Parallel vs series on your steppers

New Home Forum Updates Parallel vs series on your steppers

This topic contains 56 replies, has 15 voices, and was last updated by  Ryan 2 years, 3 months ago.

Viewing 27 posts - 31 through 57 (of 57 total)
  • Author
    Posts
  • #18150

    Jason
    Participant

    so If i went from 1/32nd stepping on the Z to to 1/16 i would just cut the steps/mm in half? 1/8 would in quarters? etc

    #18152

    Ryan
    Keymaster

    yup, you can use the prusa calc if needed.

    #18153

    Ryan
    Keymaster

    you can move faster but I did not notice any more torque.

    Interested to see if you find that as well or maybe there is more power?

    #18157

    Michael Hensen
    Participant

    any good schema on serial wiring the steppers, they now are parallel, the vicious way!.. would love to change my parallel build but hesitating to cut the wires!!

    or is there some truth in the wiring from the shapeoko forum?
    Currently the drivers are getting too hot, nearly to the thermal shutdown.. (and sometimes I hit and do not dare long runs now!

    #18159

    Ryan
    Keymaster

    First post I have a picture of a modified wiring harness.

    #18161

    ualdayan
    Participant

    I’m doing 30mm/s now, was thinking of either going to 24V or going with a NEMA23 for the Z with the goal of 60mm/s hopefully. I have a lead screw (with antibacklash splitnut) so that’s why I’m able to do 30mm/s now.

    Or who knows – I bought an Arduino Due for $13 along with 30lbs of 4043D plastic (Ahh, Filastruder – nothing like $2 a lb plastic to print with) the other day during one of Ebay’s ‘spend $50 get $10 off’ so I’m going to be trying out TinyG G2Core’s firmware which supposedly can do some amazing things with acceleration because it bases everything on constant jerk instead of constant acceleration. Maybe that better acceleration management alone can get me up to at least 40mm/s if I do 30mm/s already without problems.

    #18162

    Ryan
    Keymaster

    You are firmware and CPU limited to 8.7mm/s you can’t go above that unless you have edited your firmware and changed your step rate.

    I have been trying out smoothieware, I don’t notice a difference.

    #18178

    ualdayan
    Participant

    Yeah, I mentioned the CPU limit here: https://www.v1engineering.com/forum/topic/z-lock/ (wow, hard to believe it’s been a whole year already since I got my kit)

    I went with a 4 start 2mm pitch lead screw though for my Z so it’s only 200 steps per mm at 1/32 stepping.

    I looked at Smoothieware/board too, it seemed like it would be a lot easier to switch to, but apparently even with the increased CPU power the actual firmware for Smoothieware is the same constant acceleration as Marlin/Repetier so I wouldn’t see much benefit. TinyG G2Core’s acceleration is apparently much different – but comes at the cost of being harder to switch to – it uses JSON (Chilipeppr) instead of normal gcode senders like Repetier Server or Octoprint and doesn’t have any presupported ‘all in one’ board available that supports 3D printing AND CNC cutting like RAMPS. Seeing it’s acceleration management though is amazing, here are some videos of it in action:
    https://www.youtube.com/watch?v=tEXy6E8G154 (see how the ball swings left and right but then barely swings when the carriage stops moving)
    https://www.youtube.com/watch?v=Om0wTqFA-Dw (it runs at 50,000mm/min and isn’t losing steps – granted it can’t print at those speeds obviously)
    https://vine.co/v/hvYIdvQeZpK (swinging a big hammer around, and coming right up to but not touching a pair of buttons on both sides)

    #18179

    Ryan
    Keymaster

    How is that leadscrew nut holding up after a years abuse?

    Chilipeppr does seem real nice. I’m sure it would take some fine tuning but once it was dialed in, cream of the crop I’m sure.

    #18187

    ualdayan
    Participant

    The lead screw has been doing good – the original nut developed some wear and I ended up buying a split nut antibacklash one instead ( http://www.ebay.com/itm/322207703521 ) and as it wears the spring takes up the slack. The top part is kept from spinning by the legs coming down the side, so it can move up and down but not spin.

