Hello Guest it is December 15, 2019, 07:28:38 AM

Author Topic: Dspmc Encoders not in sync with mach 3  (Read 6685 times)

0 Members and 1 Guest are viewing this topic.

Dspmc Encoders not in sync with mach 3
« on: October 01, 2014, 11:59:35 AM »
We had retrofit a 96' chevalier 63RD made by Falcon. We used the original SEM MTS30 M4 servos driven by the original 10v servo dynamics SDFP01525-17 Amplifier. The encoders are differential and made by the English Encoder CO. they are original to this machine and operate at 5-24v, 1250ppr.
we are also Using a Vital Systems 7761 DSPMC. The Servos are tuned to the drive setup graph with the red and blue lines on top of each other and seem to drive normally and smoothly. the encoders are feeding back info to the DRO's in Mach. However...

The DRO's do not register any movement on the settings page. When running a program the graphic display in the upper right corner does not reflect the actual tool path. It  shows a broken line instead of the solid white line it normally shows when working properly.

Mach also seems to not read the program accurately when running a program. it seems to be skipping blocks of code and/or misreading blocks of code. It follows the general path but no where close to accurately.

The servos run smooth and silent but take time to settle after a movement. occasionally the servos will creep as seen on the DRO's and verified by indicator readings.

Vital Systems support says that their system is working properly and that the conflict is between the working DSPMC and Mach.
The problem seems to be that Mach is not receiving all encoder information, or is misinterpreting it.

It has been suggested that there may be a missing plugin for the 7761, Has anyone experienced or fixed and issue like this???
would it be possible to write a brain to fix this?
 
Re: Dspmc Encoders not in sync with mach 3
« Reply #1 on: October 01, 2014, 08:38:16 PM »
A question.
1. What version of Mach3 are you using?
I have had some problems with the 066 and 067 versions of mach3 and the dspmc.
If it takes time for the servos to settle down after a movement they are not tuned properly. If the servos creep there could be a tuning issue or an issue with the drives offset tuning.

HTH Mike
We never have the time or money to do it right the first time, but we somehow manage to do it twice and then spend the money to get it right.
Re: Dspmc Encoders not in sync with mach 3
« Reply #2 on: October 02, 2014, 11:44:21 AM »
I do have the 066 version you spoke of.Where can I get the right version? the drives are really tuned pretty good and move accurately during testing. I can can get them set so the DRO's stay still.next time I turn it on they are creeping. The main problem is the DRO's do not seem to be recieving all the input from the encoders.it seems to get lost between the DSPMC andMach. the DRO'S for the encoders in the settings page do not operate corectly so I am assuming the problem is in Mach somehow.I need someone that knows how Mach receives the info from the encoders and the DSPMC.please give me some things to look for.I wonder if I could create a brain in Mach to fix it.thanks
Re: Dspmc Encoders not in sync with mach 3
« Reply #3 on: October 02, 2014, 01:32:17 PM »
As you are aware the 7761 is an obsolete model since 2010 and no plugin update is done since then.  so for compatibility you should use the mach3 version that was released sometime back in mid 2010.   Our latest DSPMC versions are 7762 S and M, and are actively supported for both mach3 and mach4.

If you have internet access on your mach3 machine, we can do remote login and try to see what could be wrong.

regards,
VSI Support

« Last Edit: October 02, 2014, 01:36:44 PM by Vital System Support »
Re: Dspmc Encoders not in sync with mach 3
« Reply #4 on: October 03, 2014, 10:58:53 AM »
Thanks guys at Vital Systems.We are getting the machine online and will be contacting you.I am aware I am working with older hardware(the 7761) it is such a well built unit that I hate to toss it. I will eventually be going to the 7762 as soon as funds allow.
i would like to however get to the bottom of what is going on with this one.where do I get a mid 2010 realease of Mach?can anyone hook me up with one?
Re: Dspmc Encoders not in sync with mach 3
« Reply #5 on: October 03, 2014, 09:38:57 PM »
Hello GfactShop:

I have a 7761 Vital Systems controller, and am running Mach3 ver. 066 with no problems. However, when I first setup the machine some years ago, I used the original encoders. My system is powered by WestAmp drivers, +/- 10v, and the original encoders were BEI, 2000 ppr. When testing, the machine would not repeat to the same position. it was off a small amount, maybe .06" in 10" or 12" travel from a given start position. This error was not noticeable unless you moved to a point about 10" travel from start, and stop with a dial indicator to indicate accuracy. Everything seemed in order, and the motors ran smoothly per setup with the graphics in the software.

