[MikeTerasT]

Hello!

I know this forum is very nice and full of helpful people. But we require quite alot of assistance with setting up our new Servo system. We hit a dead end with our setup and cant figure out a way forward. Were a small hobby shop trying to make it in this competetive world and we are currently suffering alot due to the current setup what we have and our inability to get this Expensive servo system running.

If there is somebody here, who can help us up and running. And also help with adjusting/calibrating the system to its potential.
Im not sure how people will react to this offer, but to show my apprecciation I am offering a 100 €/$ Prize for the guy who gets us running. Or maybe Split it with the persons helping. But I am a man of my word. This is a real offer. I would like to offer much more, since this is so important to us. But sadly we dont have the finance yet for this. If somebody can help us. Man that would save my life

So, Intro aside, Our problem.
We built our own CNC frame with chinese cheap Stepper system. We got it working with UCCNC controller. But we never got it to work Properly. It was binding, slow, etc. So in hopes to fix it, we replaced the frame with a pro Chinese one. Same problem. So Now we went all out selling our kidneys and bought a Yaskawa Servo system.

We cannot get this servo system to operate with uccnc controller. We got the servopack connected with a Laptop(Through some ridicolous hazzle, creating an account on a japanese website through translate and downloading drivers for the newer servopacks what we have), and we are able to JOG a single servo with a laptop. But it will not listen to commands from Uccnc. And we are aware that even if it was listening to commands, there is a whole new world of "adjusting" the servos which we still know nothing about.

I added a bunch of pictures attached, with what we have and what we bought.

Yaskawa
Servopack: SGDXS-5R5A00A8002 / 200V
Servo: SGMXJ-08AUA6SC2 / 200V Ph3 / 4.4A 250HZ/ / 750 W
UCCNC Controller: UC300ETH-5LPT https://www.shop.cncdrive.com/index.php?productID=817
Breakout board: UCBB breakout board https://www.shop.cncdrive.com/index.php?productID=915

I really hope that I will find womebody here with some experience regarding our problem.

Thankyou so much!
Mihkel from Estonia.

[joeavaerage]

Hi,
post the manual for the servo drive. If is it a pulse (Step/Dir) type input then it should be very similar to diving a stepper. If however is a bus type then your in trouble.

Craig

[Vlod]

I think It uses its own serial MECHATROLINK bus protocol, it’s an open source if you good with programming… google it.
You can easy get soft to hook it up to computer and change parameters in the drive and operate the motor but it won’t help you to hook it up to cnc controller if it doesn’t support this serial bus protocol.
If you find someone to create a converter that would be a ticket.
Other way is to obtain a master unit where all drives going to be connected and controlled from, and it probably has it’s own software as well…
One more thing is these drives designed for 3 phase and might not work properly on a single phase.
It’s just my opinion.
May be someone would give you more promising ideas.

[joeavaerage]

Hi,

I think It uses its own serial MECHATROLINK bus protocol, it’s an open source if you good with programming… google it.

That is my suspicion also.

I presume you bought these units secondhand? If so can you return them.....they are not really right for your controller. As Vlod says you could probably program an interface, but its a long hard road.
What you need is Step/Dir input type drives. Did you check to see whether these drives were compatible before you bought them?

I use 750W Delta B2 series, and they are pulse (Step/Dir) input types. There is significantly more tuning and set-up to do than with a stepper, but as a broad view they are very similar to stepper drives.
They have free setup and tuning software which is a godsend if you have had no experience with servos before.

Craig

www.fasttobuy.com/flange-80mm-239nm-ac-motor-driver-kits-with-3m-cable-220v-075kw-cnc-servo-motor-for-cnc-router_p28084.html

[awerby]

Can you get your kidneys back? I doubt you'll ever get this servo system to work for you. But you can probably get the previous stepper system going with some replacement of parts. Often these "kits" sold on Ali Baba or wherever consist of problematical (but cheap) drives and motors that need a lot more power than the drives can tolerate to run optimally. Misalignment of linear rails can also cause binding. If you list all the components of that system, people here can probably help you get it running (with some swapping out). Steppers are a lot more uniform than servos, which each tend to use their own proprietary drive systems that nobody who hasn't been trained on them can figure out.

