What's new
What's new

Retrofit 93 VF2

nottamachinist

Plastic
Joined
Feb 14, 2019
Hello fellow Haasers

I’m fairly new to posting here, but a long time lurker. I just got a 93 Haas VF2 I’m planning to retrofit with Linuxcnc. I just tinker with this trade, so I don’t make any money with my machine(yet). I see quite a few people posting on control issues or seem sick of Haas holding them hostage for software.

I am a Linux guy by trade, so I’m not at all worried about getting the software part running well. I’m more curious to hear opinions from the group if it would be valuable to anyone to see this retrofit done. The plan is to have it well documented and freely available to anyone. Centroid seems like a good path if you make money with your machine and just want something push button. For those of us out there without deep pockets I was thinking this could be useful.

Thoughts?
 
It would make for an interesting read!
I've always wondered about swapping dead old iron to a modern low budget system.

Think about it this way. No one may want it today. But next year you may just have helped someone start their dream!
 
Yes Please, if you do this, please document well. I'm not a linux guy by any stretch of the imagination, I managed to get a CNC TIG welder going using it but if I wanted to update my linuxcnc distro, there is no way I could get it running again in any reasonable amount of time - so much changes with every release that non coder guys like myself can't figure it out again.
 
I am interested in doing the same but my execution might be different than yours. LinuxCNC is based upon mouse/keyboard inputs which will not be acceptable to industrial operators (especially the E-stop button).
My thinking is to replace the three board stack with a custom board that maintains all the original connectors (mine is a '91 VF-1) with a Beaglebone Black for the user interface and a Dynomotion board to do the real time controls. I do not know as yet if the Dynomotion board can give the correct output to drive the Haas servo boards.
The Beaglebone has two real time processor units that could be used for feeding the Dynomotion board G code.
I have a Saleae Logic 16 that I can use to monitor the servo board inputs.
 
Following along for sure.

My main concerns about it would be safety and reliability- Will you still be able to detect all of the alarm conditions like servo overload, toolchanger issues, etc. I considered doing this to one of my machines but couldn't dedicate the time to wrapping my head around and writing all the logic to make all of that work properly...but maybe it's easier than I thought.

Keep us updated, my old VF3 has good iron but the control seems like it wants to croak any moment now.
 
Way these old Haas mills are, they are worth more in parts than as a running machine.

Sell the take-offs on ebay, it will probably pay for the retrofit...
 
Servo overload is a feedback from the servo board. The other alarms are typically time-outs that are currently set in the settings/diagnostics page.
Many operations are a sequence of events detected by switches, feedback from the servo boards or spindle controller.
Obviously it has to work the same.
To part out the controls and replace with the newest version (if it could be made to fit) requires putting out a lot of money and waiting for someone to buy the old parts for a whole lot more than they are worth (often referred to as "capitalism").
Sort of like buying Windows... we do not support the old version anymore, buy the new one, still has bugs and will be outdated in a couple years and the income will pay the development costs many times over.
 
Curious if you decided to go ahead with this project. I'd like to see what you come up with on the servos/drives & hardware. I have a 93 VF1 that's probably getting a centroid retrofit.
 
hope you know a few things about servo tuning! People have done things like this before, but trying to sort out even the tool changer logic seems non-trivial.

The centroid conversion seems concise, I doubt you could do it better for less $$ of your time.

I've got a 92 VF2 that's going to the boneyard if one of any major components fail. Even though I have some good connections, there's simply no way to summon some of these old parts if they bite the dust.
 
I don’t yet. The tool changer scares me more than anything. I know the centroid and machmotion guys have done it before but I’m sure I will be cussing myself before it’s over.

The machine I bought looks like new but it’s missing all it’s servos and drives so it’s not like there are many other options to consider.
 
Seems if you're going to do a retrofit it would be better to replace the servos with something more easily supported, and that you can get replacements for. Cleapraths would only cost a few hundred per axis, and I'm sure save a lot of time and headache.
 
I’m trying to figure out the servos. I’m not really sure what the specs were on the originals, I can’t find anything except that they were electrocraft E728’s that are discontinued and very expensive. Google can’t seem to find me a spec sheet on them.

