PEERPSI

137.9 nm pitch I was using the fully interpolated counts when I should have been using the hardware counts. I'm learning! Thanks all!

I think that was meant to convert the motor cycle length from inches to mm. 1 inch equals 25.4 mm.

Much brain pain later I think I have it. I was using position value when I should have been using hardware counts. Thanks again.

Ixx71 has a maximum value of 16,777,215 which means I would need to change my decode from 4x to 1x in order to deal with this. Is what you are recommending a way around that?

The manual is clear. After doing the math it would appear that my counts exceed the range of the variable. I'm thinking my only option here is to change the encoder decode from 4x to 2x.

Based on the example for Ixx71 below I calculate my counts to be 30.48 mm/cycle*line/.0001379*128 counts/line=28291806 which is more then 11,000,000 then the maximum allowed per variable range limits. My hardware is Turbo PMAC2 Clipper, ACC-51S interpolator, scale with a .1379 micrometer period and my motor has a 30.48 pole pitch. What are my options? Ixx71 Motor xx Counts per N Commutation Cycles Range: 0 – 16,777,215 Units: counts Default: 1000 2. A linear motor has a 60.96-mm (2.4-inch) electrical cycle. An encoder with a 40 micron pitch is wired directly into PMAC and “times-4” decode is used. Ixx70 can be set to 1 and Ixx71 can be calculated as: Ixx71=60.96mm/cycle*line/.04mm*4 counts/line=6096 counts/cycle

The ACC-51 interpolates your line resolution by 4096x, but the result is right shifted by 5 bits (/32) . Your "counts" /mm should be your line spacing/128. Be aware that Sony scales have a lot of variation in the actual line spacing and will need calibrating. They come with a calibration sheet listing the actual lines/mm, but in use these numbers require come adjustment. In the past we used a Moore step bar to do this, but now use an interferometer. So I did a quick check with a gauge block to confirm my counts and it matches. So now I'm onto phasing/tuning the motors. We have movement but I have to get some parameter input data from Aerotech as what I have isn't adding up as far as motor pole spacing. I've also found out that the Turbo PMAC2 setup application will not work because of the inability to deal with the 51S high resolution interpolator. I expect this is also why my open loop move does not work in that application either. In PMACPro tuning the open loop works, auto tune runs. However, step tests fail once I increase the step size to 1000000 counts. It gives me an FE error. If it wasn't for actual position matching commanded very well in the other step tests I'd suspect wrong direction. Could my Ixx71 and Ixx70 be wrong? They were set based on the first few steps of the Turbo PMAC setup application. Ideas anyone?

I now have high resolution counts. I also now know the difference between $$$*** and $$$. That said I'm not sure if I'm getting or understanding the number of counts I should be seeing. My scale has a period of 137.9 nanometres. By my math that means I should have 1mm/.0001379x4096=29702683 or does the position screen show the software counts at 1mm/.0001379x128=928209? In the NC setup program for counts per unit which would be the correct value?

Thanks for that Steve. After our conversation I did read closer the 51S manual and made the appropriate entries manually. However after a $$$*** reset all returned back to what it originally was. I expect that there is some configuration file that is loading everything and I'll need to make my changes there. Working on that now. Good thing is that after our conversation I got counts and I was able to set the direction as well.

Okay, halls confirmed. Turns out the suggested M variables aren't used with the Advantage 900 system and the 900 doesn't map the halls to an M variable. That is now done and they cycle as they should. I returned my endoder conversion table back to 1/T extension of incremental encoder. Now I have counts. Unfortunately not the high resolution counts I need. When I select high resolution interpolator PMAC2 style I get no counts. Some progress is still progress! So at this point I have limits, halls and feedback. If I can just get the high res feedback we can start moving motors! Ideas anyone?

I've gone through the Encoder table setup several times, checked for my Lissajous pattern with the preamp connected to the ACC-51S and it looks good. I'm not getting any postion counts. When I try to view $78800 in a watch window I get the message Cannot read value. Anyone have an idea where I might be going wrong?

Okay limit switch variables found and limits confirmed. After talking to Steve M. I determined my cards where not stacked correctly and I needed to install the ribbon cable from the CPU to ACC-51S. Still no feedback though. Also where do I look to confirm hall wiring?

What manual or guide should I reference for a list of the M variables the Advantage 900 CNC system would use?

Encoder Addresses for ACC-51S 1 $C000 3 $C010 5 $C020 7 $C030 2 $C008 4 $C018 6 $C028 8 $C038 Channels 1 – 4 of the ACC-51S correspond to PMAC channels 1 – 4 if the ACC-51S is connected to the main PMAC2A-PC/104 board; channels 1 – 4 of the ACC-51S correspond to PMAC channels 5 – 8 if the ACC-51S is connected to the ACC-1P board. ADC Addresses for ACC-51S 1 $FFC0 3 $FFC4 5 $FFC8 7 $FFCC 2 $FFC2 4 $FFC6 6 $FFCA 8 $FFCE ACC-51S channels 1 – 4 become PMAC channels 1 – 4 if ACC-51S jumper E1 connects pins 2 and 3. ACC-51S channels 1 – 4 become PMAC channels 5 – 8 if ACC 51S jumper E1 connects pins 1 and 2. I have my 8ES stacked ontop of the Clipper then the 51S. Does this then mean that my PMAC channels become 5-8? My jumper is still set to 2-3 as per factory default. To view the values in a watch window I would select $FFCO for my first channel analog level and $C000 for my counts?

