Gregs

Hello, I spoke with Sina. ACC-24M2A units might be what you want. The ACC-24M2A, a two axis servo peripheral, can be used to control 2 channels of analog style amplifiers.

If it plugs in, it will work. Connectors were changed a couple years back, so there could be connector incompatibility between revisions. Also, there could be some jumpers that need to be changed, which will be described in the ACC-24E3 hardware ref manual.

There is no onboard 0-10V, but an ACC-28A would provide that capability. 0-5V absolute feedback is possible with option 12, but that option is not present based on your part number. (0-20mA is only available with an ACC-59E3 in a Power UMAC). Correct, for encoder loss detection, reverse the resistor packs, same as on PMAC1 PCI, and map those register bits.

I attached that document, and no, encoder loss bits are not in the suggested M Variables. There is no other quick way to backup/restore other than thru the PEWIN menu tabs. Encoder Loss Detection2.pdf

This question was answered on the other thread titled "PMAC PCI Lite Problem". By the way, please avoid asking the same question on multiple threads.

It turns out that the PMAC Lite PCI is a new enough product to have encoder loss detection, which will detect if complementary encoder signals are open circuit, e.g. A and A/ both open. Resistor packs RP60, 62, 66, and 68 have to be reversed (see under "Resistor Pack Configuration: Differential or Single-Ended Encoder Selection" in the hardware reference manual). Had to do some digging thru old app notes, but it looks like the following addresses could be mapped to M variables for indicating encoder loss on channels 1 thru 4, respectively: Y:$E801,1 Y:$E801,2 Y:$E801,3 Y:$E801,4

The problem occurs basically while the motors are being commanded to hold position, because no feedback and holding position look the same to PMAC, in both cases there are little or no encoder counts. Fatal following error should kick in during a move if feedback produces too few (or no) counts. The problem could be with the encoders, PMAC, or cabling. Not so likely for all the encoders to be failing. I would suspect the JMACH 60-pin ribbon cable which is common to all encoders. These cables are susceptible to connector contacts oxidation as they age. Sometimes you can tug on a flakey cable one way or the other to make it work or not work. One person told me his problem went away after spraying the connectors with contact cleaner. Ideally, change the cable.

You're welcome, but I have one correction. Apparently all saved setup elements go into pp_save.cfg, at default or not. I guess I was confusing this with the Generate Configuration function.

I thought that a little more explanation might help. In Power PMAC, the Save and FSave commands both take what is in pp_custom_save.tpl and add it to pp_custom_save.cfg. The difference is that the Save command also gathers all "saved setup elements" that are not at default settings and adds that to pp_save.cfg. So, like Steve said, FSave uses pp_custom_save.tpl, and Save uses both that and all non-default saved setup elements.

On a Server, there is Sys.ModbusServerBuffer[]. As far as I have been able to tell, there is no counterpart when Power PMAC is a Client. But, there are the ModbusRegisterRead and ModbusRegisterWrite commands which specify Modbus registers and corresponding PMAC variables. See attached documents. Power PMAC Modbus Command Specification 2014-01.pdf power pmac modbus design document.pdf

There are some video tuning tutorials on the Delta Tau website. Select Video Center under the Resources tab, and then Training / Tutorials. There a step by step motor setup slides that include tuning.

I attached a more detailed document.#tuning primer.pdf

Encoder position is not affected; it is always in reference to power on or reset position. Motor position is set to 0 unless using an absolute encoder in which case it is set to the current encoder position.

Just choose any unused motor number. Also, it might be a good idea to set Ixx24 so that other motors are not disabled on a fault. And, by the way, Ixx60 does not have to be > 0, unless you want to slow the execution rate.

Sounds like either it is being attempted to download PLCC programs for a PMAC with extra memory into a PMAC with smaller memory, or PLCC programs have been increased too much in size; therefore, the solution is to either reduce the size of PLCC programs or, if possible, increase the size of PMAC memory. For instance, the UMAC CPU can have either 128Kx24 or 512Kx24 for compiled/assembled program memory.

There needs to be a motor with with Ixx00=1, Ixx59=1 or 3, and Ixx60 > 0. Then, to enable the servo routine, enable motor xx. Remember that anything that disables that motor will disable the servo program. See the section "Executing the Open Servo Program" in the Turbo User manual.

Maybe the attached application note from the Delta Tau website will help.Circular Interpolation.pdf

Modbus addresses are only 16-bit values, so the max is $FFFF or 65535 and 300000 will not work in any Modbus device. I guess you would have to use a different set of addresses for passing thru Modbus.

Where or what parameter gets set to -5,000,000ish? Are there any PLCs running, such that might cause this?

See attachments.

Try using CMD "&1%%0". I believe the problem with using % with CMD is the % gets interpreted as a "format sequence" character (see under % command description in the SRM). And as Richard/Curt pointed out, Coord[x].DesTimeBase can also be used. For example, Coord[x].DesTimeBase = Sys.ServoPeriod/2 for 50% (see under Coord[x].DesTimeBase description in the SRM).

Please contact DT support for this issue: support@deltatau.com or 800-556-6766. I can't see anything wrong with the described setup. Can you send the wiring diagram? Are you using old ribbon cables? If so, it could be a problem with continuity in a connector, a common problem.

Also, there are a few PMAC functions that use Q-variables and so can overwrite values. These Q-variables are Q0 thru Q19, and Q100 thru Q127. If you are using ATAN2, Read, or kinematic subroutines, some of those variables would have this problem. See under "Special-Use Q-Variables" in the Turbo User manual. Just to be sure, I guess I should point out that Q-variable settings are lost with a power cycle or reset if they are not saved, unless battery backed memory is being utilized.

The board with the connector unplugs when the hold down screws are removed. The connector is soldered, and re-soldering would void any warranty.

As for your other question, Q variables function the same as P variables when there is only one coordinate system.