Electric Chainsaw Mill with load regulation

[wileur]

Hi
RE feed. I have found that instant reverse is very important for many reasons.
An ideal size of spindle for wire is 1 inch shaft. Most difficult thing with wire wind up is it piles up. 10-20 ft. Is a lot unless endless loop.
Even with endless loop my wire pull set up was problematic, crank by arm allowed hands on feel but wore out shoulder.
Eventually went to a tethered chain down one side of carriage. Chain path was low along carriage rail, up 90 degrees around a 3 inch idle cog, then over a 3 inch cog on a shaft and down to another idle cog, then 90 degrees to continue along rail to an end tether with tension adjust.
Chain best if just bolts, wood rest or something that allows sawdust to drop.
With wire it has to go around an idler or similar and was always getting displaced.
Chain was 100% no problem.
Wire was 1/8 SS aircraft. Did not last.
Just my experience, Hope helpful.
FYI if chain sharp all the time, you can cut 1000 to 2000 LF on one chain and 1/4 a bar. Wait for sawdust you can cut that in 1/2 easily.
I would jig sharpen, low rakes, each tooth sized, touch up 3 times in between by hand. Had but did not need extra oil. Echo used one tank fuel 1/2-2/3 tank oil.
Could keep finger on the middle of the bar 30 sec. after cutting.

Some good thoughts there! I really appreciate the time you take to share your experience!

 

[PP4218]

I'm going to use a 5.5kW three-phase motor to power the saw, a smaller motor for feeding and and a 24V geared DC-motor with wires and pulleys to lift the saw head. An optical rotary encoder will help keep track of the height.
The bigger motors will be controlled by two frequency drives I had laying around. I'll put some inductive sensors to act as limit switches to automatically turn of the saw and feeding when reaching the end of the log.

Now to my question/thought: Would it make sense to adjust the feed rate based on the main saw motor load? Since I use VFDs and modbus communication I can easily monitor current, power draw or rpm and should be able to continuously adjust the feed rate to accommodate for varying log thickness, dryness, density, knots etc. If the saw starts to bog down, the system will slow down the feed a bit. Likewise, if it cuts like through butter it can move faster.
From my experience in using an alaskan mill this sounds like a good idea, but I might be overthinking it.

If you are monitoring current and have a setpoint, don't forget to include a minimum to prevent "runaway". If using current sensing from the VFDs, or a ancillary CT, I would recommend at least two sources as inputs to your controller to provide safety redundancy with a safety shutoff in the event of sensor disagreement (minus hysteresis). Don't forget your e-stop circuit as well (it sounds like more work, but I've found that shut-offs all along the process makes control easier as well as safer).

This project sounds awesome! What controller are you using, and what IP rating are you going for?

 

[wileur]

If you are monitoring current and have a setpoint, don't forget to include a minimum to prevent "runaway". If using current sensing from the VFDs, or a ancillary CT, I would recommend at least two sources as inputs to your controller to provide safety redundancy with a safety shutoff in the event of sensor disagreement (minus hysteresis). Don't forget your e-stop circuit as well (it sounds like more work, but I've found that shut-offs all along the process makes control easier as well as safer).

This project sounds awesome! What controller are you using, and what IP rating are you going for?

Yes, I'll have several emergency stop switches as well as tilt protection.

Regarding current measurement I'll run it with manual feed to begin with, to see how things work.

I'm using an arduino nano since I had a few laying around and used to program AVRs barebones in the old days. I was considering a PLC or PI Pico, but just didn't want to spend the time learning them to the same useful level.
No points for elegance but it gets the job done.

The controller box is IP65 rated but the VFDs are only IP20, so I'll need to at minimum get some cover. Not planning to mill in the rain but you never know.

 

[wileur]

Got the parts for the bar mount and chain tensioning laser cut from 5mm steel at work.