I’m looking hard at DMM servos. They have one with a 5/8” keyed shaft that would work with the existing flex couplers. 1.3kw. I think that’s in the range of what haas puts on the newer machines. AC servos like those would run straight off my 240 single phase too.
 
Does anyone here happen to know how the old (93 vintage) haas machines were configured for axis drives? I’ve seen one place selling a retrofit kit with what looks like 1kw drives for x&y and a 1.5kw braked motor for z. I was thinking the brakes came later after haas did away with the counterweight stuff. If I’m reading the haas literature I downloaded correctly, I think mine would have had the same e728 dc Servo on all 3 axis, with no brake.

I’m close to just ordering 3X1.3kw servos just to keep things simple. I’m guessing the dmm ratings are optimistic compared to the Yaskawa drives. I think oversizing them slightly would be a better mistake than the other way around.
 
I do not know what the VF-3 had but the VF-1 servos run on 160 VDC and the servo boards are good for 20 amps giving 3.2KW peak. Yes, same motor on all three axis on mine.
 
The truth is that if you remove the Haas control and install any other, you just killed the resale on your machine. Any change is a down grade.
 
Well after what seems to be an eternity, I finally have 3 phase power in my shop and have the machine up. It doesn't quite run right, but probably more my error than the machines.

This isn't dead yet.

My plan is to remove the parts that don't work very well and replace them with more modern components. I am thinking a 93 haas vf2 isn't worth too much money, so worst case I make a 3K mistake and move on to a machine that works OOB.
 
The parts I have for the project so far are a computer with FPGA cards to run the I/O and I am currently hunting down motors and maybe new screws. These ones are pretty rough.
 
Yeah I’m not into mine much more than 3k so not too worried about resale value.

I ended up having DMM build some 1kW servos that are based on their 86mm housing. I was having trouble finding servos that weren’t too fat to fit where the old DC servos lived.
 
While I appreciate your ambition, If that machine powers up, spindle is good, axis move at all.. It is worth time and basic materials to get into good operation.

Sounds like your more a software not hardware guy.. time to get dirty!

Presuming brushed motors..
The servo motors while discontinued are simple and rugged. Brushes are available and any number of competent motor repair shops can rebuild a failed comm and winding. Same for the servo drives, old school simple.

Wires for axis crack and fail - cost more in time than parts, fixed mine with surplus aircraft stuff, inexpensive and better than factory new!

99% of backlash on axis are from the thrust bearings - under $20 from any supply house and usually only need cleaning repack and adjustment. Invest in a high quality mag-base, 0.0001" indicator and hook spanners. If your serious you will need these tool anyway. Don't rush ahead with new ballscrews without a through inspection.

Draw bar is usually stuck from neglect and or has broken washers. Tool changer has loose or broken bolts. Rails and bearings worn out, not cheap but are affordable and available and you will need in anycase.

Even the motor drive control is not too bad at around $1,000.00 to repair with two year warranty.

In the end you will have a complete understand of your machine and an appreciation of IMO the best control in the industry.

Everything you need to know is in the service manual.

When you have finished and still want to build something Linx operated.. Build a pick and place gantry to load/unload parts. I believe there are some IO connections for this sort of thing.

If you decide to go ahead with your plans - let me know what you want for the parts, for the right price - I'll take them out myself. Surgically, not like not like some offers I have seen on ebay.

My Burta ('93 VF-0) had been neglected and dismissed for junk. Labor, shop supplies totaling less than $2,000 including new vice has her making 410SS parts to 0.001" or better all day long.

Regards,
B
 
I will post some videos of what works and what doesn't.

Looks like everything except the oil cooler on the top of the motor isn't pumping oil, and the ribbon cable to the display is sensitive.

Everything else seems to function.

As far as being a software person, I have built a CNC from a grizzly machine to learn the basics. Its terribly inaccurate, but helped me to understand some basics.

I don't even know where to go to find someone who could service this machine that isn't going to cost as much or more than the machine did.
 








 
Back
Top