ClearPath VS DMM Technology Servo's

[@TeknicInc]

A few weeks ago, I came across this video. As an engineer for Teknic, I felt it was important to correct the errors (see below) so that people are able to make the most informed decisions possible on what components are best for their applications. After my initial post got rejected (because it had hyperlinks to various points in the video), I called Sam to discuss the errors I found. Since Sam's goal with each video is to accurately inform his viewers, he happily corrected some of his mistakes. There are still a few that haven't been corrected, so Sam agreed to let me post an updated comment. I’ve limited my post to correcting factual errors only. My goal is to make sure that viewers have accurate technical and pricing information to choose the best product for their applications. There are a few points that Sam and I disagree on, and we've debated those, but I would acknowledge that some of these points are somewhat subjective. (For example, Sam stated in the video that a system with a separate drive, motor and encoder is easier to troubleshoot than an integrated system. My experience is that our customers who have bought _both_ types of systems from us find the integrated systems easier. When Sam and I discussed this, he mentioned that he had changed his mind on this, but added that if an integrated motor failed, it would cost more to replace. I agree, but would add that the length of the warranty should also be factored in (3 years for ClearPath).) I apologize in advance for the length of this post; I've done my best to format the response in an organized, easy-to-read manner (unfortunately, I had to leave out timestamps because KZbin automatically converts them to hyperlinks). I hope this information will clear up any misunderstood subject matter. Best regards, Abe A. *Specification and Technical Corrections:* •The rated torque difference stated in the first table (between DYN2 drive and CPM-2311S) is less than stated. It is overstated by 24 percentage points. •In row 6 of this table (Encoder Resolution), all the compared ClearPath motors should be listed as having an encoder resolution of 12,800 counts/rev (not 800 or 6,400). The _input resolution_ is 800 steps/rev for the RLN series and 6,400 step/rev for the ELN series. So, for example, with 5mm ball screws and 800 input steps/rev, the commandable input resolution is about 0.2 thousandths of an inch per step (~6 microns per step); the encoder resolution is 16 times better. •No fan is _needed_ to operate these ClearPath motors within their torque/speed ratings over their full specified temperature range. At full load, they run comfortably within the ratings of the class H insulation and components. (As an indication of Teknic’s conservative ratings, the ClearPath warranty is 3 years compared to the DMM 1-year warranty.) A fan can be used to boost RMS torque in cases where going up a motor size is not desirable. •In the "Key Considerations of DMM DYN4" table, the DMM _native_ encoder resolution as shown is incorrect; the figure stated is the _interpolated_ resolution. The DMM motor uses a magnetic encoder; the native resolution of magnetic encoders is generally quite low. According to the DMM website, the DMM encoder uses a 2-pole magnet. So it natively generates only one sinewave per revolution. The actual encoder counts are all generated by interpolation, i.e., _estimating_ the motor position based on the amplitude of the sinewave from the hall-effect sensors-in the face of errors such as run-out, noise, offset, short- and long-term drift, sensor hysteresis, non-linearities, temperature effects, etc. These errors can be somewhat compensated for, but as stated on DMM’s website, the accuracy is only 0.1 degree (1 part in 3,600). Because of the extreme interpolation, the accuracy is nowhere near the resolution figure (the accuracy is 18x worse than the resolution). •The video states that having finer encoder resolution increases the response speed of a servo. For the DMM system, however, this is not correct because the encoder is a serial encoder. The servo gets position updates at a fixed clock rate (every 100 microseconds, per the DMM website). This means that at 3,000 RPM, you would have a positional change of 327 counts before the servo knows anything has changed. This is an unavoidable drawback of a serial encoder. •Increased encoder resolution does _not_ let you attain higher speed or acceleration as stated. If anything, as you increase resolution, your attainable speed can go down due to frequency response limitations of the circuitry. This is stated correctly on the DMM website where they footnote that the encoder resolution must be _decreased_ to attain its highest speeds. •The comparison stating that the DMM RS-485 encoder signaling is more noise immune than the ClearPath step and direction signaling is incorrect. RS-485 uses differential signaling which has decent noise immunity, but the ClearPath step and direction inputs have full optical (galvanic) isolation. Optically isolated systems are virtually immune from ground loop induced noise and common mode noise, and are significantly more noise immune than differential systems. In addition, unlike differential circuits, optical isolation provides protection from high-voltage surges and spikes. These reasons are why companies like Texas Instruments sell both a non-isolated RS-485 transceiver chip (e.g., SN65HVD1473) as well as a more expensive RS-485 transceiver chip with built-in isolation for increased noise immunity (e.g., ISO35). *Pricing, Part Number, and Dimensions Corrections:* •Wherever there is a Teknic part number listed beginning with "CPM-SCSK-xxxxx-xxxx", the part number should be changed to "CPM-SDSK-xxxxx-xxx". The pricing in the video for each of these part numbers is off by any amount between $44 and $76. Sam has listed these corrections in the description of this video as well as his dropbox spreadsheet. •Total System Costs: To make an accurate cost comparison, it's necessary to include the cost of the required DMM encoder cable and motor cable. These parts are not needed for ClearPath, but would be additional cost for the DMM setup. oThe cost of the 3m motor cable and separate encoder cable (which are both required) is $46 total oDMM's manual (pg 20) recommends adding an extra filter capacitor if using the DYN-2 dc input drive (you'd also need a mounting bracket and wiring) oIf you have significant regen (especially common for vertical axes), you’ll have to add a circuit to detect and handle over-voltage and add an integral regen resistor (see page 21 of DYN-2 manual). Automation Direct has external regen clamps for $89. oAs a side note, related to the reliability discussion, all of these items and added interconnects will reduce system reliability. •Row 7 comparison of motor lengths has errors: oColumn 5: ClearPath 3421S motor length is 99 mm, not 87 mm oColumn 7: ClearPath 3432S motor length is 118mm, not 87mm oAs an aside, DMM is a separate drive/motor - when you include in the requirements for cable access (per the DMM manual), you actually need more space (up to 188% more space).

