[IdefixRC]

Hi there,

I have recently upgraded my mill with a more powerful motor.
The motor is a 750Watt Servo Motor from GlockCNC. It is (at least from the pictures) identical to common Sewing Machine Motors out of china (e.g. https://www.amazon.com/dp/B016EJ1WB6).

The motor speed is controlled by a 10k - 10 turn potentiometer connected to the controllers 6 pin port.
Measuring across the poti (+ <> GND) gives me close to 5V (4.96v). The same is the case between + <> poti middle connector.
Increasing the speed through the poti the voltage between + <> poti middle connector will immediately drop around 2.5v and the motor starts spinning slowly. After close to 2 turns of the poti the motor speed will pick up (voltage at around 2.2V) till it reaches full speed at close to 0V (9+ turns).

Now when I hook the Controller up directly to the geckodrive as follows:

G540 VFD +10V <> Controller +5v
G540 VFD GND <> Controller GND
G540 VFD OUT <> Controller Speed IN

It will start spinning slowly as soon as power is applied (Gecko still in fault stage, Mach3 not yet started).
I'm able to control the speed but the motor pulses a bit (no spindle feedback loop incorporated)

So 2 issues:
1) Motor never stops
2) Motor pulses

Next I installed the potentiometer back in the loop and included a switch to switch Controller Speed IN in between G540 VFD out and Potentiometer middle pin.

Switching to potentiometer works fine - Manual control is not an issue.
Switching to automatic does not do anything. I can measure the correct voltage at Controller Speed IN but the motor neither starts nor does it change speed.

I'm attaching a picture of the wiring as well as of the STM8S00 based motor controller board.

Anyone an idea? Could it be that the motor controller is unable to work with the G540 PWM signal properly (not clean? saw a post somewhere that this might be an issue)?

[john-100]

you could find the G540's VFD does not work correctly using the 5V from the motor control board
from the G540 manual :-

" ANALOG OUTPUT: This is a 0V to +10V opto-isolated analog output intended for use with VFD drives. VFD OUT goes
to 0VDC while the G540 is disabled. Connect VFD GND, VFD OUT and VFD +10V to the VFD drive inputs. Make sure the
VFD drive positive voltage does not exceed +12VDC. Do not short VFD OUT to any other terminal. Do not reverse
polarity to VFD GND and VFD +10V or the G540 may be damaged. "

I would use an isolated 10V DC power supply just for for the G540 VFD
then use two resistors to drop the G540 VFD 0 to 10V output down to 0 to 5V for the motor control

only the motor control's potentiometer control voltage output needs to be disconnected


Attachment 370932


John

[IdefixRC]

Thanks John. I will give this a try.

[john-100]

for your initial tests you could use a 9V battery to power the G540 VFD circuit
it will prove the G540 will work using a 10V power supply before you buy one
like this from RS components
| RS Pro 12.5W Desktop Power Supply, 100 ? 240 V ac, 10V dc 1.25A Output Level IV 1 Output, 2-Pin UK, 3-Pin UK |

Attachment 370936

Attachment 370934

double check the motor controls potentiometer connections are not live !!
like they are on my mini lathes speed control

just because the G540 VFD workes from a 10V DC supply its easy to forget its possible to be connected to a live mains circuit

John

[IdefixRC]

Hi John,

thanks again for the suggestion.
I tried with a 10V battery and here is what I get:

Test 1: Signal line not connected to the VFD.
Measurement between GND<>G540 VFD OUT after voltage divider

- Mach3 spindle off: 0V
- Mach3 spindle on, speed 0: 0.07V
- Mach3 spindle on, speed 50%: 2.4V
- Mach3 spindle on, speed 100%: 4.7V

Looks good here.


