I recently acquired a Hammer C3-31 (the 4-function version of the A3-31), and have been really impressed with how quickly one can switch from jointing to planing mode. Simply remove the fence, flip up the tables, and flip the dust chute over. It can be done in under 10 seconds.
The part that takes quite a bit of time is raising and lowering the planer bed. For the dust chute to be positioned in jointing mode requires that the planer bed be lowered to at least ~7”, while most wood that I plane is around ~1”. The handwheel crank used to adjust planer bed height goes through 110 revolutions over the full range of travel, and that takes me approximately 60 seconds of rather vigorous exertion. I don’t really like sitting on the floor, cranking a handwheel for a whole minute when switching between jointing and planing. It’s especially irritating given how quick and elegant the rest of the changeover is. Eliminating the hassle of cranking the planer bed up and down would dramatically speed up the changeover, and make the changeover time comparable to how long it would take to walk between separate machines.
Felder offers a “Power-Drive” option on their Felder-branded machines, and for good reason! I want that. So, this is my explanation of how I built my own “Power-Drive” system for my C3-31 (which is equally applicable to an A3-31). I’m sharing this as “inspiration”, not necessarily “instructions”, so many details are omitted – but I’m very glad to share more specific details (part numbers, code, etc) if anyone wants to replicate this effort.
So: I need a motor, a way to mount the motor, some sort of user interface, and a way to power the motor.
First, choosing a motor:
I want to leave the handwheel in place for fine adjustments, so the key requirement for the motor is that the motor be able to be spun backwards. Many geared motors have worm-drive gearboxes, which can’t be run backwards.
I found that I could raise and lower the bed at a rate of approximately 80rpm (using the handwheel), and wanted the motorized version to be able to fully raise/lower the bed in 10 seconds, so a motor needs to spin at ~350-450 rpm.
I found a surplus motor made by ElectroCraft (series E240). It was rated at 60VDC (max), 5,000rpm (max), 31A peak current, 29oz-in of stall torque, 240oz-in of peak torque. It’s a brushed DC motor, so the speed is proportional to the supply voltage. Even at a reduced supply voltage, the speed was way too fast, so I’m going to add a gear ratio to slow it down.
(Note: I would highly recommend a non-geared motor for this application. I also tried a motor with a planetary gear head, and it was just so stiff when being turned backwards that I was afraid I’d damage something – like the raise/lower gears in the planer, which are plastic).
Second, the motor mount:
I considered 4 options for where to mount a motor.
1) Underneath: There is an Acme threaded rod that raises and lowers the planer bed, and is accessible through the bottom cabinet of the machine. Coupling a motor to that would be very elegant, as you could have it hidden inside the cabinet. Unfortunately, when the bed is fully raised, there is only ~1” of rod remaining under the cabinet, and it gets very close to a casting (~0.1” clearance). Perhaps a bolt could be welded to the end of this rod. The motor would obviously need to be on some sort of slide to allow it to move up-and-down, but I think a drawer slide would work well for that. Unfortunately, with the C3-31 being so big and heavy, I just wasn’t willing to try to get underneath it to work on it.
2) In place of the handwheel: I didn’t like this idea because I’d like to leave the handwheel in place for fine adjustments, and because I know many people use the digital indicator handwheel.
3) Outfeed side of table: The handwheel turns a shaft which turns a bevel gear that turns another bevel gear mounted to the previously-mentioned Acme rod. These bevel gears are exposed under the bed. So I ordered a second bevel gear from Felder (their part number 213FG) and tried to devise a way to mount a motor and this second bevel gear. I ultimately gave up on this, because the side-to-side forces were just too much, and I’d need some very precisely-machined bearings and mounts, and I’m just not a very good machinist.
4) Behind the handwheel: The option I went with was to mount a straight gear, about 3” diameter, on the shaft behind the handwheel. Getting the gear onto the shaft required some disassembly – the bevel gears need to be removed, the Acme rod manually screwed down into the cabinet base, some stop collars on the handwheel shaft removed, and the handwheel shaft slid out the OUTFEED side.
But once the gear was on the handwheel shaft, it was just a matter of drilling a few holes in one of the “ribs” on the underside of the table and “hanging” the motor by a mounting plate. The holes are obviously elongated to allow mating the gears, and the hex cap screw is in a threaded hole – since the ribs are slightly angled, this screw allows the mounting plate to be “squared” to the table.
20170309_091651 (1).jpg
20170324_104743.jpg
Third, user interface:
The Felder Power-Drive doesn’t appear to have the most efficient user interface. There is a left/right momentary switch knob to run the motor full-speed up/down, then a “jog” button to fine tune in the direction the left/right momentary knob was most recently turned. So if you overshoot, you have to go backwards, then forwards, then jog forwards, etc.