[@sheets8277]

I was gonna go with the clear path motors but now I’ve seen this video what would be my advantages over the dmm motors besides the cheaper price so confused about which way to go with servos any help is appreciated

[@mathieulemieux8287]

Such a pain to read your comment! Don’t you know that less is more!?

[@Tuffenough4u]

Man, I don't know if you're just being satirical or trolling, but if not I have to say it'a about most pro-dumbass, patronizing remark I've ever read. No one made anyone read anything and some of us actually appreciate the in depth data provided. When it comes to detailed information regarding things that we enjoy, "less" is never more.

[@madaxe79]

Aren't DMM Servos actual industrial Servos, as opposed to Clearpath being hobby level components?

[@abeamirana4892]

@@madaxe79 Hi Madaxe79, Although Teknic sells to and supports many different types of customers with various applications and goals, the main market for our products is high-volume, commercial OEM machine builders. ClearPath (and other Teknic products) are used in a wide range of commercial, OEM machine applications, ranging from medical blood analyzers, imaging machines to industrial waterjet cutting machines. We worked hard to make ClearPath price competitive and relatively easy to use (compared to many other servo systems out there), and to our pleasant surprise, these servos were also quickly adopted in the hobbyist CNC marketplace as well. That said, Teknic's products, business structure, and manufacturing/testing practices are specifically designed to support OEM companies. If you’re interested in learning more, you can find additional information on our web site in the "About Us" section. If you have any additional questions, please feel free to contact us and speak directly with a factory trained engineer (or you can leave us a message on our web site by clicking on the menu pulldown command labeled "contact". - AbeA, Teknic Servo Systems Engineer

[@Runner50783]

I can't speak anything but good about the Clearpath servos. They are on the expensive side but I think this motors are well worth it. The torque curve is "tuned" to resemble stepper motors on the (SK) Stepper Killer series. I actually believe that it is kind of a software limit as the motor construction, weight, etc is the same between the SK and HP (High power) Clearpath servos. Any how, As long as you size your motors accordingly, these motors are fantastic, tuning is easy and secure enough for a beginner, construction is top notch. I currently use this motors on my laser cutter and I absolutely love these things.

[@CNC4XR7]

No Doubt ClearPath servo's are Far better than steppers However the rpm on the higher torque motors may limit you in some applications. the PPR of only 800 or 6400 will limit you as well Vs other servos on the market.