[Vlod]

It is not impossible but very expensive and do you have 3 phase power to run them.
I would listen to Craig and go with whatever your controller supports. You already have controller and probably license for software.
Ive got hybrid steppers last year from Amazon. Nema 34 with drive for only $85 a set. Connected them to my old parallel port BOB and had them running in a day.
However if you want high accuracy and industrial strength go with servos.
The specs on the servos you’ve got are so good, they even support scales… to bad you don’t have someone local specialist to help you.

[MikeTerasT]

Hello!

The Servos are bought new. Through a local reseller who orders a lot from China (A container a month) and has some deals set up there to get quality parts. I avoided ordering on my own for the fear, of receiving faulty units.
I do not belive that Faulty units are the reason here.

If the controller is wrong, than I will replace the controller. Not the other way around. Controller with software was 200 euros. Servos with the gearboxes were 3500€

Servopacks have an option in their settings, to either use single phase or 3 phase. We chose to go with 1Phase. This cannot be the issue, since it gives alarm If you use 1 phase power in real life and have 3 phase power selected in settings, and the motor can be jogged with a laptop with the chosen RPM/s. And the RPMS are correct.

It is just not listening to the signals from the controller and we think that we are missing something or did something wrong.

Sadly the Reseller has now experience in programming or getting these Servos to actually run, Since he normally sells whole machines already assembled in China.

Binding due to bearing misalignment is not likley, since we have aligned the bearings multiple times on 2 completly different frames with the symptoms not changing

Well our "****up" is that we did not know what to check before buying. I received information from reseller, that It should be possible, since some people are using UCCNC. But our hopes were, that we will figure it out on the go. If required, we will replace controller/Software.

The manual should be accessible from here:
https://drive.google.com/file/d/13rX...ew?usp=sharing

[joeavaerage]

Hi,

If the controller is wrong, than I will replace the controller. Not the other way around. Controller with software was 200 euros. Servos with the gearboxes were 3500€

I think you are mistaken, IF the servo drives are MECHATROLINK types then yes, you can buy a controller, but It'll cost more, possibly way more than the servos and drives!
You would be better off swapping the drives for the regular Analog/Pulse input types.

The first thing I note about these drives from the manual is that they ARE Analog/Pulse input types, ie they ARE Compatible with your existing controller.

Double check the part number of the drives. If the 5th and 6th digits are '00' then the drives ARE Analog/Pulse input types?

Do you have the setup software installed on a PC ready to program the Drives?

Craig

[MikeTerasT]

Hi!

Yes, double checked the numbers on the servopacks, numbers are SGDXS-5R5A 00 A8002, so they ARE Analog/Pulse input types.

And the software is installed, with the same software we managed to connect with the servopacks, and can Jog them aswell.
Software Yaskawa SigmaWin+ Ver.7.43

[joeavaerage]

Hi,
on page 139:

You need Step and Direction inputs, in Yaskawa's terminology PULS and SIGN.

You need and enable input, in Yaskawa's terminology Servo ON
You'll need at least one output, an alarm, from the drive as well.

Do you have at least these hooked up?

Craig

[joeavaerage]

Hi,
page 143:

You need differential signaling to the drive pulse inputs. DO NOT apply open collector inputs WITHOUT a current limit resistor. You're better of using differential inputs,
you'll want the speed that they offer at the very least.

What BoB have you got? Does it have differential outputs for the servos?

Craig

[Elernaeh]

Hi, I am a colleague of Mike's, so I can provide also some information.

The servopacks are configured so, that they don't need a separate SERVO On signal, they are ON, the second they get power.

The servopacks are directly connected to the UCCNC controllers breakout board.

To me it seems, that the SERVOPACKS are talking in another language then the controller. The way the controller sends its Direction signal, is not the same as described on page 246,

Controller almost does it like the variant I underlined in the attachment.

Also im not sure on the connections from the breakoutboard. Where and how should the servopack be connected?

