Greenie problems configuring 6i25 with 7i77, using Granite Devices VSD-E drives

I'm very technical but this is still my first CNC conversion project and it is driving me a little batty. I've made good progress and the configuration part is kind of the last push (ugh). I'm having a lot of confusion on pncconf. Part of this is that I don't want to screw something up because, let's face it, there is a lot of iron flying around! Anyhow, this is the configuration:

1. 6i25 with 7i77, latest firmware applied
2. three Granite Devices VSD-E servo amplifiers
3. 7i77 ENA pins enable the drive (5VDC provided)
4. Board Name on pncconf: 5i25 Internal Data
5. Firmware on pncconf: 7i77x2 With One 7i77

What I know works:

1. If I setup the drive for analog input, I can (dis)enable the drive and jog it without any problems
2. I need about a 0.0153V offset voltage to get something relatively stable (0.0154 shifts drift the other direction)

Questions:

1. It seems like with PWM, I'd not have to worry about any offset at all. Interestingly, when I look at the SS#1 tab, I only have a PWM option, but the output is still +/-10V. So, is there a way to interface to the servo through a PWM signal as the name implies? Or is that terminology simply an indication of how the DAC is managed internally?
2. In the 'I/O Connector 3', 'Num 1' I see '7i77 Mode 0 I/O Card'. However, I want to run in Mode 3, but the UI control is disabled so I cannot even see what other options are there.
3. Per a previous conversation, I have routed my servo amplifier FAULT line to 00, 01, and 02. This was due to the fact that my amps are outputting 5VDC for HIGH and my VField is 24VDC. Apparently the 00, 01, and 02 input ports can read analog, but I don't see how to configure that in pncconf. How is that accomplished?
4. I've performed the Open Loop test but that is about as far as I've gotten. This one really isn't a question versus a 'where am I now' statement... Sometime in a distant galaxy I guess I'll come back into pncconf and enter the PID values... I assume that I'll do the same with the following errors as I determine how the machine responds. Right?
5. Moving on to the second Axis configuration screen... I'm not really sure what to put in here. I want my machine/home/origin to all be the same thing and determine by:
   1. slow jog to the limit switch
   2. back off from the limit switch until the index pulse
   How is this accomplished

1. You can use PWM with different daughtercards ( PNCConf uses the PWM terminology because
earlier cards used PWM in the FPGA card for analog out. )
2. I dont think you can configure this is pncconf but you can chose the mode with the sserial_port_0=NNN token in the:
"loadrt hm2_pci" command line in the hal file
3. Once you have chosen a mode that supports analog inputs, you should see the analog inputs pins appear
if you list the pins ( halcmd show pin *7i77* ) You will need to connect these in hal manually
4. At this point pncconf is not really going to help any more and for tuning you are better of using the "Calibration"
menus in Axis
5. I would suggest reading the LinuxCNC documentation on homing

As a general note you should expect runaways when first setting up servo systems so be very careful to
protect yourself and any others working on the machine and also protect machine hardware during initial setup
Its often safer to disconnect the motors from the linear axis during initial runs
( though you cannot do final tuning with the linear axis disconnected )


gnipsel.com/linuxcnc/tuning/servo.html

Covers the basics of analog velocity mode servo setup

Also note that a 15 mv offset should have very little influence on motion once tuned. If you wish to null this out , you can use the bias pin of the PID component. Regardless of offset, the system should never drift because the drives should not be enabled when the PID position control loop is not running.

1) output depends on the card - 7i77 outputs voltage (behind the scenes PWM is used to make the voltage)
2) number one is displayed/disabled because of legacy reasons - set number 0 to mode 3
3) pncconf does not configure analog reading of pins - must do it manually after (by adding logic components etc)
I would work on servo functionality before adding the capability.
4) Open loop test is mainly used to confirm directions of motor verses encoder (to avoid runaways) and basic functionality.
5) Did the help button text/diagram not shed light onto this page?
Machine origin is Just a reference point for the machine to calculate everything else from.
If you don't have a separate home switch, then one of the limit switches will also be the hoing switch (choose positive or negative).
If that position is also going to be the origin the that would be set to 0
the opposite switch would be set to the axis travel distance between them (watch the sign)
eg positive limit to be origin and home switch (8 inches travel):
positive travel distance = 0
negative travel distance = -8
home position (where ever you want it between 0 and -8) = 0
home switch position = 0

search direction would be towards positive limit
latch direction would be opposite
use encoder would be yes

The help diagram tries to show a similar idea other then the origin is at the negative limit.

Chris M

I have read a whole bunch of different documentation links but I have to admit that my project is spread over a very long period of time. Things go stale in my head or I finding information again can sometimes be laborious... To that end, some of the questions I posed had, I thought, some directions that I felt I needed to take. However, it is safer to ask thank to not, hence the reason I posed a few.

