Replacing old Fanuc AC Spindle Drive with Delta VFD-E and Arduino

[FlyingTool]

Hi Lrapso, Yes the Delta drive can drive any voltage motor. The 230V is just the input supply voltage, but the output is set to whatever you want it to be in the Delta Parameters of the drive. You can drive a 50V motor if you want, or a 165V motor, etc. I only suggest this route if you are willing to part with a few more thousand dollars and a lot of time. After doing all this work I found out how easy it is to repair the old Fanuc boards and I actually have abandoned the retrofit for now. I found out all my problems were due to lack of knowledge on how to keep the Fanuc boards tuned up with all their analog voltage adjustments. Someone started a website where you can find and download manuals for Fanuc. I was able to find the service manual for my Fanuc AC Spindle Drive and adjust the variable potentiometers using my oscilloscope. If you have those tools it really is quite easy to tune a drive up. And if your good with electronics it's actually quite simple to swap out the transistors in the drive as well. You can find them for sale in a lot of places and it seems you can find even more used drives pulled out of old machines every day. Fanuc used surplus supply keeps going up and prices keep coming down. Hope that helps. I'm hoping to start a wiki on the projects I've done in the shop to document it further. While I am going to stick with the Fanuc drives for now, there is a lot of knowledge I want to share about how to keep them running.

Also FYI, you have to buy the correct input voltage drive if you want to replace it. So if your shop runs on 440, then you need to buy a 440V input drive, but it can still drive a 230V motor.

Wow this changes things...I am sure my spindle motor DRIVE is alright, its just that the fanuc controller is damaged. How do I control the drive with relatively simple signals? Does it have some sort of 0-10v feature? I have attached a image of the drive... Kindly see if you can have a look.

THANK YOU

[barchdesigns]

That's what this whole retrofit was about. Using fanuc drives with either Linux CNC or EMC2 or some other NC besides Fanuc. And then taking it a step further and replacing the drives so the whole machine is retrofitted. After 2 years of researching and throwing money at it, there is a lot to learn about these old Fanuc Spindle Drives. They can be controlled 3 different ways and you can find those details by looking back through this thread. Even if you have the service and wiring manuals for your machine it is a huge undertaking by the time you understand all the signals and how the Fanuc NC was programmed for your machine.

It looks like you have the same drive I have for my spindle motor. But like I said they can be controlled 3 different ways. I have 2 identical machines, but the NC commands the drive differently between the two machines. One is a 0-10V signal and the other is a BCD over 12 wires (Binary coded digit I believe is the acronym)

At this point I am still learning but it is for fun and not so much for the need to retrofit anymore. In the big picture if I were to do this all again, I would go one of 2 routes. Either buy a whole new retrofit package for between $5k-$10k and change it all out on the machine (a HUGE undertaking and probably just as many daily running headaches after it is all done), or the other route to sell the current machine and buy a newer one that will last 5 or 10 years without fixing or retrofitting it. The latter might cost you more, but will get you up and running what you want soon. I spent 2 years with a half broken machine trying to retrofit it only to get it back to original specs now. I could have just bought a slightly better machine and had it running for the last 2 years making money.

[FlyingTool]

I am going remove the fanuc servo motors and use stepper motors for the axes...I have used them on a lathe with my own arduino sketch running and it seems to be fine.
I actually thought of replacing the fanuc spindle motor with a regular induction motor and delta drive, now looking at your post I thought I'd give only the fanuc spindle motor a try. Saves me the new motor and drive haha.

So, can you help me figure out if I can use the existing drive as well? 0-10v mode would be great!
Thank You.

[landslide1]

Has this project gone any further?

I would like to see if it is realistic to replace a fanuc spindle amp with another brand of drive. maybe it is now possible using the Jon Elson's Pico Systems Home Page serial encoder convertor and the arduino control interface developed here?

[barchdesigns]

I got this far into the project and then decided it was getting cheaper and cheaper to buy old working used Fanuc drives as spares, so I haven't taken this project any further in the last 2 years. I've got a couple spare drives now sitting around (stripped another CNC) and I'm looking into replacing the transistors and possibly capacitors on the boards to refurbish them back to "like new" condition for my machines. It seems like that is where most of the problems are and it's a fairly easy and cheap replacement.