    Even with the worn nut though I wasn’t really seeing any effect on my cutting though – just that I could move the Z assembly up and down a little without turning the screw. I don’t know if it was the weight of the tool/Z assembly holding it down during actual cutting that made it not cause problems with cutting, but I just didn’t like the play being there so I got the antibacklash nut anyway.

    #19702

    WarHawk8080
    Participant

    I made some parallel to series boards on OSHPark if interested

    Don’t need to hack up your wiring anymore…plug and play

    https://oshpark.com/shared_projects/imzyj7Kj

    #19712

    Ryan
    Keymaster

    Those boards look pretty slick, those connectors aren’t the perfect fit for the plugs we currently use but if you take off the white shield after soldering they fit perfect.

    I have some new wiring harnesses on the way, wired in series, but I might try yours out as well and see what fits better. Thanks for that!

    #20237

    Ryan
    Keymaster

    Guess What just came in the mail…..!

    #25162

    GEOFFROY
    Participant

    Hello, I am planning to build one MP CNC, and stepper motor/driver is my first concern.
    I read carefully all thread, but I don’t find anything about the link between stepper motor, driver and power supply, but if you understand that it is easy to choose what you want to do.

    The driver provides electrical power to the coils of stepper motor in right order and allows limiting the maximum current. The wiring (wires and connectors) and the motor have impendency that limit the rising time of current. The rising curve is linked to the tension of power supply.

    Due to that if you power only one stepper with a driver, the rising time of current is nearly linked to the difference of voltage between the power supply and stepper motor voltage per phase.
    If the motor run slowly, the rising time can be forgotten, but when you increase the speed this is not the case and this rising time result is to lower the average current of each pulse and decrease the torque of motor stepper.
    If you add another stepper motor in serial (without any other change) the difference of voltage between the power supply and the 2 stepper motors voltage per phase decease also (can be negative according the type of stepper motor) and due to that the rising time is bigger and the average current lower and torque lower.

    If you put in parallel, the first impact is to nearly double the current, but there are other things to take into account: wiring, and motors. Even you take care to have the same wiring for the 2 motors (length/ section / connectors…) each motor are slightly different , and the current in each of 2 stepper motor are also slightly different, and this is difficult to manage.
    Due to that “industrial” solution is to use one stepper for one stepper motor and to measure the real current in both motor (with scope) and adjust the 2 stepper motor at the same current.

    In practice:
    – You can put stepper motors in parallel if they are from the same batch and the sum of currents is under the maximum current of the driver, you must take care also of the length of the wire to each motor and keep them equal.
    – You can put stepper motors in series, if the double of tension per phase of each motor is under of tension of power supply minus 5V. For 12V power supply tension per phase of stepper motor must stay below 3,5V.
    – After according your skill you can put one driver per stepper motor, and modify the software or the hardware of the RAMPS. And if you want to add an extruder ( or more) you can make or buy an driver expander like this one: http://reprap.org/wiki/Stepper_Expander
    See stepper motors data sheet: http://www.osmtec.com/downloads/stepper_motor/stepper_motor.pdf
    Excuse my bad English, I am French.

    #26076

    ivke
    Participant

    Hello
    Is this the same wiring wiring 1
    and this wiring 2
    Thank You

    #26085

    GEOFFROY
    Participant

    #26086

    GEOFFROY
    Participant

    They are not excattly the same:
    – the 2 are serial wiring
    but
    – in drawing the 2 motors are in reverse rotation direction
    – on the wiring they are in same in same rotation direction

    #26089

    Ryan
    Keymaster

    They are exactly the same. The drawing is a representation showing the coils separately. The picture is the real world version of the drawing, for those that have my parallel wiring harness it is easy to copy that and make it series.