[@Runner50783]

I think it all comes down to real world use. I was seen a demo of the DMM servos kzbin.info/www/bejne/nKquc36LoKt9jKM and to me it was not supper impressive. As I do laser engraving, I rutinely do motion commands that exceed the 15Hz on my Clearpath SK a task in which any deviation from the command will quickly show up on the material, one thing that the Clearpaths have that DMM does not is a feature called RAS (regressive auto spline) which gives you better response than S curve motion profiles and far smoother than T motion profiles, that allows the Clearpaths to achieve quicker response with near perfect rendition of the commanded motion. Regarding the encoder resolution, the technology that Teknic uses does not require anything higher than the 12800 CPR optical encoder they use. I was comparing the reperformance I get with an other user who's got a Panasonic Minas A5 instead with a 130K CPR enconder, the Clearpaths perform far better and cost a lot less as well.

[@CNC4XR7]

I agree It all comes down to ones particular need and application. for me I felt that the Dmm servo and the DYN4 drives offered me a Better fit.

[@Z-add]

In your reply you didn't talk about the torque curve. Clearpath torque is clearly limited to a small rpm range vs dmm. Do you have anything to add to this difference. I would gladly pay the higher dmm system cost if it meant i have more usable torque at higher rpms.

[@MrWebsie]

The let down on a clearpath is the lack of encoder feedback, there is a feedback but it is only a low step count, like 16 per revolution I think. It can be selected in software. But for me and accurate current step count is an absolute must.

[@yelims20]

I bought clearpath and am happy I did. They set up easily and just plain work.

[@CNC4XR7]

Great to hear!

[@TheFeralEngineer]

I'm piecing together a DMM system right now for an SX3 mill. I'd like to argue the price point of comparing these two units. DMM recommends 2 emi filters per dyn4, they also recommend a breaker and a contractor per drive and a line reactor for 220v single phase. I'm doing a 4 servo system, two 400 watt nema 23, a 750 watt n34 and a 1kw n34 spindle motor. That all clocked in, with cables, at 2100ish dollars, not including the 700ish dollars I'm into for secondary pieces, including cables to connect from the drive to bob that I'm just now realizing I'm going to need. From there, you need an enclosure. For this system, because of all of the other components, you're looking at a 24x30x8 minimum sized box 20x20x8 if you if you use a 3 drive setup, 10 inches deep is actually recommended. Price those on the internet, they aren't cheap. I found one for 110 before shipping, but it took hours. Now, consider all of the cabinet wiring, terminal blocks, din rails, a separate 24v supply for powering your breakout board and DC accessories... Needless to say, I'm not overly thrilled by the amount of extra stuff I didn't know I needed until I bought these things... I did like the extra power, but it's coming at a steeper price than I thought it would. My next build, I may consider clear path instead, unless I'm selling the thing.

[@CNC4XR7]

Thank you for the comment. Unfortunately I have yet to wire these up. They are for my linear rail project. They do require more space as you said. I think I mentioned this in the video.The cost of bigger enclosures is higher unless you can find used ones as I have. I need to check with Dmm on the separate breaker and contactor I think they can all be on one at least mine will. I would email them on that and on the emi filters. I purchased 10 inch deep cabinets for my builds.

[@TheFeralEngineer]

@@CNC4XR7 buddy of mine has a 24x30x13 that I'm gonna pick up for 40 bucks that has a bunch of terminal blocks, a ground bus bar and some others goodies inside of it. It's uglier than sin, but it'll get the job done. I could always just toss a coat of paint on the thing if it's the end of the world. I just wasn't ready for almost a grand in secondary parts. I would've needed at least 2 contractors even if I got larger ones because my 750 watt Z and 1kw spindle motor are going to run off of 220 and the 400 watt motors are going to run on 110. The small motors can't handle 220, it'll fry them. If there's one thing I can say about this build, it's definitely teaching me a lot. I'm an AE for a global builder, but I never spent much time inside electrical cabinets unless I'm resetting a breaker or connecting a customer's company LAN, so this has been kind of therapeutic... The sticker shock is almost vomit inducing, though 🤢

[@phildalgleish2381]

Thank You for taking the time to do the comparison!.. Its a really great help in deciding on what to choose. Have a Great Christmas!