This is the breakoutboard we have.

https://www.shop.cncdrive.com/index.php?productID=914

Connected the servoback directly WITH outputs O14 and O16

[joeavaerage]

Hi,

The servopacks are directly connected to the UCCNC controllers breakout board.

But you have not told me about the breakout board. For it to be direct connected then it needs differential outputs. Is that how the breakout board is equipped?

Controller almost does it like the variant I underlined in the attachment.

YES....UCCNC is and always has been a Step/Direction controller exactly the protocol you underlined. The default programming of the drive is likewise Step/Direction, Yaskawa call it
Pluse/Sign but its te same thing. You need to ensure that you have set the drive to pulse input mode, sometimes also called position mode, and that the PULS/SIGN option is engaged.

The servopacks are configured so, that they don't need a separate SERVO On signal, they are ON, the second they get power.

In most servos that is programmable, so it could be programmed OFF at power up requiring an active signal to turn the servo ON, or it could be programmed to be ON at power up and require
an active signal to turn the servo OFF. I'd trust an enable signal supplied by the controller....your call, but all my servos have an enable signal, and the servo is OFF in absence of the signal.

Where and how should the servopack be connected?

I've already posted a pic of that (pg143). Note that there are two wires (twisted pair) to each input, so two wires (pair) for the PULS and another two wires (pair)for the SIGN.
These are expensive drives, and I would suggest you read and most importantly understand the manual before you touch them Just wiring them up willy-nilly will blow one or more
of them up.

Craig

[MikeTerasT]

Hi!

We soldered up the plug which came with the servo. We used the following wires:
Attached is a picture on how we counted the pins as a sanity check.

Pin 6 - white/blue
Pin 7 - White/green
Pin 8 - Green
Pin11 - White/Orange
Pin12 - Orange

As we see those pins are responsible for:
6 is signal ground
7-8 are pulse +/-
11-12 are direction +/-

How we connected these wires on the BOB are as following:

6-GND
7-PULSE pin from BOB, O14
8-GND
11-DIR pin from BOB, O16
12-GND

Maybe 8 and 12 should go to 5v or 12v?


Currently all cables are disconnected from BOB, is our way of connection correct, maybe something is missing?

I added an export file of all our settings, the file is quite difficult to understand clearly, if it would help, I can add Screenshots of all the parameter's.
Regards

[Andreiir]

U connected wrong wires .
If ur driver have pin 7 dir + u need to connect to 5v wich is conected to bob , else u can use 12 v , but i think u need a resistor , and if u use 24 v u shure need an resistor , if u dont use a resistor u destroy ur input interface of yaskava driver .
If pin 8 is dir - u need to connect to dir output of bob .
If pin 11 is step + connect to 5v , or 12v 24v with a resistor , to drop voltage on octocoupler input .
If pin 12 is step - connect to output of bob .

[joeavaerage]

Hi,

6-GND
7-PULSE pin from BOB, O14
8-GND
11-DIR pin from BOB, O16
12-GND

Maybe 8 and 12 should go to 5v or 12v?

DO NOT HOOK ANYTHING TO 5V OR 12V OR 24V WITHOUT UNDERSTANDING WHAT YOU ARE DOING. It a sure way to blow the inpu optocoupler
of the drive, and they are not user repairable.

Adereii's post is correct if you want single ended signaling, also called open collector signaling....but you should be using differential signaling,
that is the fastest and most noise immune system. If you choose single ended signaling, then follow the rules on page 143 to the letter with
regard to the resistance to limit the photodiode current to 9mA.

A differential output is actually two wires, when one is high (5V) the other is low (0V) and then the signal changes so the first wire goes low (0V) and the other
wire goes high (5V).

If I'm not mistaken you BoB has differential outputs for the motors....so use them. Post the manual of te breakout board.

Craig

[joeavaerage]

Hi,
looking at the pictures you have a very early breakout board from CNCRoom. It does not look anything like the current model...so I can only guess what features it has.
If I were you I would consider getting another breakout board, there is nothing wrong with CNCRoom stuff, I'd just get the latest UB1 board. The UC300 plugs in directly on top
of it....no pesky ribbon cables required, and most importantly it has differential outputs for the motors.

https://cncroom.com/en/product/ub1/

Craig

[joeavaerage]

Hi,
of the pic you posted it looks like the motor outputs of your breakout board are just single ended.