 

[wileur]

A picture of the saw head lift winch. It's a geared motor from a wood pellet burner screw. Lots of torque, it lifted my colleague of the floor...
The encoder will go on the other end

 

[wileur]

It has been a while since my last post, renovating a house and a lot of other garden work comes in the way.
But the build is progressing and it is getting exciting. I now have almost all the mechanical parts prepared for painting and assembly. I'll get some pictures next time I bring the stuff out in the light.

The other day I skimmed through the manual for the VFD again and realized it has a torque mode that i previously overlooked. If I read it correctly it will allow the motor to operate in constant torque mode, which is what I have been trying to achieve with the ideas above. The main advantage of the torque mode is that no communication between VFDs is necessary, and that simplifies things a lot. No microcontroller is needed, only a potentiometer to set the desired torque/force. If the cutting is easy the motor feed is faster and when it hits some hard knots it slows down.
Am I missing something or can it be this easy?

 

[nedsim]

It has been a while since my last post, renovating a house and a lot of other garden work comes in the way.
But the build is progressing and it is getting exciting. I now have almost all the mechanical parts prepared for painting and assembly. I'll get some pictures next time I bring the stuff out in the light.

The other day I skimmed through the manual for the VFD again and realized it has a torque mode that i previously overlooked. If I read it correctly it will allow the motor to operate in constant torque mode, which is what I have been trying to achieve with the ideas above. The main advantage of the torque mode is that no communication between VFDs is necessary, and that simplifies things a lot. No microcontroller is needed, only a potentiometer to set the desired torque/force. If the cutting is easy the motor feed is faster and when it hits some hard knots it slows down.
Am I missing something or can it be this easy?

The potential problem I see is the stability of the control loop you're creating. Even though there's no direct wired feedback between the saw motor and the feed motor, you still have to face control loop issues, but without the ability (or perhaps limited ability, depending on the sophistication of your VFD) to tune the loop. My gut feeling is that the feed motor is going to constantly hunt for a stable setpoint and never settle.

I think you'd have a better chance with a DC or brushless motor on the feed. The response of that little 3-phase motor is going to be very slow and the relationship between current and torque is highly non-linear. Again, how well the VFD can handle that depends a lot on its smarts.

Not saying it won't work, but I'd make sure I had a plan B.

Coincidentally, I've used much the same motor (yours looks like an knockoff of Oriental Motors' small 3-phase gearmotors) plus a VFD in a positioning system where speed and positioning accuracy requirements were not at all demanding. It worked fine in the application, but even with encoder feedback and a tunable loop the system was slow and not at all stiff.

I've also built machine tools where we attempted to incorporate feed thrust sensing to detect a dull tool condition or other process problems. We've had moderate success using brushless servos on the feed, but it wasn't easy to get it right and not nearly as sensitive as we'd like.

 

[wileur]

Nedsim, Thanks for your input!

The small motor is from a feed screw for a wood pellets burner. Sadly it wasn't quite up to handling the heavy saw head, as it has a planetary gearbox which happily runs backwards once power is removed. I've since changed it out for stronger one with worm gears. I also added pulleys to the wire to get a 4:1 ratio and lifting and lowering now works great!

My gut feeling is that the feed motor is going to constantly hunt for a stable setpoint and never settle.

For the power feed, I was thinking that a longer averaging time of the load level signal might keep it more stable, or maybe not?. I don't think need fast and precise control of the exact power. It's more of 'it's getting rough, slow down a little bit'.
Anyway, I'll test that when I get around to the power feed. I'm hoping to assemble the rest soon (soon has become a more relative concept the more projects I start...) and push it by hand to begin with.

 

[Czech_Made]

Following

 

[Lightning Performance]

Just got back to this thread.

