4g upgrade and speeds?

[syeven_ree]

Any advice would help!
When the 4g upgrade came out I jumped on it but since then I have tried 4 computers and just had one built specifically for the shop-bot. through all of this I cannot get the ms past 6 and js past 6 what could be the problem? I have set the voltage and done every-thing possible to increase speeds to no avail.
Any-thing past this and the motors stall.......could there be someting wrong with the board in my control box? I'm running a PRT.
Any advise would help, please.
Steve

[richards]

Steven,
What were your speeds before the 4g upgrade? I believe that the advertised increase in speed is about 2 to 2-1/2 times faster than a pre-4g machine and that the advertised increase in cut quality (resolution) is 4X better, given the fact that it takes 4X as many steps to move the same distance.

A move speed of 6-ips is pretty good. Even 6-ips jog isn't bad, although I think that by increasing the ramp a little that you'll be able to get another couple of ips.

Also, what is your power supply voltage? The stepper motor's torque is mostly affected by the amount of current available from the power supply, but the motor's speed is more affected by the available voltage. The Gecko G202 can handle voltages between 24V and 80V. I like to run my motors at least 15X the rated voltage. For instance, I would run a 2V motor at nothing less than 30V (and nothing more than 25X the rated voltage - or 50V). There is a trade-off between speed and heat - especially given the fact that the 4g is jumpered to NOT reduce current automatically when the motor is idle.

[zeykr]

Steven,

I have the 4G and have to stick with MS 6 and JS 8 to prevent stalling also. Quite happy with those speeds though ;-). I plan to try a quatech high speed serial board that I have soon to see if that offers any improvement.

In the new beta there is a diagnostics folder with a communications tester program to test comm speed from pc to control box. My results with the usb converter are 9.9834 elpsd time, 1000 packets, 34.05449 khz, 73.9%. I ran several times and got same result each time, so test seems quite repeatable. I'd be interested to know what you get with it and what those who can reliably run at MS 8 get with it.

[syeven_ree]

Mike
Thanks for the response. I was getting a max speed of 2 to 3 prior to the 4g upgrade, the voltage is set at 48. I don't need faser speeds then I get now but was wondering why some people can achieve 12ips and not me? Is it because it is a PRT not an alpha?
Steve
I will try the port speed check.

[waynelocke]

I was cutting a maximum of 1.8 to 2.5 before the upgrade. I now routinely cut at 3 to 5 since installing the 4g. I have been jogging at about 11. I have gone faster on the cutting and jogging but it is too nerve racking. I have had a couple of stalls but put in a usb2 card dedicated to the controller and have not experienced any stalls since. I have also not used it a lot with the new card or pushed it.

I know that in a recent post Ted had recommended usb but doesn't the Agent controller recommend a high speed serial port? Has anyone tried this? It seems to make sense since the usb has to be converted to serial at the control box.

Wayne

[richards]

Steven,
There are a lot of factors at play. The PRTs that used a geared motor use the PK296 size motor, which has about 300 ox*in of holding torque. The Alpha has an AS911 size motor that is more like the PK299 series. It has about 600 oz*in of holding torque.

The PRT with 4g requires 2,000 steps per revolution. The Alpha requires 1,000 steps per shaft revolution. So, the PRT will have cuts that are twice as smooth as the Alpha with all other factors being equal.

In order to take out the mystery in the math used to figure distances traveled and obtainable speeds, lets start with some definitions:

Pulse Rate - The number of pulses per second that the control computer can generate. The pulse rate seems to top out at 25,000 to 30,000 steps per second, depending somewhat on the USB to Serial converter being used.

Pitch Diameter - The pitch diameter is basically the diameter of the pinion gear. It takes into account the actual contact point between a tooth on the pinion and a tooth on the rack, so the pitch diameter is slightly smaller than the actual diameter of the pinion gear. You can find the pitch diameter listed as one of the specifications in the gear manufacturers' catalogs. A 20-tooth gear has a PD of 1-inch, a 25-tooth gear has a PD of 1.25 inch and a 30-tooth gear has a PD of 1.5-inch.

pi - The mathematical relationship between the diameter of a circle and the circumference of the circle or 3.14159.