[@CNC4XR7]

Thanks Phil, Yes it took awhile to put this together But I wanted to get the info out Dmm servos are not as well known so there's not a lot of info on them and some of the Phrases and catchy terms clearpath uses ( jerk Limiting, S curve, RAS) are pretty much standard stuff.

[@ivanabunjevcevic9364]

Very useful video. Thank you. Clearpath seems to be new craze these days, but dmm has some superior specs. I think Clearpath torque curves are like that because they are actually constructed as a stepper motor with encoder and servo control. Loss of torque at higher rpms is really not a big issue. Those rpms are for rapids anyways. One thing that I like about Clearpath is that their controller controls third derivative of path (jerk) so that acceleration is changed in s-curve manner. On top of that they have an auto tuning routine. Those two things still make them look attractive. You seem proficient in cnc stuff, so let me ask you, is there an upside of using servo motors with mach3? Mach sends them pulse&direction signals and velocity profiles and everything is controlled by mach just as it is with steppers. The loop is not closed with mach3, so it has no feedback what the following error is, and it can not compensate for multi-axis interpolations? It can only halt if error is out of bounds, but that is not very helpful. I seem to me that it is easier just to pick overdimensioned steppers, get some earplugs and start routing.

[@CNC4XR7]

Thank you for your comment! Yes ClearPath are all the craze! Thankfully I had some time to do nothing but look into other servo's while stuck away from home. DMM Technology employees their own S curve generators The DYN servo drive’s built in S-Curve generator is able to update the target position instantaneously regardless of whether the current command position has completed or not. As soon as a new command position is received, the servo drive immediately updates the servomotor target to the newest position. This function is applicable to both relative (incremental) and absolute positioning for all linear, or arc path profiles. DMM also has Dynamic Target Position Update ( DTPU ) feature which does many things The DTPU algorithm also applies a curved acceleration to maintain smooth motion. At each S-Curve transition point, the acceleration/deceleration is curved at the edges so speed is smoothly changed. This decreases motor vibration. ( ClearPath's Jerk feature) All of this information is in the Dmm Product manual. www.dmm-tech.com/Files/CTADYN4FT-1214E0_P1.pdf these features have been around for years ClearPath's cleaver advertising and catchy names have brought them to our attention. With Mach3 you are correct in that there is no way to monitor the encoder feedback so a closed loop will only report if it's Out of Bounds. Unlike ClearPath servo's Dmm DYN4 drives will output a encoder signal and allow you to run a Analog signal instead of step and direction for real time absolute positioning. another feature worth mentioning for retro fit of older machines. Mach 3 will not benefit from this however Linux cnc can take advantage of this feature.

[@SmallShopConcepts]

Ivana, The loop is closed on both the clearpath and the DMM servos to the drive when in step direction mode. This will ensure that the commanded pulses to the drive are executed by the motor via the encoder loop to the drive. As to comparing steppers and servos there are advantages and disadvantages to both although the majority of those are cost based. In this day and age i would NEVER run an open loop stepper with closed loop steppers available at moderately higher costs. But even closed loop steppers will lack the smoothness of true servo motion as well as speed (although they are faster than open loop counterparts). Additionally, a closed loop stepper generally uses a very low resolution encoder and will not have the granularity of a modern day absolute encoder servo set. As with anything the end result of your machine performance will be up to you and what you require. But for metal processing, tolerances are typically much more critical then with wood and having a motion system that can deliver becomes one of the most critical components to the build. You mentioned the clearpath having an S-curve or sinusoidal acceleration curve as an advantage, however ALL servos typically have this in some varying degree and most drives offer the ability to adjust this curve to suit, so that advantage wouyld be compared to steppers alone, and it should be noted that even the latest set of closed loop steppers i received have drive parameters to use an S curve acceleration built into the drive. However with a step and direction type control system all of the motion planning is based of trapezoidal acceleration curves (linear acceleration). With this in mind I'm not certain I would consider this a clearpath pro as without the ability to close the loop of the clearpath servo drive to the motion controller it becomes a negative for introducing slight error to the planned trajectory. It is a benefit if you are using a simple point to point type piece of equipment which is not being multi axis coordinated and you require it to have less inertial load transfer from starting and stopping to minimize shake or vibration. it is also a benefit if using a motion controller that is running in analog or step direction with feedback mode and a closed loop (not mach3 as mentioned) however at that point it isn't a benefit to the clearpath set anymore as of this writing I am not aware that the clearpath offer an output for the encoder to the motion controller. This would mean that the only benefit would be realized in more simple equipment as previously mentioned. Most modern servo drives are now equipped with a self tuning strategy as well as a manual tuning strategy which are fairly simple to implement even in lower cost units which the strategy is initiated at the drive user interface. It does of course require a brief read of the manual but at this level of equipment that should be common practice even with the clearpath models. Clearpath certainly offer some advantages in packaging, as well as when compared to an open loop stepper, however I would put them more closely competetive to a modern closed loop stepper than a full on industrial servo and drive set. they will be quieter than steppers and smoother but the rest of the pros and cons are debatable especially when a closed loop stepper costs 1/2 the price of the clearpath. these are my thoughts on the subject and admittedly I am not running either brand as the equipment i use servos with currently require significantly larger units than either manufacturer offer, however they are industrial servos and drives and are more closely related to the DMM units than the clearpath. Chris

