This 6,000 r.p.m. encoder also looks promising and can be found for $25.00 or less.
See the data sheet for more info.
JoeyB
Printable View
This 6,000 r.p.m. encoder also looks promising and can be found for $25.00 or less.
See the data sheet for more info.
JoeyB
Care needed here.
Those encoders are only quoted to 2000 ppr, which is 500 lines/rev, and a maximum of 100 kHz output. Those are fairly basic specs. Yes, the bearings can take 6,000 RPM, but you won't be getting any signal at that speed.
Cheers
Roger
Roger,
Will the encoders I posted work reliably within the r.p.m. limitations of a 5tpi ballscrew?
JoeyB
The encoder should go on the motor. Not the screw. With fast turning motors it is unlikely they would be direct drive. For a 7000 rpm motor, you would probably have about a 4:1 ratio.The screws can take upwards of 2500 rpm depending on size. If geared 4:1 then the motor could be spinning upwards of 10,000 rpm. So you could exceed the rpm capabilities of that particular encoder.
109jb,
Correct me if I am wrong. Pulleys and belts or some sort of gear reducers would add a lot of cost to this arrangement?
I was under the assumption this was going to be some sort of low cost project.
JoeyB
Those are worst case 6040's which barely meet the rigidity requirements however, stiffeners on the gantry sides help and replacing the lower outer frame assembly for a steel version and the machine improves but cutting 7075-T6 becomes problematic since the table isn't rigid enough to take the down force due to the individual extrusion strips used to make up the table, solid aluminum T's bolted to a 7/16in cold-rolled steel plate solves that problem and adding a couple more rail supports stops the rail flex during plunging.
I also don't think those specific 6040's are worth the cost since I can get a solid table and cast frame kit for $200.00 more but if you're really hard up for cash it's a good starting place if you're willing to do some modifications.
Personally I'd go to a larger DIA round rail or use HiWin 4-way loading linear rails on a 3/8in cold rolled flat-stock (with stiffeners) lower frame assembly so cutting mild steel is possible and use one of Dale Walsh's integrated air-cooled 220V 3KW BT20 ATC spindle with 2500PPR encoder (I think he still makes it) or maybe a cheap water cooled 2.2KW ISO20 ATC spindle from china.
Enough about frames, back to ODrive and package variants.
Since some gantry machines like the 6040 and larger employ dual drives to move the gantry to prevent skewing, perhaps adding two jumpers wouldn't be a bad idea so that a single axis signal (such as y-axis) can drive both motors (daisy-chain) and the second jumper to invert the direction signal of the secondary channel (so both sides of the gantry move in the same direction) which is good for dual belts, rack and pinion and dual ballscrew gantry setups.
Pulleys and belts or some sort of gear reducers would add a lot of cost to this arrangement?
Everything adds cost. Everything.
Whether the cost of belts and pulleys would be worth while would depend almost entirely on what specs you want. For sure, toothed belts are usually cheaper and work better than gears.
Cheers
Roger
These type motors have the power, but at higher rpms, not at low rpm. They are going to need some kind of reduction system to operate efficiently. If the example 7000 rpme motor is capable of 1kW, that is about 190 oz-in. If we assume a perfectly flat torque curve that means it only has 192 oz-in of torque at low rpm too. So, I'd say that motor needs a belt reduction or something on it. A belt drive doesn't have to cost a lot. I have 2 belt drives on my current stepper project and the timing belts and sprockets were bought on e-bay for about $12 per set. A suitable set of pulleys and belts for this type of servo drive would probably be a bit more money, but certainly less than about $30 per set.
I personally don't envision this as a lower cost than a stepper system, or even the same cost as a stepper system, but rather a low cost for the potential performance. As far as servo drives go I think it has that potential at a much lower cost than other servo drive systems and am watching the development to see how it flushes out.
RCaffin &109jb,
Thank you for taking the time to explain in layman's terms, it's refreshing to get a positive point of view.
JoeyB
Thanks for the heads up.
I have tried to make sure that there is as much copper as possible in the current path that is expected to carry the motor current. The copper track impedance going around the motor current loop is approximately 1 milli ohm. This is 5 times smaller than the impedance through the MOSFETs in the same current path, so hopefully the amount of copper should be fine.
I have designed the board so that you can add a single heatsink for all the MOSFETs. There is also a thermistor placed as close as possible to each of the MOSFET group (3 thermistors total). The control algorithms will also have thermal control, so hopefully it should be fine with the current design. Of course if it turns out that things get too hot, we can always beef up the tracks at the cost of a larger PCB, or thicker copper at the cost of price.
I agree that the motors will run more efficiently if you do something like 4:1 reduction gearing. But I am interested in discussing what defines the torque requirements. My hypothesis is that you need a lot more torque on a stepper system because you have to deal with even the shortest load spikes with a decent margin to avoid skipping steps under all circumstances. However, with closed loop encoder feedback, I think you can be completely fine with a fraction of the torque.
As an example, suppose you had 4Nm of torque on a 5mm pitch ball-screw, that gives 8kN of force. I am not very familiar with machining, but I don't think you persistently need almost a ton of force for most applications.
That said, efficiency is ofc better to run things cooler, so it may be a good idea to use reduction anyway.
Cheers,
Oskar
Greatest torque requirements tend to be when accelerating up to speed. This is completely dependent on the specific machine, and the performance that the user is looking for. If you want 1G accel rates, then you'll need a lot more torque than if you're fine with 0.1G accel.
This seems like it might be a nice solution for a small, high performance machine.