My problem was the encoders shared the A-, B-, Z- ground wires, with a single wire. The Vital Systems I/O has connections for A+, A-, B+, B-, Z+, Z-, and 5v.
I changed the encoders to the AMT-102 and the Differential line driver, which has all the proper +/- encoder connections, and the machine holds repeatably to .0001.

My machine was a retrofit after the original controller failed after many years. The Amps, motors, limits, etc were all working fine, but my machine was "brain dead."
The machine sat for several years until I got the Vital Systems controller, which could run the analog servos.

The mill now runs smoother, faster than the original. The AMT encoders are small enough to put inside the rear motor case, and are connected to the wire loom with an RJ45 connector, wired into the original encoder loom, which did have all the wires for A+, A- etc, but was not required by the old system. The encoder connects to the differential line driver with a 5 wire connector, and the RJ45 connector, which has 8 wires,  plugs into the other end.

The AMT encoders are DIP switch settable with 16 resolutions from 48 to 2048 ppr.

The encoders are available from CNC4PC.


John


Re: Dspmc Encoders not in sync with mach 3
« Reply #6 on: October 05, 2014, 08:26:00 PM »
Thanks I will check the encoder thing out. I have much error in repeatability also.we are considering new encoders also.mine are EMC 755 HV 1250ppr.Quad ,A,B,Z 5 volt and Grnd, 8 wire. seem like good encoders.anyone know about these encoder good or bad? I have the grounded to COMM.is that bad? seem much more stable than hooked to the 0v line.
On Mine under settings the DRO's do not move,are they suppose to move?the crosshars on the program run screen are not following the tool path.under program run it seems to read the G code,X,Y,Z even feeds and rapids.it does not read the actual coordinate move amount. a circle will come out 30% larger than it is suppose to come out. we are set at 51020 pp/unit.
Re: Dspmc Encoders not in sync with mach 3
« Reply #7 on: October 05, 2014, 09:30:37 PM »
Hi Gfact:

I don't know anything about the EMC encoders. For the 5v supply to the encoders, you should setup a 5v supply, and not take the power from the I/O board. When I setup my machine, I kept getting errors on one axis. Upon checking the voltage at the encoder, I was reading about 3.5v and not the full 5v. I changed over to a separate 5v supply for the encoders, and input the 5v on the I/O board where the encoder wires connect. This cured the problem. It was a couple of years ago, so I can't give any details, but the low voltage was a problem as I remember.

The AMT-102 encoders with the differential line driver made a big difference in the machine. I have had zero problems in homing, and accuracy since.  Also, my original machine wiring was very high quality, with shielded cable everywhere, and individual wires in the loom with foil. Also, each pair of all the signal wires, i.e. A+/A-, were twisted pairs with separate foil wrapping on all the pairs.

When connecting the RJ45 connector to the loom wires, I make sure to use the twisted pairs of the RJ45 wires match the twisted pairs for all the signal wires.  You have to do a continuity check from the I/O board, all the way back to the RJ45 connector to make sure there are no mis-matched wires.

From what I have read about grounding, it is not a good idea to share power supply grounds because of ground loops. This can occur when there is a very tiny voltage difference in grounds, causing a current, which then creates a ground loop. Since the encoders are only 5v., the grounds are more vulnerable than the 24v signal wires that the dspmc uses. The dspmc uses a 24v supply, and the encoders can have a separate 5v supply. Also, make sure the AC wires are properly connected, i.e., Hot, Neutral, chassis Ground green wire.

Do not connect the DC negative to the chassis ground, as that can introduce noise into the system.

From your description of DRO's not moving, crosshairs on the run screen etc., it sounds like there is a major malfunction in the program behavoir. You might try re-loading the program, and also make sure there are not any old programs running in the background. The forum has various accounts of some programs interfering with Mach3.

I would suggest getting a new hard drive, and start a clean install from the ground up. Considering the cost of machine, tooling, and time spent overall in setting up a finished machine, the hard drive is a small expense. You can re-install XP from a disc, but Windows 7 will have to be a new program. Newegg has Win7 on sale for 85.00 this week. This way, you know exactly what the computer environment is.

That is all I can think of at this time.

John