[@TeknicInc]

Hi Ivana I’m an engineer with Teknic. I came across your questions here, and I wanted to reach out and answer those that I could. I want to start out by making a clarification. ClearPath motors are not designed or manufactured like stepper motors. They are truly three-phase, permanent magnet, AC sine-wave commutated, brushless servo motors with an integral encoder used for commutation and positioning. They sometimes are mistaken for stepper motors because they accept step & direction commands, which are common in many stepper systems. The ClearPath SDSK series (Step & Direction Stepper Killer) can add to the confusion because they were designed as stepper motor replacements. This means that the SDSK torque-speed curves can look similar to stepper motors (especially the ClearPath “S” winding) but with 2-3x the power when compared to a similar sized stepper. Teknic’s “RAS” function (Regressive Auto Spline) is significantly more advanced than a simple S-curve algorithm which only limits the instantaneous rate of change of acceleration. The distinction here is that S-curve smoothing will still leave abrupt changes in the torque (which can result in overshooting and suboptimal performance). The RAS algorithm, on the other hand, is a 4th order polynomial spline that limits the rate of change, of the rate of change, of acceleration. It should also be noted that the RAS does not negatively affect multi-axis coordination. Just make sure that you use the same RAS setting for all axes. If you’re interested, a Teknic engineer created a video detailing the differences between the RAS and other forms of jerk limiting. It is an interesting (and relevant) topic, and you can find that video on Teknic’s KZbin channel. Feel free to contact us directly if you have any additional questions. Thanks, Jon K.

[@TeknicInc]

Hi Chris, We receive a fair number of questions about using the ClearPath motors (and their RAS feature) in coordinated multi-axis applications - so I wanted to reach out and address some of your concerns. Let me start by saying that you are absolutely correct; morphing the velocity vs. time trajectory can have a non-negligible effect on your actual position at any point in time. Similarly, any deformation of the mechanics during motion can have a similar effect. In an ideal world, the mechanics on any machine would be infinitely stiff and capable of responding to high-frequency, high-energy changes in the velocity vs. time profile. In this case, command smoothing wouldn’t be necessary. Any good CNC controller will create the appropriate coordinated profiles for the axes to follow. However, these controllers frequently do not take into account how the jerk (or jerk-derivative) component of a profile will affect the mechanics and the cutting of the part. Many times, this high-frequency energy generated in the command not only negatively affects the cut quality, but can also be seen and heard it in the mechanics as they vibrate and ring. To get around this, many machine operators will have to lower their accelerations and feed rates to avoid exciting and/or flexing the mechanics. This approach is more of a band-aid for the problem rather than addressing the root cause of the problem. The RAS is a user-selectable, 4th order polynomial spline that limits the rate of change, of the rate of change, of acceleration. This algorithm is more sophisticated than S-curving which only limits the instantaneous rate of change of acceleration. S-curve smoothing still leaves discontinuous changes in the commanded torque which will cause errors in the actual machine trajectory (and thus in the finished part). Because the RAS limits the next higher order derivative compared to S-curve or jerk limiting, it removes the instantaneous changes to the torque command. This will generally improve the machine’s ability to follow the command accurately, and if the RAS is set appropriately, the trajectory changes will not be noticeable in the finished part. There is an optimal RAS setting for each type of application. The ideal setting removes enough of the high-frequency energy to improve the cut quality but without negatively altering the output of the part. Here are some typical RAS settings used by our industrial OEM customers for different types of CNC machines: Laser CNC 9ms-16ms Wood router 16ms-24ms Plasma cutter 24ms-44ms A Teknic engineer created a video detailing the differences between the RAS and other forms of jerk limiting (including S-curving). It is an interesting (and relevant) topic, and you can find that video on Teknic's KZbin channel. Thanks, Jon K.