    #27450

    UnLtdSoul
    Participant

    Hey,
    I just had to put in my 2 cents here. I’ve had my unit up and running about 5 months, and doing daily runs on it about 3 months. Ted, thank you for a great design. My unit is 3×4 foot and is highly Modified. I would post my mods, but, I just don’t have time to.
    First mod is I put mid way supports on the outer rails on both X and Y axis. But, I wasn’t happy that the gantry had too much flex. So, I made an Extreme mod of the Gantry. First, I put 2 pipes on both the X and Y (only on the gantry, not on the outer fixed rails). I kept the center rollers on the outer rails the same – only one pipe is attached to the center rollers, the 2nd pipe is screwed to the main pipe. On one axis I put the 2nd pipe on the top, on the other axis i put the 2nd pipe on the bottom. The pipes are screwed to each other, and i used 4 screws – one at each end and several along the length. The two pipes – one on top of the other, has a lot more stiffness. But, due to there now being 4 pipes going to the Z gantry, 2 on the X and 2 on the Y, i had to totally redesign the Z axis and gantry. I went with HEAVY Duty – using 3/4″ thick oak plywood plates, one on top, one on the bottom. And redesigned the rollers to account for the 2 pipes on each axis. It is heavy and bulky, but, it is SOLID. With the dual pipes on both X and Y I still get flex in the gantry, but, far far less. Previously I just had to put a slight pressure on the gantry and get very noticeable flexing. Now I have to exert a fair amount of force to get a slight flex. Much more solid.
    At the same time I switched to an 8mm lead screw rather than the 5/16 threaded rod. The Z is much faster – can run it at 2000 mm sec. BUT – with the lead screw there isn’t much resistance, so the weight of the motor would make the Z axis drop straight down when the stepper motor would turn off. I ruined a number of bits and have holes at my home position and all over the spoiler due to the motor just dropping straight down. Finally, my latest mod was to add a Counter Weight system to my Z axis. That adds another 6+ lbs to my gantry, but, so far i have not seen any issues with that. The counter weight added some bulk, but, it has taken all the weight off the Z axis motor. With power off I can literally spin the Z axis and just as easily move the Z up or down – freely. And, yet, it no longer drops on it’s own. (Like I said, my unit is highly mod-ed).

    At one point I had updated the outer center rollers with the newer design that has built in braces to keep it from bending. But, I had only updated the Y axis (because I had damaged it at one point). The pressure of the timing belt on the older center roller caused the motor plate to bend after a while. But, like I said, i had only replaced them on the Y axis. After running the system for several months – all of a sudden it started messing up on my v-carve work. I finally found the X axis was skipping. It was binding. With power off I could freely push the Y axis, but the X was binding. In the end it wound up being the older Center roller, the X axis that i hadn’t updated – it was leaning over. That slight tilt was causing the belts to bind up enough to make it skip when it was carving. I then printed the newer center rollers and that took care of the binding.

    But, before I figured out what was causing the X to skip, I looked into what I could do to up the current to my motors and was looking at wiring separate drivers for the X Y motors rather then running them in parallel. I am using 84oz torque NEMA 17’s that max at 2 amps. But, the drivers max at just over 2 amps as well. I had set the trim pot on the X Y drivers at 1.1 volts, so they would output 2.2 amps. While researching what i was going to do, i found this thread. Good thing I did.

    It was so easy to change over to series on my design where the wires come out of the cable chain on the center roller, i just disconnected the wiring for both motors there and easily changed it from parallel to serial. WOW – finally, the 84 oz motors are running at their full potential. WOW – what a freakin difference. I can now plow through oak with ease. Hitting rock hard knots on pine is no longer an issue. I am now 3D carving pine, maple, and even red oak skipping the 1/4″ end mill rough cut, and just directly carve with the 1/8 ball nose bit (at 45 degree angle) into maple at over 4000 mm/sec with no problem. But, just to be more sane, i try not to run it over 2500 mm/sec. I tried 3D carving in solid bamboo and skipping the rough cut, and it was cutting it very clean running at 4000+ mm/sec, but then the bit broke, so I think i was pushing it a bit too much. Now the issue is not the lack of power or solidity. I just have to learn the limitations of the bits and material. (I tried V carving in laminate flooring which has HDF backing, and that stuff is like milling steel. My carbide tip bits go dull before i can complete 1 piece. And, that was running them at only 900 mm/sec. I will try slower, but, i am thinking HDF may be too hard – not worth it).