Test 2: Signal line connected to the VFD.
Measurement between GND<>G540 VFD OUT after voltage divider

- Mach3 spindle off: 0V
- Mach3 spindle on, speed 0: 0.07V
- Mach3 spindle on, speed 50%: 1.8V
- Mach3 spindle on, speed 100%: 3.6V

VFD however does not recognize the signal somehow. The LCD which changes as soon as the spindle starts turning in manual mode remains in the "spindle off" mode (trying to find a manual for the controller online now to see if there is something that can be deciphered from the LED message).
No speed setting or repeat restart of the motor controller (with mach3 spindle engaged, or spindle off or gecko disabled) changes anything on this state.
In addition completely removing the Poti (motor controller +5v not connected) does not change anything either.
Could it be that the STM8 chip is not able to recognize the signal from the g540 and discards it as false input as it is different compared to that of the poti ?

Update: Found a lot more info on the Consew CSM1000 in the Taig forum. In essence the control voltage should be between 2 and 4V, below 1.4v it starts to throw the E7 error.

Since the CSM1000’s speed is controlled by a voltage it is easy to customize start/stop behavior.
Basically there is a 10kOhm resistor to ground so adding for example 1000uF on the input pin to ground it will take 7 seconds from full speed to stop when the input is disconnected and the capacitor gradually is discharging.
If the voltage is allowed too much below the normal 1.4V “Stop” condition voltage the controller thinks the Hall Effect sensor is bad and it will display an E-7 error. In that case if accidentally triggered, turn off the power and wait a long time until the on board capacitors are discharged and it works again.
In other words, the control voltage should vary from 2V to 4V to get 0 to 4500 RPM. See may previous post with the data for the details.

[john-100]

OK
now I have that information it looks like you need to add a DC bias to the G540 VFD output

may be some thing like this -


John

[IdefixRC]

Update to the update:
I found some interesting info in the Taig Group on Yahoo. The VFD is a CSM1000 equivalent.
Speed is typically controlled by an optical sensor connected to the VFD. If the voltage on the control line drops below 1.4V, the VFD will issue a hall effect sensor warning. With the Poti this is not a problem and it will clear as soon as it is advanced.
In order to not get the error at all with the Poti I wired a 70k resistor between VFD 5V and VFD IN (signal) - this resistor is not shown in the drawing below. With this the controller does not show the error and speed advances smoothly as soon as the poti is adjusted.

With the G540 however it does not work.
Inline with the measurements above, the voltage at VFD IN drops as soon as the line is connected (when using the 5V from the VFD or when using the 10V battery you suggested above).
What works though is to connect the G540, enable the spindle via Mach3 at around 40% so that 1.4V are measured between VFD IN and GND and powering the VFD up. No error and speed can be increased through Mach.
BUT: voltage at VFD IN tops out at around 2.8V never going beyond it hence speed idles in the 20% range of the motor. in addition the voltage fluctuates heavily causing the motor to pulse.

Below is the full schematic of the Poti controlled setup including the circuit inside the VFD till the signal hits the STM8S chip which must govern the speed setting.
The resistors inside the VFD are in my opinion responsible for the issue. As I have no idea how the gecko operates internally, question would be how to best proceed?

[IdefixRC]

Thanks John,
That would take care of the min voltage
However would it solve the voltage drop issue once the vfd is connected (see my previous post and the circuit I posted).

Thanks,
Eric

OK
now I have that information it looks like you need to add a DC bias to the G540 VFD output

may be some thing like this -


John

[john-100]

the diagram in post 7 looks like I expected but the control voltage range is a little different from the few controllers I have worked on
( they required 0-5 or 0-10V )

if the control voltage is too low at 100% you could reduce the resistance connected to the G540 VFD output terminal - change the 4K7 to 2K2 for instance
Attachment 371086
or




adding an op-amp configured as a unity gain buffer could help , as it will reduce the current taken from the junction of the two 4K7 resistors that sum the VFD control voltage and the DC bias
(Gecko used a LMC6032 CMOS Dual Operational Amplifier in the G540)

from an old post here on CNCZONE I found the circuit of the G540's VFD circuit

Attachment 371088



John

PS

circuit using dual op-ampbuffer

Attachment 371090

[IdefixRC]

Thanks a lot for the help John. I got it to work after playing around with the circuit and changing some parts around the input. :banana:

Next I will get the final 10V power supply which I want to install in the cnc controller housing, test the manual/auto switchover (currently it goes into error mode as the connection is lost - maybe a 50-100k permanent resistor will help here) and draw up the final circuit and share it here for others to be able to replicate what I have done. Should have it all working over the weekend :-)

[john-100]

great to see you have made progress

I am sure others will find your final version useful

John

[IdefixRC]

Success !!

