[Marty_Escarcega]

Marty,
- if you haven't already, set the inverter to display Hertz while running. this is a good debug too. (along with the fluke meter on the analog output of the allin1dc) to quickly see what the inverter is outputting when you command the spindle to spin.

- make sure the spindle overide button on the operator panel is set to 100%

- verify that the analog output adjustment switches are set to the 0-10 vdc setting a per the install manual as well. (fyi.. the allin1dc can be set to output 0-10vdc, 0-5 vdc, +-10vdc or +-5vdc for the spindle control output via the dip switches on the allin1dc and the plc program).

Hi Keith,
Yes, Teco VFD is set to HZ
Spindle Override is set to 100%
SW3 verified:
1=off
2=on
3=off
4=on
5=off

I did a short video that may help. I don't know where to look in the control, not sure if its the recent AUX programming that Scott did. Could we possibly start from square 1 on the spindle control and work out from there? Again, it sure would be of help to me to understand how the control, controls the 0-10V output. Sure would be nice for a wizard to ask, Call S1500, what did you get? Enter the RPM and the control adjusts the output voltage. :-)

https://youtu.be/pBcLKTe8-ug


Again, I hope its me, but again, I can't fix what I don't understand....

Thanks!
Marty

[cnckeith]

marty,
i would go back to basics and set the gear range parameters all to 1:1 (this eliminates gear ratios and plc program mcodes and aux keys setup errors/misunderstandings)
and verify that when you command the max rpm (set in the control) in this case 3000 rpm that you do indeed get 10vdc output showing on the fluke meter.

be sure to put the lathe gear head in the gear ratio that is 1:1 with the spindle motor. VERIFY this physically. put a mark on the spindle pulley and a mark on the motor pulley. turn the spindle 1 revolution and make sure that the motor rotated the same.

once 10 vdc is achieved when commanding max rpm (3000) then you can go mess with the gear range parameters to get them to match up with the actual gear ratios that you want to use (set up a hi, med and low no need to use any more since a vfd is being used) verfiy the gear ratios physcially, do not trust the made in twaiwan indicator plate, put a mark on the spindle motor pulley and a mark on the chuck and rotate by hand and calculate the actual ratio and that is what you put in parameters 65,66,67

[Marty_Escarcega]

A follow up for today on the speed control issues I was having.

I did a remote session again, this time with Mario. We did what I should have checked first, the analog output voltage from the All In One DC spindle output with nothing connected to it other than a Fluke digital multimeter. We commanded 3000 rpm, the max machine spindle speed, which should have yielded 10VDC. Instead it was 8VDC. We lost 2 volts with the VFD connected. Mario suggested testing the VFD with a 9 volt battery and I let him know I had a bench power supply to test with. I set up the bench power supply with 10VDC max output. Connected it to the analog inputs to the VFD, had the Centroid control up and running so I could monitor the spindle RPM.

I applied 5VDC to the VFD and got 30hz. I ran the bench power supply up to 10VDC and the VFD responded as directed with 60HZ. (No voltage loss) I thought I wonder if my CAT6E cable was a bit too long and had voltage drop over the small gauge wire. (You might recall, I had used a long cable when I was trying to chase down encoder quadrature errors, when I removed the VFD from the cabinet and set it on the ground). I got a shorter pair of wire and connected the VFD up to the spindle analog outputs of the control and commanded the speed 1500rpm and 3000rpm. I got 5VDC and 10VDC respectively on the Fluke DMM. SO, its not control issue, rather a wiring issue. 10' of small gauge wire presented a problem with voltage drop.

So, I proceeded to shorten up my CAT6E cable from the AIO DC to the VFD control inputs. Went to test it....guess what, X axis quadrature encoder errors are back. I got another single shielded twisted pair cable I had which happened to be 14gauge and use it for testing. Spindle RPM appears to behaving now. Checked with digital multimeter while commanding 1500 and 3000 rpm and got the correct 5VDC and 10VDC respectively.