I have the PICO system here as well to test with, but again I haven't touched it in about 2 years now. I was just thinking about offloading all that stuff actually... I learned how to tune up all my Fanuc boards and so I have been found it seems easier to just maintain the Fanuc boards vs replace everything with newer stuff.

I was running the Fanuc AC Spindle Motor with the Delta drive and it worked fine for about a year, but I couldn't do tapping because I never got the encoder signals back to the drive correctly. Then once I figured out how to tune up the Fanuc Drive I have just stuck with Fanuc.

[landslide1]

thanks for the responce.
I'm not at a level I can rebuild a drive component by component so an off the shelf drive is appealing. finding quality components is getting harder and I’m popping IGbt modules so the diy approach is adding up fast.

modern drives are about half the cost of rebuilding a drive which makes it attractive to me. I'm supprised a VFD company hasn't already made an alternative considering the numbers of fanuc controls still being used.

[barchdesigns]

I was surprised a VFD company doesn't have a drop in replacement as well, that's why I did this DIY approach for a few years. Ultimately I ended up finding sources for cheap new transistors and caps and fixed up the Fanuc Spindle drive while I was still working on the new drop in replacement. If you are consistently popping your transistors, maybe it is time to look at the cause of WHY? I've been sourcing the new transistors and rebuilding the drives for about 10% of the cost of a new one. I'm not sure what drive you have, but my Fanuc AC Spindle Drive is really easy to open up, unscrew the wires to the transistors an replace them. I got a bit lucky and grabbed a few spare boards from working machines in the last few years for a good price. It's a bit easier and not so stressful when you have a spare Drive to work with on the bench while the machine is still making parts in the background

FYI, My XYZ Velocity drive units can be picked up used and working for under $100 now. A used Fanuc AC spindle drive can be found for under $500 if you look around. Put a few hundred $$$ into replacing the components inside, set the potentiometers while looking at an Ocilliscope and it's not that bad to get through. I'm self taught, and if I can do it anyone can. The machines are working better than they ever have (knock on wood) these days.

It sounds like a lot of work and knowledge, but in my shop by myself I found it was easier to fix up the working Fanuc stuff vs replace it with something newer and figure out how to interface it. Plus with all the interfacing it was starting to look like spaghetti wires.

Good luck which ever direction you go.

[mactec54]

thanks for the responce.
I'm not at a level I can rebuild a drive component by component so an off the shelf drive is appealing. finding quality components is getting harder and I’m popping IGbt modules so the diy approach is adding up fast.

modern drives are about half the cost of rebuilding a drive which makes it attractive to me. I'm supprised a VFD company hasn't already made an alternative considering the numbers of fanuc controls still being used.

If you are damaging your IGBT's then your Braking circuit is the problem or even just the Braking Resistors could be the faulty parts

[barchdesigns]

The hardest part of interfacing the Delta drive running the Spindle Motor was using the existing encoder feedback from the Fanuc Motor back to the Delta drive. We had to build a custom circuit that took the 2 analog signals from the encoder and turned them into 0-5v square waves for the Delta drive. There is an electronic chip called a Schmitt trigger which we had to use to do this. About this same time building the Schmitt trigger on a breadboard and getting it to work was the same time I learned how to fix up the Fanuc Drive so this was where I left the Delta Fanuc interface project. It's now all sitting on a shelf.

[landslide1]

my board is the a06b-6059-h003 quite common. and yes for a broken board it's about $500.

can you recommend a cheap oscilloscope that I can use to test the spindle motor encoder signal? That was the last error I was getting excessive speed deviation or spindle encoder fault.

I have checked the motor and cable megged it myself then got a shop to test them and they said it was fine and they ran it to test. I measured the braking resistance and it was 15ohms which what was quoted in a thread on this site.

Do you have a good source for parts? The Chinese IGBT I bought was rubbish and I’m better off buying second-hand ones but they can be blown also.

[mactec54]

my board is the a06b-6059-h003 quite common. and yes for a broken board it's about $500.

can you recommend a cheap oscilloscope that I can use to test the spindle motor encoder signal? That was the last error I was getting excessive speed deviation or spindle encoder fault.