After a lot of try and error and with the help from John here at #cnczone.com I was able to interface the CSM1000 motor controller with my #Gecko#G540 controller and control the speed of the motor via #Mach3. Due to the small voltage range in which the controller operates, I will still need to invest in a more reliable power supply as mine has a quite big voltage variance each time it is powered up throwing the tuning out of balance.




Below is the final working circuit (small modification to John's drawing above as I had serious issues getting the max voltage to the right level, plus addition of the resistor required to allow switching between manual and automatic mode) and a description of how to set it up.
I'm also attaching the CSM1000 manual which explains the error codes and how to change motor direction and CSM1000 circuit diagram (both thanks to the guys over at the Yahoo Taig Owners Group)

[Ron Thompson]

Any idea on how to reverse this with a digital input from the G540? Thanks!

[IdefixRC]

Hi Ron,
I do not see a possibility with the G540 only.
It will probably require a microcontroller in between. E.g. To act on a direction input pulse and control the CSM1000 reverse function (through the push buttons).
One major downside however is the resolution required to get it to spin at the right speed for threading. The control voltage range is very narrow. Small changes in control voltage result in a large change in speed. That would need to be aolved first in order to get the exact speed to sync spindle and feed. Not aure if this will be doable with thw provided controller or whether a different one would be better (in which case I would get one with step and direction input right away)

Any idea on how to reverse this with a digital input from the G540? Thanks!

[Ron Thompson]

Thanks, for your response. I thought that might be the case, but I wouldn't know how to do it. As to threading, all that is required is a reasonably steady speed. Even if if varied, the feed of the other axes are slaved to the index pulse, so should follow.

[Nilez83]

Hi guys,

I'm looking into using the consew type motors for my cnc converted mini mill and lathe but I am not using a gecko G540. I'm using the average Mach3 ebay breakoutboard that is seen everywhere. It does however have an analog 0-10V output for spindle speed control. Does this mean I can also use this circuit for controlling the motor speed? If so I don't really understand if the 10 volt power supply is necessary because the description of the board does not state it is needed for operation of outputting 0-10V at the ports.

Thanks in advance,

Niels

[john-100]

it will depend on which BOB you want to use and whether or not the consew potentiometer connections are live (hot)

the G540 makes it easy to control live (hot) speed control boards as the G540's VFD circuit is isolated from the rest of the G540's circuit

if the consew speed control potentiometer wiring is live CNC4PC C14 board could be an option

one chinese BOB I have reverse engineered is this

Attachment 386754

the PWM to analogue converter circuit is isolated from the PC printer port
but is connected to the same 10V regulator that powers the input circuit for the 5 inputs !!!!
mains voltages from the consew board will be a danger to you and damage the BOB

John

[Nilez83]

Hi John

Thanks for the reply, Darn I think that's provably bad news then. It does say the board is isolated from the pc side of things but I don't think it is from the stepper driver inputs etc.

[Ron Thompson]

I don't think it matters. I did it with the G540, but the Consew only has a little less than 5 volts on the potentiometer. I did wire it straight to the G540, but results were poor (erratic). I used a voltage divider and a 10 VDC supply similar to what some fellow did on the Taig Facebook group. That works pretty well. Here is a link to that group, and I am pretty sure the instructions are in the group files: https://www.facebook.com/groups/524365887660604/

[Ron Thompson]

Here is the schematic from the Taig site. I only used the bottom, as I didn't care for the switch. I used what I had on hand which was 10K pots so I doubled the value of all resistors to get the same voltage drop.