Hello,
I have cv and backlash comp enabled. My units/min show a steady decline until I come to a dead stop over 15-20 minutes. The slow down includes jogging as well. The shuttle accel is set very high (don't know why, it was that way when I first set up) but it cuts smoothly. I set the accel down to the recommended speeds and it stops moving smoothly and picks up a jerk I can't tweak out. If someone could tell me how to keep from bogging down I would greatly appreciate it. 

What you seem to be saying is that, when you are running a G-code program, the program sets off and seems to run at the speeds set for G0, and G1. Then, as the program progresses, over about 20 minutes, the respective speeds gradually slow down, until the whole machine grinds to a stop. This happens with CV on and with backlash compensation enabled.

Does it only happen with backlash enabled - i.e. if you cancel backlash compensation then the thing runs at the proper speed throuhout the program, or without CV on it runs OK

I was recently using my lathe with a program of about 20 mins. It is set to CV and backlash (quite a big backlash) was on and I did not notice any drop in speed (although my G1 speed is only 4 ins per minute on steel).

What % is your backlash speed. What speeds are you expecting in the first place. You are mentioning acceleration, which I assume is the setting in the Config/Motor Turning/Acceleration. What is this et at for the motor speed you are expecting.

It sounds an odd fault, but there must be a reason - somewhere.

You are correct and it only happens with backlash enabled and the machine grinds to a halt. CV doesn't seem to have an impact. I've tried 5%-200% backlash speed. Though I have tried tweaking the motor accel I was refering to the shuttle accel.  I have a max speed of 100ipm but speed seems not to matter. I've been using a pocket wizard at different speeds trying to figure out the problem.
Thanks for any help, Shepard

curious and more curious - what pattern are you trying to cut/mill/turn or whatever.

Jim

Just contouring a simple clover leaf shaped bracket.  Once I discovered the problem I switched to a 3" pocket so I would have more feedback as I tried tweaking.

I think we could do with some help from the team on this one.

I have just done a short program - g0 x1,g0 x.3, g0x.5
Run without backlash - perfect
Run with backlash (set at 20%) - runs to x1, runs back to 0.9215, long delay, runs to 0.3, runs back to 0.3795, long delay, runs to 0.5
and this is only on one axis. The delay was several seconds and seemed to be standard, regardless of the distance setting for backlash.
I had wondered if your problem was on small incremental movements, if your backlash exceeded the move, would the machine be able to calculate.
I must admit this has me beat. The backlash - which I have set up on my lathe works perfectly, and is accurate. It will, for instance stop if cutting a ball shape, pause on the crown, apply the backlash, then cut down from the crown, with perhaps the faintest of marks where it has paused. The backlash on the mill needs some expanation, I think. In the scenario above, why in one direction does it run 0.0215 and in the other 0.0205 - before applying backlash. If this is part of the backlash the two should be equal. I have tried with backlash greater than and less than this figure.  If your backlash is less, then the delay is short, but if the backlash exceeds this figure then backlash can be some seconds delay.

I can see if you are contouring a complcated profile, where all axis have to stop and apply backlash compensation, that the machine must seem to grind to a halt.

All my observations are on Mach3 program alone - the machine here in the office is not connected to a machine.

Anybody got any observations on this, please

Jim

Try setting your shuttle accel. to .05-.1 and your backlash speed to 200% and let me know if it helps.

Brett

Changing the shuttle acceleration cuts down the time the axis travels after (reversing) before stopping, before applying backlash, and changing the backlash speed helps cut down the time the machine takes to apply the backlash. I would NOT however, on my own machine, like to take the backlash speed above 100% in case I missed steps - (but that is on MY machine)

On a shuttle setting of 0.1 and 100% backlash speed, although the machine hesitates, it is certainly much quicker off the blocks.

Whether that is going to answer Sheperd's difficulties - I await to find out.

Thanks Jim, those numbers are just the ones I have seen good results with on my machine and others seem to agree, they work pretty good. All machines will be different though. My machine is slow and relatively heavy for the steppers that run it. This means that I have a slow accel. rate. It also means too that I have a much higher safety factor for backlash comp. than I would if it was really light. The reason being, backlash comp moves shouldn't be moving anything except the drive mechanism, not the table, Gantry, etc. This is the reason you shouldn't see skipped steps at 200%. The backlash is only taking the dead space moves, this generally will not take any where near the torque required to move the components of the machine plus tool load.

Brett

When I set the accel to those speeds the speed of the reversing axis caused a jerk I can't get rid of. I lowered the amount of backlash the axis needed to compensate and solved the problem. But the amount of backlash now compensated is less than the measured backlash. Im still trying to find the sweet spot but i'm getting close.