E-Stop is an excellent first step! When Delta Tau first set up the NC system they defined a whole new set of M variables, ignoring the suggested m variables in their manuals. For my work I have created an offset set of the general motor M variables starting at M1100. This allows most of the examples in the manuals to be used with only slight modification. My order of testing would be to first verify feedback. setup a watch window to look at the AqB counter of the ACC 51 and make sure it is getting ttl counts. Add watch entries for the ADC channels of the 51 and see if sufficient amplitude analog signals are coming. Next look at the flag inputs to verify your halls and limit/home switches. Verify the encoder conversion table by observing the position display window. Verify the phase counter is counting. Open loop enable the drives and six step the motor verifying phasing direction and amplifier function. Find a copy of the Hall Effect Phasing app note and follow it's procedure for determining the initial hall angle. Some preliminary gains: Ix08,Ix09=1 Ix30=100000 Ix31=2000 ix32=2005 ix33=50000 I've been trying to verify my axis limit switches, they work to the Omega convertors I'm using to isolate them from the card. My Sunx LS work on 12 volts and I understand the card expects 0-5. I expected to be able to see them as well as the halls in the View Connector window but it would appear they are not available there. I confirmed 5 VDC to my scale preamps from the 51S board, previously I confirmed sufficient amplitude at the preamp from the scales, however I'm not seeing any counts in the position screen. I considered using the watch screen but I'm not certain what address I should be looking at. I'd love to have the time to read the manuals cover to cover and figure it all out for myself but too many outside distractions. Every time I sit down to troubleshoot another fire breaks out some where else! However, today marks the start of a new week!

Aerotech linears? BAL or BL 10-40? Those are good drives. In recent years they have been paying a lot of attention to biasing them for true AB operation. The bandwidth is not the greatest (1KHz), but workable and it's certainly an order of magnitude greater than the position bandwidth will be. Also, pay close attention to your encoder signal levels. Saturation is much more harmful than a little S/N degradation that comes from a low amplitude. BL 10-40. I was actually considering buying their NC system complete but they couldn't put together a respectable pendant or jog station. Finally put power to it today. Nothing worked save for the E-stop loop and my hydrostatic servo system. I love a challenge! I suspect the axis feedback failure is due to jumper settings, external versus internally powered. I/O I'm not sure yet. My Omega modules are showing triggered but the connector view shows nothing. One step at a time!

Thanks for this. I was able to get my E-stop loop sorted this morning as a result. I'm hoping to have limits, halls and feedback sorted by the end of the day. Moving axis by Monday! The linear motors and drives come from Aerotech, they are used on other machines we have here with success.

Thanks for the advice so far. My system includes a Turbo PMAC Clipper, stand alone not PC card. My accessory boards are 8ES and 51S. It is a lathe and I'm using Sony B78 laser scales with a BD95-T32 preamp that produces 1vpp. Sunx NPN inductive proximity sensors for limits, they were already on the lathe. I have several years experience with Delta Tau PMAC's as the majority of our lathes in this particular department use them for NC control. I've also built several machines previous to this project using CNC systems from a couple of different manufacturers. Unfortunately getting to go on a factory course isn't an option for me. I will however follow the training material as provided in the link. Right now my concern is that when I use the Turbo PMAC setup application it is asking for my attached accessories. I'm not finding either of them as an option. I also understand there is a certain way they must be stacked. I have the 8ES directly ontop of the PMAC with the interpolar on top of it. I understand that with the high resolution 51S it is to be set up in the encoder tables rather then as an accessory. My other question that I haven't been able to find a quick answer as to where to pick up the E-stop loop that is built into the Advantage 900. I've wired in a relay that will control a contactor feeding power to the amplifiers for my axis and spindle. As for how my I/O is wired, I'm using Omega I/O converters to isolate the loads/signals from the PMAC card. They are very flexible and allow for a wide range of voltages on either side. To connect everything I'm using Phoenix cables and break out modules. They have connections for the 5 volts for instance. This should come from the PMAC card via the connectors or does it need to be seperately connected direct from the power supply. When I power up the PMAC and issue a reset from PEWIN the card responds but if I view the connectors the icons for voltage are not lit as I would expect.

I have all the manuals but there is so much information in them and no real guide as to start here with this program or this step.

I have an Advantage 900 system with Turbo PMAC-2 Clipper. I've built CNC retrofit systems before but this one doesn't seem to have any sort of guide as to where to start. There seems to be three different software packages for configuration, one for the PMAC, one for the NC and then the NC/CNC software itself. Any advice or guides available would be very helpful. My system consists of scales for the linear motors, x and z. A spindle with traditional encoder. Nothing complicated at all.