Not sure if I have an issue or not but thought to get anyone's input.

I've re-worked an older General Gorilla router machine(and yes there's no support) with an XHC Usb 4 axis card and larger drivers for the NEMA 34s.
We don't know the exact pitch on the rack and pinions but when measured on calipers the pinions are 16 tooth and 0.741 in diameter. Metric = 18.83. The rack measures approx 8 teeth per inch.

I'm now set to the lowest micro-stepping (400) on the driver and when using the Mach3 calibration feature I get around 19K steps per (inch).
It seems accurate when running but those numbers look higher than normal to me. 

Does anyone see an issue running it this way?

Is there any Gorilla owners that have done a similar upgrade?

Is there anyone with a similar high count machine or is it just me : ) ?

Thanks!

Ian

That does look funny.
Is there any belt, or gear reduction other than the r&p itself ?

I have servos on 2 machines, 50,000 pulses per inch.

If it is moving accurately, fine. That's the primary concern.

Russ

No gearing or anything. Direct drive to the racks.

It does move accurate but can't do fast. Lose counts that way.
Current velocity is 160 with accel 4.

What stepper driver do you have ?

It's this one- 2DM860H

http://www.steppermotorcanada.ca/stepper_motor_driver

Using this as the BOB.

https://www.rovercnc.com/collections/cnc-electronics/products/mach3-usb-4-axis-motion-controller-4th-generation

I'll assume that you are positive of the microstepping dip switch setting (400), and that you powered it down to secure the setting.

Looks as though the microstep is set higher than 400.

Russ

One other thing I've seen with the little dip switches.
You sometimes need to click them back and forth a time or 2 and make sure they end up in the position you desire.
If they aren't slid completely to the end, they might not actually be where you want them to be.
Then power down, of course.
Russ

Hmmm, 16 teeth per rev and 8 teeth per inch on the rack = roughly 2 inches (8 teeth per inch) per rev.

Now you need to know how many pulses you have to send the motor for it to rotate 1 revolution. Your micro steps will be considered in this part. And, you need to know the actual number, not guess and change until it moves one rev. No telling how many steps it skipped moving that one rev when guessing like that.

If your steps per do == 19,000 and your motion device can support the frequency of pulses, then nothing wrong with it. There are lots of systems running higher resolution than 19,000 steps per inch.

Just some numbers for you........

19,000 steps per inch X 12 inches = 228,000 pulses per foot

228,000 X 160 feet per minute = 36,480,000 pulse per minute

36,480,000 \ 60 seconds = 608,000 pulses a second or 608 kHz

Depends on what your using to generate your pulses but some external motion devices can reliable send millions of pulses a second (MHz).

The parallel port tops out at about 50kHz or less in most setups.

I think I see whats happening.
DIP switches set at  1, 1, 1, 1 is 400
         "                  0, 0, 0, 0 is 40,000  Which would be near 20k per inch (1 rev = approx 2")
You think they are on, but they are all off.
The controller only goes to 400 khz.
You can do the math to determine the best setting for speed.
Russ

4000,    1 - 0 - 1 - 0  would give you a resolution of about .0005"
Pretty good res. for a router, and would speed you up quite a bit.

And double check the other current dip switch settings as well, they may be inverted as well.