I have checked the motor and cable megged it myself then got a shop to test them and they said it was fine and they ran it to test. I measured the braking resistance and it was 15ohms which what was quoted in a thread on this site.

Do you have a good source for parts? The Chinese IGBT I bought was rubbish and I’m better off buying second-hand ones but they can be blown also.

Post the original IBGT spec's manufacturer

The Braking circuit has a lot more parts to it than just the external Resistors usually when you have IBGT problems it is related to the Braking circuit

[landslide1]

Thanks for the assistance. there isn't much info on fixing these boards so I'm only doing what has been mentioned in threads on this site. now that you have sugested it it looks like TM1 , DM1 and C1 are involved in the breaking circuit. The manual dosn't give much info on testing these. I guess I'm looking for shorts or breaks in the diodes?

[barchdesigns]

Wow, This sounds exactly like the same problem I had a while back. I was getting speed deviation errors. Turns out NOTHING was actually bad. Really......it was just adjusting 2 potentiometers on the Spindle drive while watching the oscilliscope. Once I learned how to adjust a few things on these drives, it fixed all my problems with this machine. I'd have to look for it (maybe I already posted it somewhere too?), but I am sure I have the manual for the spindle drive somewhere, and within that it has a description of how to adjust these 2 pots. Basically they adjust the width of the square wave signal coming from the motor so the drive has a good signal to work with.

The posts to hook up to are underneath that spindle orientation board. Just loosen the lower screws and it hinges upward. Under that you'll find a 0V post and the 2 other posts for channels A and B signals coming back from the motor encoder.

I use a $200 Handheld Oscilliscope. It does everything I need and more for what I do with these CNC machines, and the price is right. https://www.amazon.com/SainSmart-Poc.../dp/B006J4FZMO

There really isn't much you can do without the Oscilliscope though. Best investment period I have made for maintenance of my CNC machines. Amazon reminded me I bought it 6 years ago....it's still going fine.

[barchdesigns]

Last night I got a Fanuc Alarm 2 on my AC Spindle Drive (LED 2 was lit up on the drive) "Excessive Speed Deviation" So I thought I would post how I use the oscilloscope to tune in the encoder feedback from the motor today. If you look at your AC Spindle drive you will see the smaller circuit board near the bottom, this is the spindle orientation circuit. 2 screws come off on the bottom of that and the board hinges up. Look for 3 posts under there on the main AC Spindle drive board marked ch7, 0v, and ch8. The oscilliscope reads between 7-0V and also 8-0V. These are your Channel A and B encoder signals coming back from the spindle motor. These wave forms need to be square waves at 50% duty cycle. That means the mountain of the square wave should be the same width as the valley. Usually what happens to me is the electronics drift over time (weeks or months) and my square wave gets to look more like 90% mountain and 10% valley. So I need to adjust the 2 potentiometers (RV 18 and RV 19) to get those square waves back in line.

once your oscilloscope is hooked up, start the machine and turn the spindle on around 1,000 or 1,500 RPM to start. Watch your oscilloscope and adjust the pots ever so slowly to see how it affects your square wave. RV 18 and RV 19 adjust the width of each channel's square wave.


FYI, the encoder signals coming directly from the motor are a sine wave and can't be used directly for encoder signals for any AC drive (whether it's Fanuc, Delta, or others). The Fanuc drive has a circuit on the board that takes the sine wave and turns it into a square wave (I think using a schmitt trigger type circuit).

[landslide1]

thanks for the how to.

I finally got an oscilloscope and tried testing the motors encoder. initially the duty cycle looked good @ close to 50% but the duty cycle could vary 10% either way depending on how fast I spun the spindle. I can't yet drive the spindle with an S command so I was following the manuals check procedure.

the other thing I noticed was that the voltage offsets for the encoder were a fair bit off. one was over 5 volts and the other was 4.6V but had a Vpp of 4.8V had a –ve zero offset instead of the +ve offset described in the manual. The manual states a vpp of about 4.1V with a 0.4V zero offset.

I didn’t have any luck capturing the RA RB signal as I can’t get a handle on the threshold function yet. The signal indicated the same type of error in the signals with one higher than the other.

