[bbox]

Has anyone figured out how to use Mach3 to control the spindle speed in the Super X3? I looked at the motor wiring diagram and could not figure out a way. It looks completely different from a regular X3 mill to me.

Thanks, Bill Box

[devers6]

This would seem to be the Holy Grail for some of us here. I have been looking for a solution for some time, without results. Here's what I have done so far:

1) I communicated with Syil America to see if they would sell me their solution separate from their CNC conversion kit. The emails from Richard were friendly, but he seemed intent on trying to sell me the complete electronic package including breakout boards, drivers, and power supply. I was only interested in the interface board that plugs in between the SX3 driver board and the front panel, but could never get a price on just that piece.

2) I also communicated with Arturo from CNC4PC.com. He has a solution for the regular X3, but has not yet created one for the SuperX3. I emailed him some schematics that user OldRaven from this board had traced out, and he seemed to think that would help him figure it out. He was also going to communicate with Seig to see if they could provide him with the necessary info. I have not seen any results from this so far.

3) There are several posts in the Syil support forum on this board that speak to this issue, but none with a solution. One user replaced the entire SX3 driver circuitry with another 3 phase amplifier that was externally controllable, but I didn't see any details as to manufacturer, cost, etc.

So apparently there are many of us waiting patiently for a true solution. Welcome to the club. Hopefully the extra noise on the demand side will prompt some more attention.

[cjdavis618]

From what I can tell, the Boards are being created for them by this guy.

http://translate.google.com/translat...H_enUS271US271

So I doubt we will be able to order them through Syil or anyone else. The package that he has looks exactly like what is in the units I have seen pictures of.

I may have some useful info, but I need to compare some things first.

Devers6, I sent Arturo some specs from Grizzly's manual, as well as the photos of the control boards. He replyied back with some diagrams, but they still didn't match up with what I have for pinouts etc. It looked more like an edited version of the one on the website. Would you mind sending me a copy of those diagrams you mentioned? I have something that may be of interest to you also.

I just sent him another e-mail offering to send him an SX3 for a test platform. Maybe he will accept and help us get this figured out. There is no doubt he can certainly figure it out. He has the contacts with Seig to get the original Diagrams and such, but as of yet, hasn't been able to.

Maybe having the mill there to look at will solve this for us.

[John S.]

cjdavis,

Can you post a pic of the main SX3 spindle board.

I have some here but it's a UK one and could be different.

On the UK machines there is a main spindle board that has a daughter board plugged into it. Then on a ribbon cable there is a front panel that has the stop start revs up / down and tapping - then a speed DRO board.

On the KX3 CNC direct from Sieg it used the same main board and daughter board but the other two boards are missing and replaced by a small speed logic board that has it's own power supply and supplies 0 - 10 volt to the main board.

This AP2 board is all that is required.



John S.

[cjdavis618]

Sure. Here is my Grizzly sx3 board. So where can this AP2 board be purchased, and does it allow us to keep the original manual functions. Such as the control panel and the tapping features?





Wiring Diagram from Grizzly.

[John S.]

The AP2 board comes as standard on the KC1 and KC3 CNC mills from Sieg.
I don't know if it can be purchased separately.

It controls spindle speed and direction only and the KX series being full CNC's don't have the tapping feature.

I can't help but to me it looks as if that ribbon cable from the control panel circuit board, which is the same as the KX setup, will have two of those wires carrying 0 - 10 volts or 0 - 5 volts ? for the variable speed.

John S.

[Crevice Reamer]

...to me it looks as if that ribbon cable from the control panel circuit board, which is the same as the KX setup, will have two of those wires carrying 0 - 10 volts or 0 - 5 volts ? for the variable speed. John S.

I've been thinking that myself. I wonder if Cnc4PC could make a speed control board that just plugs into that ribbon cable and then (with its OWN ribbon cable) attaches to the daughter board.

CR.

[cjdavis618]

The problem is that we do not have a simple ribbon cable though. For some reason, that was changed to 18ga stranded wire in the molex. I went through the system yesterday with my DMM and found that almost all the cables on that run shows between 6.3 to 7.3 volts. None changed when using the buttons on the board to control the spindle.