Can you tell me whether they source or sink current?.

In particular I see that the board has a 24VDC input supply.......do the motor outputs source 24VDC to the drivers. If they do STOP IMMEDIATELY. You absolutely MUST follow the current limit advice
if the breakout board sources current, you CANNOT afford to blow the optocouplers.

Craig

[joeavaerage]

Hi,

Lets say for a moment that te breakout board sources current.

7-PULSE pin from BOB, O14
8-GND
11-DIR pin from BOB, O16
12-GND

Then you have hooked it up correctly, that is to say the positive current flows from the breakout board (O14 and O16) to pins 7 and 11, through the input resistor and on through
the photodiode to pins 8 and 12, which you have grounded. That should work. A couple of caveats.....if the breakout board supplies more than 5V then you must use an inline
resistor. Sorry to keep harping on about it but if you dont do this you will wreck your expensive new purchases....and it will not be covered under warranty. DONT DO IT!!!

My guess is however that your outputs are active low, that is to say they sink current also called open collector. I hope this is the case because then its most unlikely that you have already blown the photodiodes.

In this circumstance you have wired the photodiode backwards and it would not work, and would explain why you cant get them to go.

Hook pin 8 to 5V via a 300Ohm resistor (from the breakout board). Now the current will flow into pin 8, through the series resistor (internal to the drive), and through the photodiode and out to pin 7. Hook pin 7
to O14. As O14 'sinks' current....when O14 is active, ie ON and sinking current the photodiode will be passing current and the drive will 'see' the signal.

The way you have it wired at the moment you are trying to pass current in the wrong direction. A photodiode is still a diode, it will pass current in only the forward biased direction, and that
is contrary to how you have wired it.

A further note, at the risk of confusion, is that internal to the drive there is a series resistor but also a parallel diode, not the photodiode. While photodiodes are diodes they have very poor reverse voltage
breakdown characteristics, and thus manufacturers put an anti-parallel diode that prevents large reverse bias voltages from ever getting to the photodiode, this is done in interests of reliability.
If you were to measure the current you would find approximately equal current in either direction, and so using an ammeter is NOT useful for determining the polarity of the photodiode.
From a practical point of view you can ignore the anti-parallel diode, it plays no part in the signaling.

Craig

[joeavaerage]

Hi,
after having a closer look at your pics I now see that you are using a CNCDrive breakout board UCBB. A dual port breakout board. It is still a current part number.

The outputs are definitely open collector types, or rather open drain types, given the output device is a MOSFET. So in my previous post I suggested that you needed to reverse
the direction of your connections to the drive is correct. Do not forget an inline resistor of 180Ohm to 470Ohm from the UCBB 5V supply.

That should allow you to get your servos to work. The downside is that typically single ended signaling is restricted to about 200kHz, whereas differential signaling is good to at least
500kHz. Even if you don't need to signal that fast to start with, you'll eventually want to signal faster to take advantage of the servos speed and resolution. So my suggestion about getting a CNCRoom UB1
is still a good idea, but you can get your machine running without it, and maybe upgrade later.

I have the electronic gearing in my servos set to 5000 pulse/rev (for 1um linear resolution) and I expect the servos to go to 5000rpm when doing G0 rapid traverses for a signal rate of 416.667kHz.
My servo driven spindle has an 8000 count/rev encoder and at max rpm of 3500 rpm it requires a signal rate of 466.67 kHz. So you can see that you can certainly use the extra signaling speed differential
signaling offers.

I use Mach4, which is not dissimilar to UCCNC, and I use an Ethernet SmoothStepper as motion control, not disimilar to your UC300. I made my own breakout board, because electronics is my thing.
It has six axes (X,Y,Z,A,B and C), all differential outputs for max signalling to servos, plus all the usual inputs and outputs for home/limits/probes etc and a dedicated relay and PWM output circuit
for the spindle.

Craig