I assumed the drive track motor would be DC or DC brushless. My thing was planning on using a rope drum drive like a winch to allow a natural clutch like slip when the load got heavy. It will have drawbacks like wet weather and morning dew if you leave the drive rope attached 24-7 and your drive exposed to weather.
K.I.S.S. was my first and last thoughts there on moving the carriage along. The rest you gents have worked out. DC drives are usually more dependable and far easier to control in linear fashion. 12v, 24v or 36v is the question I have. 12v or 24v is my easiest out here.
Decided to swap my manual lift head to electric drive verses the manual on it now. Keeping the rack drive in the overhead with new custom made plastic bushings so the steel interslide wasn't all slop or use bearing guides. Swap the chain from the hand crank drop down mount straight over to an electric gear reduction motor 12v or 24v. My log bed will likely get a 12v hydropump on the log turner and wedge board. One motor for two tasks. The 14hp single commercial engine stays for now. It's pretty easy to swap that out for an electric motor or vtwin engine if needed. The main systems are all manual so far as produced by RR5032. Having electric will be nice at home. Solar pv and battery banks are in the works. Already have a 60" bar or a 41 to use. I think having a completely separate chainsaw carriage and track will be well worth it. Using as much as will work on both is fine by me.

 

[Lightning Performance]

Got the parts for the bar mount and chain tensioning laser cut from 5mm steel at work.

This will mount up pretty good from the looks of it.

 

[wileur]

I finally got things together enough for the first test run! The mill actually runs and ripped the (too dry) log! While there still a things to do I'm quite happy to have gotten this far.

In my eagerness to test it I overlooked the importance of securing the log. It moved a lot during the cut and the beam turned up quite crooked. Log clamps similar to Norwoods Portamill PM14 are lined up for welding next.
Second, it gets very heavy to push the saw after a while. But a feed of some kind is coming, that's the reason for this thread after all.

I don't have much reference, but milling speed was a bit slower than expected. It might be due to very dry wood, too high rakes or that the log kept moving around. More tests are coming.

As you can see in the pictures, fresh paint really makes a difference. Everything but the bar and chain are repurposed pieces or from scrap metal. Especially the motors look very fresh now.

The oiler is gravity fed, with a 24V pneumatic valve that is controlled from the main motor VFD. When the saw turns on the valve opens. The normal bar chain oil was a bit too thick to flow well, but vegetable raps seed oil works fine.
The oil nozzle on the bar mount is made from an M8-screw that I turned to size and bored on my lathe. I managed to get hold of a ca 1910 Bradford engine lathe which has proven very useful for everything from making oil nozzles to wheel axle shims and washers. (There is a good picture of the same model here https://doit101.com/Lathe/chapter1.html)

I've also had time to experiment with the iNVT VFDs and realized that I don't need the fancy control box with microcontroller. I made a new one with just the minimum of buttons, start, stop, up, down, left, right and emergency stop.

Again, a big thank you all for valuable input!

Stay tuned for more!

 

[nedsim]

Very nice. In case you haven't seen this, a little more info on your lathe.
http://www.lathes.co.uk/bradford/
I'm a lathe junkie -- if lathes were smaller I'd have as many as some of the folks on this site have chainsaws.

 

[wileur]

Thanks, I did indeed find the Bradford page. My lathe only has the importer's nameplate and it looked like a German model I've seen before, so I started withe German sounding names. Lathes.co.uk is great but it took literally a couple of weeks of clicking through all the links before I found Bradford.
I'm glad to have a piece of history in my garage.
My welder is another old thing, a 200kg ASEA three-phase motor-generator beast from around 1940 that can handle 200A continously.

A few photos:

 

[Fundyheather]

A 'right slope' on the tracks might allow gravity to provide a comfortable feed rate. Then you are 'just' winching the carriage back up for the return stroke. 'If' you ever get this right, the winch system on the elevated end that brings the saw carriage back to the starting point, could also drag another log up to the start position. There would have to be a small cart matching the carriage rails to keep the log high enough to miss the bunks as the log gets winched in from the low end at the ground. If you somehow achieve the saw feeding itself, by gravity, vfd's or whatever, then you want a significant time interval as it cuts to stack boards, file a chain, shovel sawdust, or sit with coffee. Points to a big change in how you work the mill, suggests a lower, slower power build out over all.