensive?
Regards,
Dom
-
SmoothStepper is about $150. Go to www.warp9td.com
Regards,
Ray L.
-
I have never had a computer that was 100% stable at over 45KHz.
I use SmoothSteppers on all my machines, the Lathe definitely needed it to get the rapids as I have industrial type AC Servos and drives with 2000 line encoders(8000 count). I did use electronic gearing of 1:2 to get 3m/min rapids but wanted more and with higher electronic gearing the cogging was noticable. When the SmoothStepper came out I used one and it is now capable of 20m/min with no electronic gearing although I just run at 10m/min as that is plenty fast enough.
Hood
-
Note that just because a computer can pass the driver test at a given speed does not mean it can operate Mach at that speed.
If you right click in the bar on the bottom of your windows screen, you get a window where you can choose task manager . . which has a performance tab . . which tells you the CPU load. When it maxes out, you have problems.
-
I'm a relative newbie and I'm still learning. I've been reading this (and similar) threads and I want to make sure I understand something. Is speed of axis travel the only reason to increase the Pulse Frequency/Kernel Speed setting?
I followed the Tutorial Video and set mine at 25,000 and things seem to work fine. Is there any downside to having this setting too low (other than limiting max speed)? I'd like to understand the implications (pro/con) of increasing/decreasing this setting?
That also brings up the topic of motor tuning. I have my accelleration ramping up very gradually to make my moves as smooth as possible. Same question here. Is there any downside to having your acceleration set too slow?
-
I don't know of any disadvantage to using the lowest pulse rate that gets you the speed you want, but setting it arbitrarily high is counterproductive as you gain nothing, but narrow the gap between 'it works' and 'it doesn't work'.
On the other hand, setting the acceleration arbitrarily low can cause a host of problems, depending on what you are cutting. This forum has a very large percentage of hobby users who cut soft materials at very low rates. If you are in that category, then you are probably OK with very low accelerations.
For short moves, a slow acceleration Will not allow the motors to reach full speed and the net effect would be an overall slower travel, but also a very inconsistent travel as longer moves would be faster and one would imagine this makes the CV operation far less effective.
The big problem is going to come in if you are cutting 300 series stainless or most tool steels and/or you are using carbide inserts. You want to maintain a certain chip load per tooth especially with inserts as if you stop cutting and ride up over the material, it gets extremely hard extremely fast and now you have a hard-as-hell 'crust' for the next tooth to cut through . . . which it may or may not be able to do.
It does not seem logical to persons new to cutting metal, but tiny chip loads (which you will get during slow accell and decell) can cause an incredible amount of heat build up in the tool and the part. The re son is that the if the chips are the right size and are cut at the right speed, the lion's share of the heat will go overboard in the chip itself, instead of sticking around to make our tool bit turn pretty colors :(
Personally, I would run the acceleration a little below where the motors cannot tolerate. The faster the acceleration, the more consistent your feed rate will be on average. Mo' better to tool life and finish.
-
Thanks for the reply simpson36! You confirmed my guess about the pulse rate.
As for the acceleration, I need to ask again. I am familiar with the concept of "work hardening" from my machinist days (years ago). I should have given more information or asked the question differently. I will mainly be cutting wood (pool cue inlays) and the occasional piece of plastic or aluminum (fixtures/jigs). I'm not a production shop, so cycle time is not my main concern. I'm more concerned with the hightest level of accuracy and longest life (least wear-and-tear) I can obtain with the machine. My question about acceleration was more "what's best for the machine". Could too slow an acceleration perhaps actually be bad for the motors?
-
Thanks for the reply simpson36! You confirmed my guess about the pulse rate.
As for the acceleration, I need to ask again. I am familiar with the concept of "work hardening" from my machinist days (years ago). I should have given more information or asked the question differently. I will mainly be cutting wood (pool cue inlays) and the occasional piece of plastic or aluminum (fixtures/jigs). I'm not a production shop, so cycle time is not my main concern. I'm more concerned with the hightest level of accuracy and longest life (least wear-and-tear) I can obtain with the machine. My question about acceleration was more "what's best for the machine". Could too slow an acceleration perhaps actually be bad for the motors?
I'm not an electron guy, so that question is outside my knowledge, but I can think if no reason slow accelleration would hurt the motors. My understanding is that Steppers draw full current all the time so it would have no practical effect on those motors, and a servo motor would simply draw less current to accellerate the load more slowly, so theoretically it would run a little cooler. I don't think this is significant though.