The two questions for which I really was seeking confirmation were whether or not the PWM reference actually implied the ability to output PWM (reason being... my amps can support that) or if it was merely the way that the analog was generated. So, the latter was confirmed by you folks, which is great. The more pressing question was how to get this into Mode 3, something which I'll try to do today.

I recognize the pncconf is a beta product and am certainly happy to accept that this means I will have to modify the generated file accordingly. The limits of pncconf are a little harder to ascertain for a newbie, however. It can all be a bit daunting. Subsequent conversions would/will likely be considerably easier.

I recall reading the homing procedure but I don't know where. What I have learned from you guys, however, is that I have probably maxed out the ability of what pncconf is designed to provide. That is great and perfectly fine. Having confirmation that I need to move to the next step is a big win.

I did not see that tuning tutorial so that is a great link. Thank you.

I hope to move forward today. I have to admit that although I understand the concepts of how HAL works, I do have a bit of trepidation there. I need to get my limit switches tied in properly and the servo fault reset correct as well. I'll be happy when the servo system is all tidied up. After that it will be the spindle and then accessories (coolant, etc).

As I sit here, there is one thing that I am wondering about though. For the spindle, I am not planning on using the ENA line. The main reason is that I have to have separate relays for the FWD and REV. Ordinarily, I would run the relay power through ENA and then have outputs control the FWD and REV relays. However, I'm using a relay board and all I have access to are the control inputs to the relays. As such, I am using two OUTPUT pins for the FWD and REV. However, inside of HAL, I would like to logically tie the ability to energize the FWD/REV output ports through the spindle ENA signal. I'm assuming this is possible but I haven't really looked into it. I may have questions about that once I get to the spindle (bit off in the future).

Regards,
Alan

2) number one is displayed/disabled because of legacy reasons - set number 0 to mode 3

Doesn't this contradict what Andy is saying (that you cannot chose Mode 3 in pncconf). Either way, I can say that all UI controls within Num 0 of the 'I/O Connector 3' tab are disabled, so I can't select mode 3 there.

Did the help button text/diagram not shed light onto this page?

I had seen this originally but forgot it was there. However, I am/was aware of the direction specifications of a machine so that part wasn't especially helpful for me. The origin and home data confirm what I assumed to be the case. But, in general, I wasn't having problems with pncconf in this area so, although I think the diagram is great, it wasn't something that I personally needed all that much. That said, when I start to get into the manual configuration of the homing sequence, I'll almost certainly be coming back to the drawing.

Best regards,
Alan

2) sorry i tested the latest version of 2.7 possibly with bug fixes that are not released. What version of 2.7 are you using.


help page) i ask only because I'm always interested in making things clearer/better if possible. As i have being playing with linuxcnc for ten years or so - it's hard to remember what needs to be explained to less experienced new user or the best way to do that.
Pncconf was also built when I was very new to python programming and building user screens - not that I'm sure i could do a whole lot better now - it's a lot of info to display.

Chris M

Not a problem at all, Chris. And, yes, I am guessing that once I get a running machine, I'll look back on this all and say, "what the hell took you so long?" As a green bean, however, I can say that it is all a bit daunting. I'm very technical but there is a lot of intimidating stuff in all of this and the ramifications are pretty scary. And, to make things worse, I can't very easily remove pulleys so that I can test without actually moving an axis. I'm converting a Hurco and the axis are configured as follows...

Z axis - motor has a pulley but the encoder is on the leadscrew so I'll lose all of the feedback if I pull the belt.
X axis - direct drive motor with encoder on the rear of the servo
Y axis - similar to the Z axis, except removing the belt would be a nightmare to boot

The Z axis is the most reasonable for me to pull apart but it still would be pretty difficult. i don't have many good options there.

Regarding the version... That is a good question. I looked for a version in the Help file but didn't see anything. Before I go hog wild and starting things that, perhaps, I should not right now, it is probably safer to ask... "how do I find out the version?"

start linuxcnc from a terminal it will print the version in the terminal (before you pick the config)

or start linuxcnc with AXIS click on AXIS's help/ about AXIS - it will version the version.

Chris M

Well, that was pathetically easy. <sigh> But, even an easy answer could spell doom at the nascent stages, so I'm happy I asked.

Anyhow, I'm running: 2.7.12

I remembered that I have a servo with an encoder hanging off the back. Unfortunately I only have one and no extension cords. So, this trick will really only work with the X axis. Furthermore, I'm not sure how this will work once I start getting into the homing process. I'll have to run over and pass a screwdriver along my proximity switch, but assuming that works properly, the Y and Z (which I won't have powered up) will be problematic. It seems like what I really want is to have three motors all disconnected from the axis; that isn't very feasible for me, unfortunately.