The moral of this story? It is imperative that proper wiring practices be followed. There is a write up in the All in One DC Installation manual about it. That includes VFD placement, power cable placement to and from the VFD as well as control wiring. None of which are Centroid's responsibility, and are of the installer. I'm sure the majority of support calls to Centroid are related to poor wiring practices of which I am a victim.

At this point it is time to clean up the power wiring in the cabinet to the VFD and from it to the motor, and ground some shields at one common point. I hope this resolves the quadrature encoder errors, so that I can move on to finishing with the gearbox and speed tuning, which, I assume will go much smoothly now that I'm getting proper voltage to the VFD.

I want to thank Mario of Centroid support for some extra effort. While not his responsibility, he took some time to find the manual on the TECO CV7300 VFD and call me back with a suggestion. I appreciated that. I've got nothing but good things to say about Centroid support staff.

I understand they include a couple of hours of remote assistance to purchasers of their controls, and reserve the option to charge $90hr for support after that. A value in itself when you get stuck. I completely understand Centroid can't be responsible for user installed issues and their time is worth something, They did steer me in the right direction again.

Just reporting back my findings for the day in the hopes that this helps someone in the future.

More soon....
Marty

[Marty_Escarcega]

marty,
i
once 10 vdc is achieved when commanding max rpm (3000) then you can go mess with the gear range parameters to get them to match up with the actual gear ratios that you want to use (set up a hi, med and low no need to use any more since a vfd is being used) verfiy the gear ratios physcially, do not trust the made in twaiwan indicator plate, put a mark on the spindle motor pulley and a mark on the chuck and rotate by hand and calculate the actual ratio and that is what you put in parameters 65,66,67

Keith, I was a little ahead of you on this one and did exactly as you suggested to verify ratios. Thanks for the suggestion,
Marty

[cnctechml]

Marty,

I'm glad to hear that you got the issue resolved with getting the correct voltage from the ALLIN1DC to the vfd. Personally, I don't think that 10' of CAT6E cable should have a 2V drop. I am wondering if there may be something wrong with the cable itself like a short or a nick somewhere in the cable. You could probably verify this by doing a continuity test. You should be reading a value close to 0 ohms from one end of the wire to the other end of the wire.

[Marty_Escarcega]

Marty,

I'm glad to hear that you got the issue resolved with getting the correct voltage from the ALLIN1DC to the vfd. Personally, I don't think that 10' of CAT6E cable should have a 2V drop. I am wondering if there may be something wrong with the cable itself like a short or a nick somewhere in the cable. You could probably verify this by doing a continuity test. You should be reading a value close to 0 ohms from one end of the wire to the other end of the wire.

Well, I thought it odd as well. Funny part is when I had this 10' of cable, hooked up, VFD in the cabinet with the AIO DC, it resolved my quadrature errors. I shorten the cable up to about 24", Quad errors return. Now that is frustrating!

[Marty_Escarcega]

Well, today's progress, or lack there of. I pulled the VFD out of the cabinet, set it back on the ground, I also ran new wiring to and from the VFD and to the motor, installed a ground buss for a common ground point.
The machine jogs and moves around fine. No axis or spindle quadrature errors UNTIL I command the spindle to start up or if it does start and run, when it is commanded off and it gets to 0hz. I used CAT6E stranded cable from the control do the VFD. There is still some voltage drop which drops the rpm say 30-40 rpm. I used a separate cable and regained that rpm loss. I still guess that the small gauge wire is causing a bit of voltage drop. Though i don't think this has anything to do with the encoder quadrature errors. Spindle encoder quad errors happen occasionally, but more often than not. Again, these errors happen when either the VFD is commanded to start or commanded to stop and the VFD reaches 0hz
I'm leaning towards the TECO CV7300VFD but hoping for other opinions.

Thanks guys,
Marty


https://youtu.be/qpMTnG9TQGg

[cnctechml]

Marty,

It definitely sounds like it's an electrical noise related issue. I would make sure that all of your encoder cables, that's the axes encoders, spindle encoder, and handwheel encoders, are all shielded twisted pair cables to minimize any electrical noise from getting into those signals.