While wood, plastic and most aluminum will not work harden on you, wood will definately burn and plastic will melt if you loose your feed rate. Softer grades of aluminum will often cling to and clog the tool bit if it gets too hot, so I would still go with the fastest accelleration the motors can reliably tollerate in order to achive the most consistent feed rate.
-
Another reason that fast accell is probably better is because if you had slow accell you may be bothered with corner rounding whilst in Constant Velocity mode. If that was the case then you would be using Exact Stop which would put more strain on the hardware.
Its probably not really worth bothering about but I would certainly prefer to compromise my rapids to get faster accell rather than the other way around, even if its just from a machining point of view, and you will probably find the faster accell/slower rapids would make the cycle time faster than the other way around.
Hood
-
Thanks guys!
Great feedback, but, guess what? It generates more questions.
Back to my intention for this machine. I'm looking for the greatest accuracy I can get for my pool cue inlays. The more precise the fit, the less likely it will be that you will see the "glue lines" around the inlay. In the research I've done so far, I read that Constant Velocity mode may create "rounded corners". This would be unacceptable, so I was thinking I would use Exact Stop mode. However, I agree with simpson36 that this would potentially cause "burning" due to the zero chip load during the exact stop. In addition, Hood makes the statement "using Exact Stop which would put more strain on the hardware". So now I'm torn on what mode to use.
So here's my questions. Can you elaborate on the "strain on the hardware" statement when using Exact Stop mode. I can understand the strain on the hardware during high speed start/stop operation. Since cycle time is not a priority for me, would this statement still be true if I tuned my motors for a relatively slow rapid rate? Conversely, can I eliminate the "corner rounding" aspect of Constant Velocity mode by setting my acceleration as high as possible?
-
If the machining speed was fast as is often the case in materials such as wood then exact stop can often cause very jerky movements, especially if the code is made up of lots of small segments. The machine itself will play a part in how things are affected, if its a very rigid machine then its unlikely to be an issue, if its a light fast machine then rigidity will be less and problems may be seen.
As I said above its probably nothing really to worry about, you can always tune the motors differently as you see how your machine performs. Steppers will never have ultra fast accelleration but the better it is then the better it will perform in CV.
I suppose in a less rigid machine even having a fast accelleration could cause problems so best not to think too much about it and go get the machine finished ;D
Hood
-
I can point you to a feature in Mach that you will then need to study up on and experiment with.
The CV feature does have some controls that you have access to for tailoring the behavior for specific needs/jobs. I am no expert in these settings and while I have been known to occasionally cut the cheese, I don't cut wood so I'll have no specific advice for you.
You can define within CV, as I recall, at what angle the CV will disengage. With this control, theoretically, you can have CV active to avoid the shaky movements that Hood described while still having the ability to, within reason, keep your corners sharp.
I cut a lot of aluminum with small (sub 3/8") end mills at 7,000 RPM so I have similar concerns to people cutting wood at say 20,000 or more, and thus far I have only encountered on precision shaped pocket that required expressly invoking exact stop mode.
CV does not round off corners randomly. Is is proportional to the feedrate. Sort of like driving your car; the faster you are travelling, the sooner you need to start turning and the arc you can sustain is going to be bigger the faster you are going. So IF you are not able to obtain a completely satisfactory result with the CV controls, you can still 'tune' a particular cut by slowing down the feed rate with a proportional reduction in spindle speed to prevent burning at problem spots in the process.
The above is not an ideal solution, but then again there is no logic in dissecting too closely a problem that may not even exist, so as Hood suggested, there is a time to talk and a time to just get in there and see what she'll do for you. Methinks you are at the latter.
-
I was not aware that you could tailor CV mode. I will have to search on that subject to see what I can find out. My guess at this point is that I WILL use CV mode, but keep my feedrates low to minimize it's affects.
You're both correct. At some point I'll have to "just do it". I just wanted to take advantage of the knowledge and experience in this forum and not "reinvent the wheel".
Thanks again to both of you!!!!!!!
-
Best thing is to try and make your machine as rigid as possible, that way you will have less issues.
Hood
-
Hi Bill,
You may want to send Bill a message. Maybe he'll share with you some experience on how he achieves his great inlay work. Here is a link to his posting on cue sticks.
http://www.machsupport.com/forum/index.php/topic,3348.msg25656.html#msg25656
After you get your machine done, you may want to try and find a reliable setting where velocity
equals acceleration setting and experiment some with your machine.
RICH
-
Thanks for the link Rich! I'll definitely be getting in touch with Bill!