Yaskawa V1000 / Mach3 modbus connection

[Mmpie]

USB RS485 modbus to X200

Someone with a different VFD but exactly the same issues. The solution may as well be written in Chinese for all I know about electrics.

Guess that means buying another RS485 convertor? Noise has to be the issue, it's the only explanation. Though I'd have thought a double shielded cable would have put paid to any interference...

Maybe if I bump up the voltage to the com port in BIOS it will make it a little more noise resistant? Is there a limit/recommended voltage to send to it?


Sent from my iPhone using Tapatalk

[A_Camera]

I think you are trying to do too many things without knowing why. You say it worked BEFORE you started messing around with brains? In that case it is obvious that the the brains is wrong OR you changed more than just added brains. Go back to that stage, deactivate brains and start over. Noise is NOT the problem if you can control the VFD manually. Once you can control your VFD and spindle with the Mach3 manually you can start experimenting with brains, but make sure you KNOW that the communication is still OK and manual control still works. I would NOT experiment with the voltage level in the PC because that might fry your PC. Have you tried to analyze the actual data which is sent from the PC? I think that the problem is mainly not at the VFD or the receiving end, but the sending. It is very likely that you send rubbish which the VFD does not understand so it is not answering. Get a communication port analyzer and check the data. Get a 5V power supply. A phone charger is very good for this purpose because it gives clear 5V. That is better than manipulating BIOS, I had no idea that you change the voltage level through BIOS, but I think that is not a good idea. Also, what is the problem with uploading video? Just upload to Youtube and place the link into the "Insert Video" here.

Electronics is difficult if you have no knowledge, no instruments and don't know what is what.

[Mmpie]

I'm back to bare basics now - fresh install on Mach and the VFD wiped to factory settings so I can't have accidentally/unwittingly have changed something bad now. No brain, I am still playing in modbus test.

I will start spindle, write, no error
Change frequency, write, no error for maybe half a second, then receive timeout
Every command now I have to sit clicking write rapidly until eventually it goes through. Receive timeout is just permanent.

If you can understand. This I why I think it's noise related. Ok when the drive isn't running and while it starts picking up but once it's running it starts going wrong.

YouTube, of course. I was trying to upload straight to forum.

Edit: YouTube keeps giving error when trying to upload from my phone. Will have to wait til I can get to a PC.


Sent from my iPhone using Tapatalk

[A_Camera]

In post #39 you are saying that:

It was working at first before I tried writing my brain but once I tried to actually use it and the crashes started the PC isn't getting any data back at all.

Try to get back to this point. I don't know what else you have done after.

[Mmpie]

In post #39 you are saying that:



Try to get back to this point. I don't know what else you have done after.

I see. Just to clear up then.

At first I copied the modbus config and brain from the post I linked to in my OP. That worked just fine, except the spindle RPM dro didn't update and there was no spindle run/stop command via modbus, that was still hooked up via digital input. The output frequency was reaching the PC though as tested in modbus test and the output current DRO updated perfectly ok.

So then I deleted his brain, added a config for spindle run/stop and wrote my own. My own brain just made Mach3 crash when I had the inputs on. Played around a bit and couldn't get it working so yesterday I deleted Mach3 and restore the VFD back to factory settings.

So right now the only thing I have done is enable modbus in Mach3, set up the motor parameters and modbus parameters in the VFD, and in modbus test try to test spindle run/stop, set frequency and read addresses 0023 - 0027.

No configs, no brains, no dicking about with any other settings.

I'm going to bring the PC and VFD home with me tonight so I can actually sit at it for a few hours rather than a few minutes here or there when I get chance, then maybe I will discover the problem.


Sent from my iPhone using Tapatalk

[Mmpie]

I think I have found a/the problem. When ever S+ or S- is linked to R+ or R- on the VFD it kills the connection. Unhook S + or - to kill the send/receive connection and the run/stop/frequency commands work exactly as they should, except obviously no data coming back to the drive so it's in a permanent state of receive timeout.

If the S and R terminals are linked nothing works at all.

Edit: Ive got it working. 4 wire RS422 into the PCIE serial card via a rs232 converter. No errors no timeouts just perfect. Seems my drive doesn't like RS485.