[@shaneclendennen1929]

Sam, I'm thinking about going with servos over the stepper motors I have on my small mills. I saw where you switched from DMM to clear path. Why was that?

[@CNC4XR7]

Hi Shane I didn't really switch I have Both The Dmm Servos will be on my Linear rail conversion. When I get back on it. but the power modules are to big to fit in my existing machine power cabinet so for that machine i went with ClearPath.

[@SmallShopConcepts]

very good information Sam! well documented and very useful in decision making process for newcomers to servos! I may in the future convert from my closed loop steppers to the DMM units! Chris

[@frijoli9579]

Very well done. My DMM servos and drives will be here Wednesday for my Lathe conversion. I decided to go servo for the spindle too. I will use this as a C axis as well!

[@CNC4XR7]

That's Awesome! I plan on converting My lathe as well also with a servo spindle. what size spindle and Axis servo's did you go with? With a spindle servo you will have the ability to do rigid Tapping.

[@SmallShopConcepts]

I think that as long as you use a spindle encoder or feedback pulse to the control unit that is granular enough to coordinate dual axis sync'd motion you can achieve rigid tapping, no need for a specific servo spindle motor, how ever the benefit would be the ability to do precision indexing if you had additional axis' on the table/bed for running a rotary tool.

[@kentvandervelden]

Another advantage for the Clearpaths is the sophisticated tuning and diagnostics software. Also, the software control version of the Clearpaths (the SC series?) has a unique advantage of the SDK Teknic supplies, which can be used to readout motor status as well as perform control. Both of these advantages may only be required in a small percentage of applications. But, disadvantage wise, the cost of the Clearpath servos go up so fast once options start being added, and they may just be software enables. Maybe the reduced torque curve of the Clearpath is because of challenges of dissipating heat from the integrated drive? Also, the higher end speed of the DYN require 220VAC? If limited to 120V the top end speed of the two would be more similar... I think... though not saying anything about the torque. Fascinating comparison Sam.

[@CNC4XR7]

Thank you, Well I did obtain higher than rated speed with 110v 4K rpm but this was on the bench. The 220v just made wiring simpler no need for a DC power supply and another part that may fail.

[@kentvandervelden]

Same experience here on the higher speed at 120V, but over that I see a lost phase error. Higher the voltage, the quicker those fields can be energized, and that may be the biggest advantage of the DMM AC drive over the 75VDC unit. Cool stuff.

[@TeknicInc]

Hi Kent, I’m an engineer with Teknic; I came across your comments here and wanted to reach out to answer your questions. To start, yes, the ClearPath SC is indeed the Software Controlled version that uses Teknic’s C++ (and C#) SDK. There are numerous versions of that product, but you’re absolutely correct; the full featured model (and/or the highest encoder resolution) is not a requirement for every application. It was for that very reason that Teknic, instead of creating a single product with all of the bells and whistles, created multiple versions of that product for applications where not all of those features are required. Similar to how a car manufacturer has multiple trim package options for a single make and model, this approach lets our customers pick and choose the options that make the most sense for their application, without being forced into the highest trim option. As for your question about reduced torque - it does indeed have to do with heat, but also with the bus voltage. As you mentioned, the ClearPath NEMA 23/34 models were designed to accept up to a nominal 75VDC input. The DMM products are taking in the 120/220VAC signals and internally converting them to a higher bus voltage (~150-300VDC). Teknic actually has stand-alone drives capable of AC input (essentially acting like an integrated drive/power supply), but as you thought, the components required to address that higher power are larger and require increased heat dissipation. The increased component size leads to a larger footprint product with increased costs, which is something Teknic felt didn’t make sense for the NEMA 23 and 34 motors. Speaking to this point, Teknic recently released a line of NEMA 56/143 motors (which are physically much larger motors than the NEMA 23/34 family), that actually take in the 110/220VAC (single or three phase) power. It was this larger size that allowed us to embed the higher powered AC-DC converters in the motor while still keeping the drive electronics cool (and output close to 8.5hp). I hope this information helps clarify your questions. Feel free to contact us directly if you have additional questions. Thanks, Aaron B.