.....But, I am questioning my DMM now. There were some flaky things happening with it yesterday to make m think it isn't reliable. I am picking up a fluke today and will re-test.

I have pics of the actual Control board and the Spindle Speed board now as well.

Here they are.

(note. the red and black lead at the top corner go to the Safety shield switch. That is not a voltage input)

The plug on the lower right is the connector for the Spindle readout. The upper is the one that comes from the main control board.

[arturod]

Chris,

Thanks for making me aware of this post.

As I understand this, the external control board generates an analog 0-5vdc that goes through the ribbon cable. So the thing would be identify the wire that carries the analog signal and the ground. This does not sound like a very hard thing to do. Our speed control boards come with a pot, that would make it easy to take down the signal from 0-10 to 0-5vdc.

Arturo Duncan
http://cnc4pc.com

[neilw20]

On the Syil SX3 CNC the cable from the front panel that went to the bottom of the motor control PCB next to the bigger green connector is daisy chained to a PCB with 2 relays on it. The relays are for M7 and M8.
The input to this PCB goes to the Break Out Board which sends the DC control voltage and direction signal. You need a revision 1.3 Spindle Control PCB to go with the piggy backed motor control setup.
The 1.3 PCB using the BOB DC signals, fakes front panel stop/start/faster/slower buttons, and the whole arrangement cannot be made to work properly. The control strategy is flawed/dumb/stupid read that as badly engineered by east crystal. A redesigned 1.3 type PCB is all that can fix it. Picture of 1.3 PCB and BOB included here FYI.

[cjdavis618]

Thanks Arturo and Neil.

I think at this point, it would be difficult to do anything but "simulate" the key presses on the mill. At least for full functionality.

I guess like any design session. We need to decide whether we are after the full features of the mill, I.e. forward, reverse, tap or simply forward speed with rpm control.

Obviously, at the very least, I would like to have the ability to control forward speed with the mill, but not removing the manual controls for simple manual tasks such as drilling and tapping. These would require the reverse and hand control button for tap. If there were a way to "tap" into the forward only controls of the mill, that would certainly be a start.

I am putting some fence up at the moment, once daylight is gone, I will be back in the shop getting some voltage readings on the new meter.



Neil, does your mill still give you all of the Sx3 functions? Tap and all? Along with Mach control of the spindle? If you do have Mach control of the spindle, do you have forward and reverse?

[phomann]

T

I guess like any design session. We need to decide whether we are after the full features of the mill, I.e. forward, reverse, tap or simply forward speed with rpm control.

Hi,

It should be too hard to produce a DigiSpeed to go in-line, between the 2 pcbs, if there is enough interest. Would need a little reverse engineering, but should result in maintaining the original functionality plus a mode to be controlled by Mach3.

Neil, I presume you have one of these mills?

Cheers,

Peter.

[neilw20]

Yes I have a factory CNCed SX3 with 3600 RPM belt driven spindle, but have other jobs too finish.:drowning:

I have spindle control from CNC
From M3 S3000
I have to wait at least 10 seconds, more like 20 while the moron cpu pushes the buttons trying to get the speed right by trial and error.
M4 will reverse after about 30 seconds after it tinkers with the buttons.
CNC operation of spindle is of no use in it's existing form.
M5 stops it eventually.
All the time the spindle runs, the beeper beeps, but you can't hear it above 1500 rpm.
The display show 1800 rpm for 3600. Chinese said multiply by 2. Yes, I can believe it.