Sent from my iPhone using Tapatalk

[Dan911]

For RS-485 on your drive manual reads..........

parameter "H5-07 = 1"

dip switch S2 needs on.

[A_Camera]

I think I have found a/the problem. When ever S+ or S- is linked to R+ or R- on the VFD it kills the connection. Unhook S + or - to kill the send/receive connection and the run/stop/frequency commands work exactly as they should, except obviously no data coming back to the drive so it's in a permanent state of receive timeout.

If the S and R terminals are linked nothing works at all.

Edit: Ive got it working. 4 wire RS422 into the PCIE serial card via a rs232 converter. No errors no timeouts just perfect. Seems my drive doesn't like RS485.

Sent from my iPhone using Tapatalk

That's great. Anyway, I think I pointed out the manual and how to connect several times but you said the wiring and configuration was right...

Regarding RS422 and RS485, I think it should work with RS485 as well, but once again, CHECK that you did what the manuals require you to do, but at the VFD end and also at the PC end. On the other hand, if you have an RS422 connection then that is better than RS485 since it is a balanced connection and is basically immune against noise. There is no difference seen from Mach3 side, Mach3 has no idea about the hardware interface and doesn't care about that part. It wants data and does not care about how it gets the data as long as the software protocol is right.

Anyway, at this stage, if you have not already do it, you can start with the brains, assuming "I've got it working" means you can start/stop the spindle and control the rpm.

[Mmpie]

It's really strange, I can't understand it, I'd tried RS422 before and both protocols with/without RTS enabled with no joy. It was definitely wired correctly as per the manual. I just put it back in its home where there is already the 2 wire cable I made tacked to the wall and RS485 works fine now too.

The only possible explanation I can think of is a strand of wire or something must have broken off and managed to short S+ and S-.

Everything works fine now anyway so happy days. Thanks for all the suggestions/input.


Sent from my iPhone using Tapatalk

[A_Camera]

It's nice to know that it is working well now, regardless of what caused the problems before. But as I said, even if RS485 is working now, keep the RS422 is my advice. If I had that possibility I'd use that because it is much better than RS485.

[Mmpie]

Well I spoke too soon. Now I've got round to hooking everything up, spindle, motors, fans etc it's not playing nice.

RS422 is a permanent CRC error.
RS485 works a bit but keeps having the odd receive timeout and CRC error once every one or two seconds, and also when I set the frequency to 300hz+ the PC reckons it's doing -25000hz.

So there must be an issue with noise because nothing has been changed, the VFD is correctly setting the frequency I tell it to but the return signal is getting scrambled somewhere. PC and VFD on their own worked happily.

I think 0-10v analogue is far superior


Sent from my iPhone using Tapatalk

[A_Camera]

I think 0-10v analogue is far superior

Definitely not superior, but if that is the only one you can make it work with then that is what it must be. Otherwise Modbus is definitely superior to analogue control.

[Mmpie]

Well a bit of fiddling with timeouts, response times and refresh rates now has got everything working both inputs/outputs and no crashes. I hope the tighter RPM control, on screen feedback and ability to automatically pull the estop if the spindle starts struggling turn out to be worth it!

Thanks again everyone.


Sent from my iPhone using Tapatalk

[A_Camera]

Cool...

A lot of ups and downs (now it is working now it not...) but hopefully it stays this way now.

Anyway, I really think you should install a communication analyzer to check that your messages are sent, understood and responded as they are supposed to be, each without error. If there are many missed messages or other errors in the communication than you should fix that because that may cause problems, for example bad response times or no response at all to commands. It can be a safety issue if you rely on Modbus and don't have a reliable communication line.

[Dan911]

Well I spoke too soon. Now I've got round to hooking everything up, spindle, motors, fans etc it's not playing nice.

RS422 is a permanent CRC error.
RS485 works a bit but keeps having the odd receive timeout and CRC error once every one or two seconds, and also when I set the frequency to 300hz+ the PC reckons it's doing -25000hz.

So there must be an issue with noise because nothing has been changed, the VFD is correctly setting the frequency I tell it to but the return signal is getting scrambled somewhere. PC and VFD on their own worked happily.

