635??   Well thats going to be 2540 pulse per turn of the ballscrew or 12700 pulses required per inch.  Quadrature encoders output 4 times their rated lines per rev value.   635 cant be the total count as 635/4 is not an integer.

So at Machs max pulse rate of 45Khz for the printer port you will get (45000/12700)*60 = 212"/min. at 25Khz 118"/min.

I think you may have to eventually go for an external motion device to get back your 400"/min, like the G100 or NC-Pod or simiilar.  Or lose some resolution and use a drive with a pulse multiplier.

Wayne....

Here's the part # off the encoder 9006-2257-635.  Maybe 2257 = pulses/rev?  Darek has worked with these, he said 635 was the lines/pulses/count don't remember which.  He also mentioned that it was a bast**d/non standard.

Mostly I do small parts, really don't need 400 ipm, 200 will most likely be good enough.

635 Although on 1st look seems pretty weird.   but if yo look at the numbers you end up with, it makes sense on an Imperial ballscrew machine.

12700 (12.7mm = 1/2") 25400 (25.4mm = 1")

Wayne...

635 Although on 1st look seems pretty weird.   but if yo look at the numbers you end up with, it makes sense on an Imperial ballscrew machine.

12700 (12.7mm = 1/2") 25400 (25.4mm = 1")

Wayne...

So, now in motor tuning I put 25400 in for steps/inch?  That gives me a max velocity of 106

Your steps/inch will be 12700

635 x 4 = 2540 (Quadrature output of Encoder)

2540 x 5 = 12700 (5 tpi)

That will give you close to the figure I quoted above at 45Khz.

Wayne....

Your steps/inch will be 12700

635 x 4 = 2540 (Quadrature output of Encoder)

2540 x 5 = 12700 (5 tpi)


That will give you close to the figure I quoted above at 45Khz.

Wayne....

As you can see, I am easily confused.  Plugged that in, now I'm unconfused...............temporarily.

I was just running my eagle from an NC pod and you are going to need more KHz then that can give I could only get to just over 130 IPM and had to reduce the PPU to 10,000 to get 427 IPM I was running at 25000PPU and 500 IMP with the G100. This machine gets more hardware then most.. it is the test bed that I run all the hardware on

I have 50,000 steps per inch when I done have a step multiplier set in the drives.. that is the best I can do on the mill that I test with

Just wanted to tell you that there is a 75KHz limit on the NCPod... If you get a Galil it is only limited to the encoders and I think it is about 13MHz that it will read them at. It is a Nice fast way to go if you have the hardware that needs it. You could run your old drives with a galil too...

Better then that is the Delta Tau stuff that we will soon be programing for... I think they will take in up to 40 MHz on the encoders and output at 4MHz if you would like to run step and direction... The PC104 will also run in velocity mode and torque mode as well.

Well that is about all I can tell you about the things we are working on getting running... as you can see there is going to be GREAT speed in the future with true closed loop! IT just takes time for us to get the work done

Brian

There is a chinese motion control card coming on the market sometime over the next few months.

Some of the features are as copied from the Datasheet below  Cost is said to be several hundred dollars.


Wayne....

32-bit PCI bus, Rev. 2.2, 33MHz
􀂗 Pulse output r rate up to 5MHz
􀂗 6 Pulse/dir output modes: OUT/DIR, CW/CCW etc
􀂗 2~4 axes linear interpolation
􀂗 2 axes circular interpolation
􀂗 Multi-axis continuous interpolation
􀂗 Position/speed change on-the-fly
􀂗 13 home return modes and auto home search
􀂗 Hardware position compare and trigger with
auto-loading FIFO
􀂗 High speed position latch function
􀂗 Programmable acceleration and deceleration time
􀂗 Trapezoidal and S-curve velocity profiles
􀂗 28-bit up/down counter for incremental encoder
􀂗 Multi-axis, simultaneous start/stop
􀂗 Programmable interrupt sources
􀂗 Supports up to 5 cards in one system (20 axes)
􀂗 Hardware backlash compensator
􀂗 Software limit function
􀂗 Easy interface to any stepping and digital servo
systems
􀂗 All digital inputs and outputs are 2500VRMS
isolated
􀂗 Manual pulser input interface

Brian

Thanks for all the info.  My $ are limited.  I just ordered the breakout board.  Now I have to wait 'till I can order the pixie's.  I will get along with whatever these will accommodate and send Santa a wish list.

Wayne,
That looks like a nice card

ardor,
I know about the money thing In the past I would save a month to buy a drive so I could test to see what I liked the best...