IMAG001.BMPIMAG002.BMPIMAG003.BMP

[barchdesigns]

Your waves look good coming from the encoder. No need to test RA and RB as you are already seeing the square waves created from them and they look fine. It has to be really far off (like 10% valley, 90% mountain) to be any issue. And at low speeds even at 10% 90% it should work as the machine can still catch the blip of the 10%. Only at higher speeds will that become an issue as it can no longer see the blip because it doesn't have time to catch it. The mountain has to be there for x amount of time for the machine to catch it, and at max RPM it's close to the limit and more important to have a 50/50 wave form. Otherwise the mountain is not at 5V for long enough for the machine to catch it.

Honestly I would look elsewhere if you can't use an S command. If you're not able to command the spindle I would start checking the circuit for "machine ready". Do you have the electrical book for this machine? Have you checked your relays (assuming you may have relays in the back of the machine) to see if the correct ones are lit up?

For instance on my "Thing 2" machine I have a relay that goes on and off for "machine ready" state. It's random when it happens, but I have narrowed it down to a intermittent thermal switch somewhere. Could be in a motor, could be in a drive, but what I do for now is jumper the 24v relay to "trick" the machine into thinking all is good and is ready for machining. DO NOT CONTINUE TO RUN THE MACHINE LIKE THIS, it's just for testing to narrow down where and what the problem is. I know my drives and motors are not getting hot so I believe in my situation it's a bad thermal switch. These thermal switches also exist on the transformers for the machine.

I'd be happy to talk on the phone for a few minutes if you think it will help, send me a private message and I'll give you my phone number. I'm no expert, but have been doing this over 10 years myself and have figured out a lot of stuff to keep my shop up and running. Manuals are you best friend!

Benjamin

[Karolkaz]

I read it several times over, and I must say that I admire your brains and systematic problem solving skills.
I am troubleshooting an old Fanuc 10M, and would like to take advantage of your lesson first: try to fix the existing ( 1985 vintage) hardware, as it was built to allow easier maintenance and in many respects built to outlast many modern drives and systems. My challenge: I have no service manuals, not even the wiring diagram (just tracing cables physically by touch). You mentioned a location where these service manuals are available. Would you be kind enough to share? Thanks.
Also, if you don't mind, I was curious whether you considered the popular alternative of swapping the whole system with Centroid, leaving only the motors, and why was this option discarded? I already saw two success stories with these Ajax retrofits online, and was scratching my head, as it costs over $2k in initial investment.

[barchdesigns]

I read it several times over, and I must say that I admire your brains and systematic problem solving skills.

Thanks, this community has been a lot of help.

I am troubleshooting an old Fanuc 10M, and would like to take advantage of your lesson first: try to fix the existing ( 1985 vintage) hardware, as it was built to allow easier maintenance and in many respects built to outlast many modern drives and systems. My challenge: I have no service manuals, not even the wiring diagram (just tracing cables physically by touch). You mentioned a location where these service manuals are available. Would you be kind enough to share? Thanks.

cncmanual.com
Search for "fanuc" and it comes up with just about everything. I'm guessing the old manuals are past the copyright protection and are why they are available online still. The manuals were key for my understanding how to interface.

Also, if you don't mind, I was curious whether you considered the popular alternative of swapping the whole system with Centroid, leaving only the motors, and why was this option discarded? I already saw two success stories with these Ajax retrofits online, and was scratching my head, as it costs over $2k in initial investment.

I'm open to any and all retrofit options until this old fanuc hardware finally needs replacing. I hadn't come across any Centroid stories when I started this project a few years ago. I'll check them out, thanks. Sounds like a fair price. It seems every year there are more options and easier ways to retrofit.

[Karolkaz]

Thanks!
One of these success stories: . It caught my eye, because I own a very similar machine.

[mactec54]

Thanks!
One of these success stories: . It caught my eye, because I own a very similar machine.

This does not have much to do with the spindle motor and drive, this spindle motor can be run by any VFD Drive you want to use, a Fanuc Spindle motor is a different story and would not be worth messing with if you had a Fanuc spindle motor on a machine like this, you would just change out the Fanuc motor for a regular 3ph Ac motor and any VFD you want to use