I think 0-10v analogue is far superior


Sent from my iPhone using Tapatalk

Whether you use analogue or modbus either needs to be setup and configured correctly to work properly.

Neither 422,485,usb or serial are an answer and full proof to a excessively noisy system and should only be considered a secondary precaution. Addressing the excessive noise is the first.

1. What cable do you use from VFD to spindle, is it shielded and grounded on end?
2. Does your spindle have 3 prongs or 4, if 4 is it actually used? (most aren't)
3. Does your electrical enclosure have a ground star point.?

It wasn't until I addressed these issue I was able to use modbus successively and reliably.

and also when I set the frequency to 300hz+ the PC reckons it's doing -25000hz.

When I experienced my noise issues never once did it change a settings or read wrong, I would most certainly trouble shoot what's causing this.

A_Camera gave you the best advice at this stage, use a serial monitor to look under hood of what's happening, it can speak a thousand words.

[Mmpie]

I'll try that. I still have the odd timeout but it doesn't seem to be causing an issue, so it's not perfect but it works and I ran it for quite awhile earlier without a hitch. I'm not sure what else I can do now to fight that occasional timeout.

I have a 3 core CY cable between spindle and VFD. The braid is terminated at the star point via a drain wire and is only rolled back just far enough for each conductor to be able to reach its terminal.

The spindle has four prongs, I just got a new spindle with a four prong connector over and will eventually put a 4 core cable in and earth it directly, right now I just have an earth cable from the machine frame to the star point. This will probably wait until I get round to a closed loop cooling system as my router is in an enclosure with the coolant lines going through a silicone sealed hole. Doing anything with the cable chain is an absolute pig of a job inside the machine enclosure as there's little room.

Yes absolutely everything is to star point, as referenced above.

I think the negative frequency reading in modbus test is a bug, because I went ahead and wrote the brain and when using view brain the frequency is correct.


Sent from my iPhone using Tapatalk

[Dan911]

Sent from my iPhone using Tapatalk

Hi Mmpie,

I plan on adding the Yaskawa V1000 settings to a manual I'm putting together so I downloaded and took a better look and came back here to see what you used. I'm not sure if you gave up on Modbus but by the looks on settings your using for Mach3 I wouldn't doubt it.

In your Modbus configuration you have your run address(1 )and frequency address(2) being sent as single holding registers, it clearly states in manual and what I suggested on first page of this thread they need to be sent as multiple registers. I could make no sense of the input registers address's you are using(36, 39).

Dan

[Mmpie]

Hi Mmpie,

I plan on adding the Yaskawa V1000 settings to a manual I'm putting together so I downloaded and took a better look and came back here to see what you used. I'm not sure if you gave up on Modbus but by the looks on settings your using for Mach3 I wouldn't doubt it.

In your Modbus configuration you have your run address(1 )and frequency address(2) being sent as single holding registers, it clearly states in manual and what I suggested on first page of this thread they need to be sent as multiple registers. I could make no sense of the input registers address's you are using(36, 39).

Dan

Those settings were copied from someone else, linked in the OP, I just added the spindle run command as he was using an analogue or digital start/stop. For what it's worth the spindle commands worked absolutely fine on their own separate configurations but the inputs caused Mach3 to crash presumably due to a pile up of CRC errors and timeouts. I changed the start/stop and frequency to be under the same configuration after it was pointed out that's a better way to do it. 36 and 38 are the frequency and current output monitor addresses, as the pic states.

As per my previous posts, all working now as it should, both inputs and outputs.


Sent from my iPhone using Tapatalk

[Dan911]

Those settings were copied from someone else, linked in the OP, I just added the spindle run command as he was using an analogue or digital start/stop. For what it's worth the spindle commands worked absolutely fine on their own separate configurations but the inputs caused Mach3 to crash presumably due to a pile up of CRC errors and timeouts. I changed the start/stop and frequency to be under the same configuration after it was pointed out that's a better way to do it. 36 and 38 are the frequency and current output monitor addresses, as the pic states.

As per my previous posts, all working now as it should, both inputs and outputs.


Sent from my iPhone using Tapatalk

Than it does accept Single Holding registers even though manual differs. Happy to read up and running all good!

Dan