[Marty_Escarcega]

Marty,

It definitely sounds like it's an electrical noise related issue. I would make sure that all of your encoder cables, that's the axes encoders, spindle encoder, and handwheel encoders, are all shielded twisted pair cables to minimize any electrical noise from getting into those signals.

The X and z axis encoders were Centroid factory wired. I noted all had the shield drain wired soldered to the DB connector shells as they should be. The original spindle and handwheel encoder cables are twisted pair. I disconnected the handwheels encoder cables and had no effect on the issue. It's oddly coincidental that the machine can be jogged around fine with no errors until an attempt is made to start the VFD or if it's running, when it comes to a stop at 0hz, not during deceleration. VFD is set to 15 secs of deceleration, and seems only when it hits 0hz do I get the fault.

[Marty_Escarcega]

So, I didn't wire the VFD control on this machine. I simply replaced wire for wire. But I started looking closer at how the VFD is wired to the AIO DC control and it doesn't look right to me. The control was originally ordered NON VFD. (So it has its FWD/REV contractors and it appears some of it is still wired, see the red wires in the picture, 551 and 552, SPIN DIR COMMON and SPIN ENABLE. If I understand the control correctly the outputs are simply contact closures. These should be coming/going to the VFD not the contactors? I have attached a photo of the AIO DC outputs and the page out of the TECO Manual for reference.

Here is what is wired
TECO on left, AIO DC on the right:
S1-->Spin CW/NC
S2-->Spin CCW/NO
S6-->Spin RES
AGnd-->Spin Speed Com
AIn-->Spin Speed
24V-->Spin En Com
R1A-->Spin Fault
R1C-->TB1B-14 (AIO-DC Common)

If you don't mind, let me know how you would wire the TECO CV7300 VFD

Thanks
Marty

[cnckeith]

Marty,
You said it yourself... you didn't wire it.. so I think its time to simply run the correct encoder cables and terminate them properly as per the Allin1dc installation manual.
You have probably already spent more time fussing with someone else's poor installation on the encoder and vfd cables then it would have taken to simply run new and proper encoder and vdf signal cables.
encoder cable (and vfd cable) is cheap! (centroid sells it for $1.25 per foot) so no need to waste time trying to figure out someone else careless installation. rip it out and run good cable using good grounding and shielding practices.
On page 13 of the allin1dc installation manual you will find the build you own encoder cable instructions. here is the link to the install manual. http://www.centroidcnc.com/downloads...all_manual.pdf
on page 32 there is general advice on building an electrical cabinet to control rf noise.
wiring up a spindle motor to the allin1dc instructions start on page 45 of the install manual. both reversing contactors and vfd hook up is covered here.
also very useful individual circuit schematic for the allin1dc are here in this zip file (along with common complete system schematics) here is the link http://www.centroidcnc.com/downloads...ematic_set.zip
take a look at the PDF called. " allin1dc individual circuit schematic set" take a look at page 7. you can use this to cross refercene to the teco using the teco manual to determine what terminal labels on the teco correspond to Spin Forward, Spin Reverse, Analog In, Analog out etc...etc..
Best regards,
cnckeith

[cnckeith]

i took some photos of the centroid recommended cable for VFD signal and Encoder (or scale, or mpg..use same cable) please find below. (click on it for larger view)

[cnckeith]

you'll have to be sure to terminate the drain wire properly for the shield to work.
the ground that the drain wire is connected to must also be verified that it is a good ground connection for the shield to work.
the machine itself should have a machine to ground rod to earth connection (its a mistake to just count on the ground wire coming from the electrical box)
a quick google search will reveal several good web pages that go over proper machine tool grounding in detail so i won't repeat it here..
Attachment 331140
Attachment 331142
Attachment 331144

[cnckeith]

thought you'd like to see some encouragement.. below is a photo of an Centroid Allin1dc retrofit upgrade electrical cabinet of a Lathe (that was originally equipped with an Anilam.)
The cabinet has THREE VFD's inside it! One for the three phase spindle motor, one for the three phase coolant pump and one for the three phase tool changer motor. why so many VFD's?...so i can run the whole machine off single phase power! small VFD's are cheap.. and effective, and better and cheaper than running the machine off a phase converter. the machine runs great.
Attachment 331146

[cnctechml]

Marty,

In addition to all of Keith's suggestions looking at the photo of the outputs on your ALLIN1DC, I see that you do not have anything wired to the Spin Res. Com. terminal. You will need to install a jumper between that terminal and the Spin En. Com. terminal. If not, then the inverter will never get voltage to the signal to reset itself.

