Start by setting your steps/per and a modest velocity,  there are no real starting figures as every setup is different, I would start with the latest autotune and see how it gets on.

http://www.cs-lab.eu/en/artykul-4-Download.html

Set the axis in the middle of the travel and press the autotune button, the axis will move + and - many times until it finds an optimum setting,  You may need to hit reset and re-run it if it faults.

As a guide my settings for the X & Y axis on my small mill are 16384, 200, 1000,500 and Z is  77000, 87, 1000, 1, but they may be of no use at all on your machine.

Graham

2.0 Firmware will be soon available. Soon also wider documentation for CSMIO/IP-A will be ready.

Below you will find short 'step-by-step' description how to tune the PID.

The maximum allowed position error can be adjusted by 'Max Following Error' parameter.

Parameters:
-------------------------------------
kP - proportional gain
kI - Integral gain
kD - Derivative gain
kVff - Velocity feedforward gain
Max I accumulator - maximum vaule of integral accumulator
Max following error - maximum allowable position error

Manual PID tuning is usually performed in the following way:


- set kI, kd, Deadband and kVff to 0

- set kP to low value (~500)

- set Max following error to large value - for example 10000 or so

- JOG the axis and increase value of kP until first vibrations occurs

- decrease kP with about 15%

- set Max I accumulator to 100000

- JOG the axis and increase kI value until vibrations occurs.

- decrease kI with about 15%

- kD in most of cases can be leaved '0'.

- Now JOG the axis and increase kVff value. Observe the position error on the chart while axis is accelerating or decelerating. Find kVff value when error is smallest.

- while axis is switched on, but not moving click icon near 'DAC offset' to automatically adjust DAC offset. This prevents some 'knocks' during machine switched on

- last thing - check the maximum position error value, than adjust 'Max following error' 4 times larger and everything should be OK.

Happy to see these are working well !
Only wish the "M" was faster than 125k. It's perfect for a current app. otherwise.
Costly jump from $275 to $720 when only wanting the higher 4M speed.

Thanks for starting this topic,
Russ

It is a bit of a shame but its not too bad if using servo drives that are capable of electronic gearing. Normally I dont like electronic gearing but I was considering the M for the wee lathe and was going to put up with having it. Only reason I went for the S version for the wee lathe was the M does not support threading at this time.
Hood

I understand you Hood.
I'll be gearing servos w/ 2500 ppr encoders if I go w/the M. Otherwise, I might try the USB-SS I have on hand.
Then again, adding the BoB, Line drivers, missing out on the 24vdc I/O and the analog .... it all adds up.
I also prefer not to use gearing but the resolution is not that critical for this app. Gearing was also not too bad on my mill w/the AB drives.
Thanks

Suppose it all depends on what gearing and screw pitch you had on the machine. With the wee lathe I had direct drive on X with 2500 line encoders and 5mm pitch screws so that would mean I could have got a max  of 3750mm/min which would likely have been plenty but even if not then a 2:1 electronics gearing wouldnt have been too bad. The X would have been worse as its 2000 line encoder, geared 2:1 but only 2.5mm pitch so max rapid would be 1172mm/min Likely fast enough to get  by on the short X but think I definitely would have used the electronic gearing on it.

What really put me off electronic gearing was the big lathe  when I was using the PP. To get the rapids I really wanted I would have had to have about 6:1 if I recall, could only get 45KHz reliably with the PP.
Hood

Saw a new plugin on the downloads page last night and gave it a go this morning. Tried the threading ( 2mm pitch thread on a 16mm dia piece of mild stee ) and the pullout was instant and it looked every bit as good as any thread I have done on the big lathe, so very happy with it.  For some reason the spindle override is not working so I was not able to vary the speed during the thread to see how it performed.

Overall I am delighted with the CSMIO so far, I have found a few small issues but I am sure they will be addressed soon.
One thing I have found is when going to change any config settings in Mach, make sure Mach is in Reset or you will get things hanging. No big deal as I am in the habit of doing that now but took a few mins to work out why it was hanging sometimes and not others.

Another thing I have found is if you have limits and home switches sharing on multiple axis  it will trigger an E-Stop when the axes home. Having Homes shared between multiple axes also has an issue where as soon as the first axis hits the switch it will not back off but instead the second axis starts homing.
If sharing a home and limits but keeping each axis on its own input then things seem fine, you have to do as the manual states however, disable the Home Switch Safety option in General config.
 As the CSMIO has plenty inputs then sharing would not normally be done anyway and in fact it would be better to have separate home switches as that way you can take advantage of the Index homing capability.

Another thing that I found was if you hit a limit switch then you can reset but cant back off if you have the MPG module. With the module connected I could not select keyboard jogging to try it as the MPG module kept changing Mach back to MPG mode. I found however that if you have a switch in line with the power to the axis select then you can select between MPG and Jog. Once in Jog mode you could then back off the limit no problem.
The manual shows a momentary switch in line with the power to the axes select switch but it doesnt really say why you need it and only says its optional. I never use keyboard jogging so instead of having a switch there I think I will have a switch that I will use as a limit override switch, if a limit is hit and I wish to back off I can press and hold that switch and it will allow me to jog off via the MPG.

Hood

This is great news with the threading, Hood. Thanks for posting. Hope those minor issues get sorted too sometime soon.

Got a CSMIO IP/A with the MPG and threading module last week. Plan to use it on the lathe. Don't know when I will actually get to do this though as I have a lot of work to do first.

What I didn't like from your description above is the lack of ability to use jog keys in conjunction with MPG as I am used to having jog keys. Or is it only when a limit switch is tripped?

Dan

Merry Christmas to all

In August I wrote CS Labs asking:

"The EMCO Compact 5PC that I'm converting has 2 simple opto sensors on the spindle - One for speed that reads a 100 pulse-per-revolution slotted disk and one for indexing that reads 1 pulse-per-revolution for indexing. Could I not simply use 2 of the CSMIO/IP-M digital inputs to read these opto sensor outputs and configure them for RPM and Index sensors through MACH3? Having to use the CSMIO/IP-S + ENC module for this appears to be overkill for this simple project."

They replied:

"I think about something for threading in low-budget projects. Currently we have a lot of work with firmware and new functionalities for CSMIO/IP-A and IP-S, but function you're talking about + PID controller for spindle speed(torque control mode) should be available also with near firmware upgrades. "

As I have not seen anything on a -M update, on 5 Dec I wrote:

"Do you have an update for the status of the CSMIO/IP-M firmware update that will add spindle control (speed and index) to the -M  controller for lathe threading?"

Has anyone heard from them regarding this?

Dan
 Here is a vid I took this afternoon of the Conect threading 25mm Dia mild steel and 2mm pitch, as you can see the pull out is instant. http://youtu.be/8c1sndp-554
Below is a pic of the thread.

Regards the selection of jogging, I had a thought last night and tried it out this afternoon. The rotary switch I am using for axis selection has 12 positions but I had the lock tab so only 2 could be selected (X and Z)I moved the tab round one position so I now have 3 positions on the switch, the last not being connected to anything thus it effectively cuts the power to the X and Y axis select inputs. When I chose that position I was then able to select the Jog mode and use the keyboard to jog.

Hood