Pulses Per Rotation or Steps Per Rotation - The number of pulses required to rotate the shaft of a stepper motor one turn.

Gear Ratio - The number of times the motor's shaft rotates to turn the gearbox's output shaft one time. Those PRTs that had gearboxes normally used a 3.6:1 gearbox on the X and Y axes, although 7.2:1 were sometimes used on some machines.

The first formula that I use finds the distance traveled per pulse: (Pitch Diameter * pi / PPR / Gear Ratio) = Distance traveled per step.

Example for a PRT with 3.6 gearbox and a 30-tooth pinion: (1.5 * 3.14159 / 2000 / 3.6) = 0.000657 inches per pulse.

Example for a PRT with no gearbox and a 20-tooth pinion: (1 * 3.14159 / 2000 / 1) = 0.001576 inches per pulse.

Example for an Alpha with no gearbox and a 20-tooth pinion: (1 * 3.14159 / 1000 / 1) = 0.003152 inches per pulse.

The other formula that I use finds the jog speed available. It uses the number found in the examples above multiplied by the Pulses Per Second available: (Distance traveled per pulse * Pulse Rate) = Inch Per Second jog rate.

Example: PRT with 3.6 gear box and 30-tooth pinion and 25,000 PPS: (0.000657 * 25000)=16.41 ips

Example: PRT with no gear box and 20-tooth pinion and 25,000 PPS: (0.001576 * 25000)=39.39 ips

Example: Alpha with no gear box, 20-tooth pinion, and 25 000 PPS: (0.003152 * 25000)=78.79 ips

Obviously, since the jog speeds show speeds much higher than those obtainable on most machines, other factors have come into play. I think that the most overlooked factor is the stepper motor itself. Although a stepper motor might have 300 oz*in of holding torque, that same motor might only have 50 oz*in of torque when running at 25000 pulses per second.

In order to get the highest performance out of an existing motor, you would want to verify that all of the current that the motor can use is available to the motor. That is usually determined by the value of the current set resistor on the Gecko stepper driver. You would also want to determine that adequate voltage is available to the stepper motor. The Gecko driver requires at least 24V but not more than 80V to operate properly. You can use a voltage that is up to 25 times the motors rated voltage (as long as that voltage is > 24 and < 80). I've found that 15 to 20 times rated voltage is a good compromise between speed and excessive heat; however, different motors seem to have their own 'sweet spot' when it comes to voltage. Probably the most important factor, when it comes to tuning the drive train, is to verify that all mechanical parts are adjusted to their optimum level.

[jeffreymcgrew]

Steve: After I did the 4g upgrade, I too had some stalling problems.

What helped a lot is to go into the Properties on 'My Computer' in Windows by right-clicking on it. Then go to the Device Manager by clicking the button. When it opens, go to View on the menu, and pick 'Show Devices by Connection'. This swaps the view to show thing based upon how they are hooked to the computer instead of by category. Now you'll need to click on lots of plus signs to expand the tree out until you get to the USB bridges and the Shopbot, but this will let you see what else is sharing whatever USB Bridge the Shopbot is on. I had much better results once the Shopbot was placed on it's own bridge and not sharing anything. My problem was that the laptop I was using had an internal webcam that also shared one of the USB bridges and was causing problems. So even though I didn't have anything but the Shopbot hooked to the USB ports on the computer, it was still having stalls because of the stupid webcam. So double check that the Shopbot's on it's own Bridge first...

[syeven_ree]

Hey Guys
Thanks for the help, I have 25pitch gear and nothing sharing my usb with the bot. What if I install a seperate usb card for the bot?
Steve

[zeykr]

FYI:
Put in a quatech qsclp-100 high speed serial card last night and tested with the shopbot speed test program.

As reported above, my results with the usb converter were 9.9834 elpsd time, 1000 packets, 34.05449 khz, 73.9%.

With the quatech serial board they are 7.984 elpsd time, 42.585 khz, and 92.4 trans efficiency.

This is on a Dell with a 2.6ghz pentium 4 and 512M ram. I did aircut a 2d file for a while at ms 8 js 10 with no stalls, but have not run long enough to be confident in those speeds.