[cen199]

One of the problems is the VFD is on the floor with relatively small ground wires attached. The drive must be mounted on a grounded electrical panel, preferably the same one as the rest of the control. Ground the drive and panel with the biggest wire you can fit. Run all ground wires to a single main ground bus bar. It is not useful to work on grounding and shielding problems until the drive is mounted in its final location.

[cnckeith]

thanks for the eagle eye marc.. marty, you'll find this required relay common connection on page 7 of "allin1dc_individual_circuit_schematics_set.pd f"
Attachment 331148

[Marty_Escarcega]

Huh, I responded this morning but must have forgot to hit submit....I really need to start drinking coffee!

Gentlemen,
First thanks for all the suggestions!
After what I noticed yesterday afternoon, that things didn't look quite right with the VFD wiring, (and I spotted a couple of loose wires soldered to other loose wires and taped with mylar tape! I thought I need to clean up the mess, so Keith, yes, we are on the same page. I am considering removing the FWD/REV contactors (CNT3/CNT4) and their related wiring to clean that up. There still connected to the rotary disconnect. I will then follow the AIO DC for lathe with Inverter print and wire the inverter correctly.

CEN199, are you saying if I can stick a #6 ground wire in there I should? Electrically not necessary as the motor is 7.5hp, draws 19amps, I have #10's running to the motor to feed it and a #12 stranded for bonding. Are you suggesting LARGE wire for shielding purposes? I will remount the VFD back in the control cabinet. It must have just been plain dumb luck the other day when I was testing and had no faults with the VFD out of the cabinet and on the ground.

Keith, I will measure up what I need for cable and call in an order. I've had good luck with CAT6E, its 23AWG, twisted pair (even the pairs are twisted at different rates) stranded, shielded with drain wire. BUT, I will go with the stuff you recommend. I know the hand wheel encoders and their cables have no effect on the VFD/quad encoder error because they were disconnected and still the errors like clockwork presented themselves upon VFD start up or command off but not until it seemed it reached 0HZ.

Anyway as pointed out before, new control, should get all new cable and new encoders whenever possible. Otherwise one might well be chasing their tail. The handwheel encoders and spindle encoders are DIADUR/HIEDENHAIN ROD 426. They can be face mounted or have a groove near the face which is required to mount them to the handwheels on the apron. I'd gladly replace these as well if there are reasonably priced suitable bolt in replacements.

Today is study the prints, RTFM, measure up cable, place an order, mount the VFD and study how the thing was wired and is supposed to be wired....

Thanks again gentlemen,
Marty

[Marty_Escarcega]

Point taken, but I'm not about to drill a ground rod into my garage slab. :-) The power to the machine happens to come directly from the service panel only 10' away. 240V dedicated circuit as is the 120V to the control.
Good info Keith.

[cen199]

CEN199, are you saying if I can stick a #6 ground wire in there I should? Electrically not necessary as the motor is 7.5hp, draws 19amps, I have #10's running to the motor to feed it and a #12 stranded for bonding. Are you suggesting LARGE wire for shielding purposes? I will remount the VFD back in the control cabinet. It must have just been plain dumb luck the other day when I was testing and had no faults with the VFD out of the cabinet and on the ground.

The ground should be large to provide the lowest possible resistance. The larger wire circumference will also improve high frequency conductivity due to skin effect. This will allow the ground to drain off more noise. With the larger ground and panel mounting, you will reduce mysterious errors due to both radiated and absorbed noise. Assuming there are no other problems with the spindle encoder wiring, this will cure the quadrature error.

Under ideal conditions, the ground does not flow any current. I assume you sized the ground for fault current. Using the heaviest ground possible (rather than going "by the book") will make your integrations much easier in the long run.