I only use it in auto spindle mode if it is a long cut and I intend to walk away and let the wife shut it down.(nuts)
As there is no tool changer you have to do something after inserting a tool and setting the zero I press the start button, set the speed, then cycle start the program.:tired:
The Chinese can't fix it because communication between Syil and East Crystal does not work, or they have given up. They say it works!
Just push the button and wait! What a crock of 5hit.(flame2)
I had a problem with the power supply and after overcoming communication problems with Syil I had a power supply from China in 4 days.
The electronic/mechanical/integration thought paths do not exist. Syil don't appear to be able to write a spec (if they ever did) and East Crystal have proven they don't care about the scope of something. They had to make the buttons stop start, etc. but nobody made any rules.
Until it is up to speed, even the stop button is ignored!:argue:
NO SPINDLE SYNC SO NO SOLID TAPPING. IT NEVER WILL WORK UNTIL REDESIGNED :boxing:
grumble grumble. I have a circuit of motor board and that explains why it can't work. I even offer to fix it for free if Syil gave me the Atmel code and they did not even respond (to that part of the email). I bet Syil don't have the code, and never even knew they should own it. EC are big and could do it properly, but it is all $ based, and nothing else. They even saved $5 using an Atmel to control the motor instead of the original well designed Motorola BLDC.

:stickpoke

Cheers fellas. :cheers:

[phomann]

Hi Neil,

Thanks for the reply. I was think of inserting a new DigiSpeed board between the main powerboard that wants a 10Vdc control signal and the control panel board.

In the manual mode the signals would be passed straight through. In automode, the 10V would be generated from the DigiSpeed based on Mach3 input.

Obviously to do this, I'd need access to a mill or at least the information. Your mill could be the test bed.

Cheers,


Peter.

[neilw20]

If you visit me I can give you enough hardware (partially damaged PCB's) that will work well enough on a breadboard.
I am in Deer Park.
PM Me if you wish.

[extrapilot]

I had a quick look at the voltages/waveforms at the actual ATMega MCU on the daughtercard during mill ops. I have to say, it is confusing- and I do a lot of MCU work. Anyway, in checking the 6 analog/digital converter inputs, none of them seemed to have a DC signal, except I believe ADC0 which showed 2.5V at any RPM. Presumably, that is just a reference voltage then.

So, assuming that this is not a simple voltage level control, it is going to be digital signaling. There again, lots of strangeness. This MCU has I2C, SPI, and UART interfaces, but none of them shows anything like what I would expect to see. Pin 18 (slave select) is being toggled every 5ms, which suggests that the ATMega is being signaled via SPI. But I did not see valid data on the clock line, which sort of implies that data is not being clocked in. And, at least if the docs match this pinout, the PWM drive pins overlap the SPI pins, so if they are using native PWM to control the motor, it would be a hassle to use SPI anyway. One possibility is that they only switch out of PWM when they are receiving a message, and that is just a few milliseconds duration max. But the Slave Select pin, whih would typically trigger that, is in use (I think) by the PWM output. So maybe they are issuing an interrupt on another pin to kick it into SPI mode briefly...

Normally, I would say it is not typical to see SPI used in a noisy environment like this mill- lots of EMF and all. But the screen on the LCD/control board seems to indicate a SS line and a CS line, which implies SPI.

So, sorry, not much to add- other than- Ill look at this again later.

Best
Rob

[extrapilot]

Correction, pin 16 is the slave select, and was the pin I saw being toggled every 5ms, just in case anyone is not bored to death yet

[cjdavis618]

That is very good info Neil. Thanks.

Maybe that is why the syil X3 was short lived. I wonder what the components of the X4 are made from?

It may be easier to replace the entire controller board since we know the motor is simply VFD controlled.

[neilw20]

The control in SX3 is BLDC 1KW using hall sensors.
Concept OK. Implementation needs improving.
Without some testing I believe the input to the big motor PCB is digital, so no amount of snooping with a meter will figure out how the speed is controlled.
The are serial links to the display and the keypads.
As I said, the Ver 1.3 board is where it is all screwed up. It does not communicate with the big PCB in an orderly manner. Atmel code may be poor too. Not sure yet.

As for the X4 CNC
I will tell you in about 2-3 weeks.
I am inward goods inspection/critic for new incoming machine... soon.
Bet I will make a neat list, before they drop the ball this time.

[neilw20]

Exrtrapilot.

The ADC0 signal is the motor current measured across R414.
It will go low if the current is too high.
It's fed form an opto, which gives cycle by cycle current limiting during acceleration I think.