I wanted a user interface with tactile feedback: ie – a knob that could be turned left/right to go up/down, and the speed of the motor would vary based on how hard the knob was turned in a given direction. Essentially, I want a potentiometer with spring return-to-center.
From what I can find, these do not exist. Apparently, when this is implemented in consumer products, it’s just a regular potentiometer with an external spring added. So I found a toy RC car controller with one of those big knobs on the side and hacked it apart – it had the potentiometer and some external linkages with a spring to give me the desired action.
On the C3-31, at least, the “control panel” on the jointer/planer side is just an e-stop button. The black plastic housing that it’s mounted in is 2 pieces that are symmetrical – the “front” piece is the same as the “back” piece, just flipped upside down. So I removed the e-stop button from the “front” piece, filled the hole with black epoxy, then swapped the front and back pieces (since the back piece had no holes drilled in it), and drilled new holes to mount the e-stop and the control knob side-by-side.
(I’m pretty bummed that I had to put two screws through the front to mount the control knob, but I tried to obfuscate them with some black paint…).
20170323_161004.jpg
20170323_161016.jpg
20170324_104617.jpg
Finally, the motor power supply:
There are plenty of cheap DC motor speed controllers available (on Amazon, ebay, etc). Generally, these are not complicated devices – there are good “single IC” solutions to DC motor control, so these are often just a single chip put onto a circuit board with connectors, etc. The problem to solve is interface: I need a way to connect the left/right variable speed control knob to a speed controller AND have it go forwards/backwards.
Extremely conveniently, Texas Instruments makes a $49 demo board for one of their speed controller ICs complete with 60V external transistors, a MSP430 microprocessor, and a potentiometer. I bought this demo board, replaced the on-board potentiometer with wiring ran to my control knob, and hooked it to a 240V-to-48V DC power supply. The demo board comes programmed with code that interfaces it to a PC GUI program (via a USB cable), but the board also has a programming header to allow you to re-program the microprocessor. I wrote some new code that varied the motor speed and direction based on the control knob position and programmed the demo board with it.
The nice thing about the microprocessor is that it allows a lot of control over how the knob responds. For instance, I added a wide “window” of deadzone around the knob center, so it takes a very deliberate turn of the knob before it responds. Likewise, the knob position can be scaled (either by a constant or some non-linear function) such that it moves very slowly to a light touch but very quickly for a harder turn.
I opened the main electronics enclosure in the C3-31 to tap into the existing 240V supply. I put a toggle switch on it and added a liquid-tight fitting for a wire that runs to a new electronics enclosure containing the PowerDrive electronics. This second enclosure contains the demo board, the 240V-to-48V power supply, and a small 48V-to-3.3V power supply (to power the demo board microprocessor).
20170323_211753.jpg
20170323_211726.jpg
20170324_125155.jpg
Other notes:
1) The spring-return potentiometer was a huge pain. It might be better to just have “up”/”down” pushbuttons and implement the variable speed in software (ie – pushing the button causes the speed to be slow, initially, but speed up the longer the button is held down)
2) The motor is likely oversized, but I wanted lots of low-speed torque to allow very fine positioning. If you only wanted a high-speed up/down, I’d suggest a smaller motor.
3) I wish there had been room for a larger-diameter gear on the handwheel shaft. The “full speed” is too fast, and a little more low-speed torque would be nice.
4) The size of the motor and the large gear slightly interferes with the ability of the table to go all the way down. The original maximum depth was ~9”, and now it’s 7.5”. I have never planed anything even close to that thick, though, so I don’t think I’ve lost anything. I replaced 2 depth stop screws with longer ones to make sure I don’t accidentally crash the motor into anything. The demo board has built-in programmable over-current limiting, which stops the motor if I hit the top or bottom stops.
5) Mr. Chris Parks and I have been discussing how this might be fully automated. I already have a Wixey DRO on the machine, so conceivably this could be interfaced to the microprocessor and a more complex user interface could be added (perhaps a touchscreen with keypad) that would allow the machine to be commanded to a specific thickness setting. The problems I see with this are:
a. A GUI touchscreen (or even mechanical keypad and display) gets complicated in a hurry.
b. It would require some sort of control system – likely a PID loop – to be able to vary the speed without overshooting or oscillating around the desired position. Also tricky.
If I have time, I’m definitely going to try this, but at the moment, I’m getting ready for a move across the country, and just don’t have time right now.
Well, that’s it! It works! I likely spent 100x the time and effort I would have ever spent just manually cranking the handwheel, but this was fun and I love the result. Here’s a video of it in use:
I’m glad to answer questions, share part numbers, code, etc.
Reply With Quote
Originally Posted by Malcolm McLeod