    As far as the main difference between parallel and serial – it makes sense. When the motors are in parallel the current is split between the 2 motors. For illustration purposes, say the motors are 2 amps at 4 volts on each coil. But, our drivers can only output 2 amps total. So, when it is wired in parallel the driver outputs 4 volts so that the current is 2 amps. But, those 2 amps are divided in parallel, 1 amp to each motor. Both motors see the 4 volts, but, each motor only gets 1/2 the current the driver is putting out.

    In series, the driver is still set to output 2 amps, but now the load is higher resistance, so the drivers have to output more volts. Even though the driver is out putting double the volts, in series each motor only sees 1/2 the voltage, but, the full amperage passes through each motor. So, the full 2 amps now passes through each motor. The driver is now out putting 8 volts at 2 amps, and each motor only sees 4 volts each in series, but, at the full 2 amps. So, we just doubled the power to the motors, as you said, for Free. I haven’t turned my drivers down yet. I should, because they are still set to 1.1 volt, or the max of 2.2 amps. But, the motors are rated for it, it’s just the drivers are running on the hot side. I should be able to turn down to .8 with no major loss – but right now, i am very pleased with the added power. I noticed the X and Y run smoother.

    I have added IR leds to my DW660, so I can make a Tach and view the RPM. Took off the hand tightner on the DW660 and printed a small holder for the IR leds, and painted a mark on the motor shaft just above the collet. I still yet have to program a separate arduino to view the RPM, but the IR sensors are out putting signals. The next mod is to read the RPMs and create a PWM controlled AC speed control to automatically control the RPM on the DW660 spindle. I got it designed, just need some spare time to build it.

    #27451

    P3DCNC
    Participant

    You should post some pics of your mods..

    #27724

    UnLtdSoul
    Participant

    sorry it took a few days – but, here are a few photos highlighting the dual pipes and the modified Zaxis-Gantry. Also the counter-weight system. Using a DW660 so far so good, if i were to do it over again, probably go for 1.5 HP or larger router. But, i am thinking once i get a decent tach synced pwm controlled ac speed control hooked up on it, it should work even better by maintaining rpm with or without load.

    I put descriptions on image file names

    #27730

    UnLtdSoul
    Participant
    #27736

    Ryan
    Keymaster

    That is pretty crazy.

    I like the dual rails.

    #31672

    Joshua Dukate
    Participant

    Does anyone have an accurate wiring diagram for series wiring for the mpcnc? Each time I wore them they either run in the same direction or not at all any help will be awesome thank you.

    #31695

    Ryan
    Keymaster

    Accurate meaning?

    I have a wiring harness for sale so you don’t have to worry about it, the picture a few posts up shows exactly how to modify a parallel cable, you should be able to translate that to any stepper color you have.

    If you show a picture of you steppers I could give you a color chart.

    It breaks down to two loops running from your control board through each coil then back to the control board. the coils are usually in order on the plugs. if you look at the hand drawing and substitute you plugs for the 12ab wiring it like that should work.

    #32026

    Mbef
    Participant

    Interesting thread!

    I was having stepper skipping problems on both X and Y, using our cheap little Wantai 42BYGHW811 steppers, which I had lying around from when we built a Cyclone. Just before I hit BUY on some stronger steppers, we came across this thread, so I tried out series connection.

    I haven’t tried out actual milling yet in the new configuration, but just with a strength test, the results are really impressive. I’m using a calibration device that we built for helping with the initial commissioning of the MPCNC, which is held in place with elastic bands.

    With steppers in parallel, the bands were stretched by 65mm, but in series, they were stretched by 186mm. Assuming the bands stretched fairly linearly (), we’re seeing a factor of three improvement! And – it feels like it too; before, you could shove the gantry out of position with a good push, but now you have to really lean into it, and the gantry crossbars are bending visibly before it skips. I started to measure actually how much force this was, but I’m struggling to measure it without dismantling the bathroom scales. It’s a lot.

    Gotta finish the wiring, then we can test it out for real. More to come…

    #32032

    Ryan
    Keymaster

    Awesome, I do feel it is the best option.

Viewing 27 posts - 31 through 57 (of 57 total)

You must be logged in to reply to this topic.