[@TeknicInc]

In addition to responding to Kent's questions, I wanted to clarify one point about the AC and DC drives. The discussion regarding the pros and cons of AC and DC input systems is commonly held at Teknic because we manufacture both AC and DC input ClearPath servo motors, along with stand-alone AC and DC input servo drives. While there are positives and negatives to both types of systems (wiring included), generally speaking, the component count in an AC-input system is at best similar to a DC-input system. In the case of a multi-axis machine, the component count is usually far greater. Regardless of the type of input voltage, these servo systems all internally use DC voltage to spin or commutate the motor. This means that an AC to DC conversion has to occur. The same components needed to make this conversion (e.g. transformers, bridge rectifiers, capacitors, etc.) in the DC input system are going to be located somewhere else in the AC input system. In a single axis system, the difference is whether these components are located in an external power supply, the servo motor, or servo drive. For the AC-input drives, these power supply components are built directly into the drive. For DC-input systems, a separate supply is used. So in a multi-axis machine, the most common implementation is a single DC power supply powering many (if not all) of the axes. When using the AC-input drives, however, we don’t have a single power supply; we essentially have one power supply per drive, thus increasing the total component count and possibility of a component failure. Thanks, Aaron B.

[@kentvandervelden]

Thank you Aaron, fantastic information that's going directly into a saved file. Coming back here I also found Abe's comments which are amazing! Discussions like these are fantastic for learning. I purchased a DMM servo recently, they provide KZbinrs with a discount, and I'll write LinuxCNC extensions for it similar to those I've done for Leadshine closed-loop steppers. I'll be able to get all the real-time diagnostics information that I need, but comparing to the SC software interface, leaves me wanting.

[@joeygonzo]

Testing the for accuracy would interest a lot of people in the real word

[@CNC4XR7]

I will try to put something like that out however there are many variables that would throw the findings off.

[@joeygonzo]

@@CNC4XR7 make it simple . Machine some channels, holes and strips at the same speed and see which parts come out more precise

[@ClockwerkIndustries]

The majority of this video is why i decided to go with DMM servos over clearpath when i upgrade my steppers in a few weeks here. great video!

[@CNC4XR7]

Thank You! yes these are great motors.

[@joeygonzo]

how did it go ?

[@SuperYellowsubmarin]

It would be more interesting to compare the specs for a specific application. Take typically a X or Y axis for a tabletop CNC with ballscrews and dovetails. You Don't want to go near even the rated speed of a DMM. In this case, the Clearpath would have the advantage of offering a 1000rpm max, high torque, short length servo for the same power output.

[@CNC4XR7]

Of course each application will vary, Dovetails offer more resistance than Linear rails. The Clear-path servos maybe slightly shorter However the torque of the DMM servos is maintained up to 3000 rpm before any noticeable drop.

[@SuperYellowsubmarin]

Hey @@CNC4XR7 . Well that was my point, in many cases I suspect that this will be irrelevant. I think there are very few CNC applications where 3000+ rpm speed is necessary, let alone amateur CNC. I could only think of a spindle motor. That was a good review but maybe too much of a spec war.

[@CNC4XR7]

@@SuperYellowsubmarin I agree for most cnc machine's over 3000 rpm is not needed as you mentioned. however with 4 x 8 CNC routers having rapids @900 IPM is very helpful.

[@SuperYellowsubmarin]

@@CNC4XR7 Right. I'm thinking mostly desktop, knee Mills and machining centers.

[@Foomanlol]

@@SuperYellowsubmarin If you don't need the max speed you could get away with using a smaller motor with it geared down and get more torque or using finer pitch. I run 5mm pitch ballscrews on my machine and I use all 3000 rpms giving me about 600in per min rapids while having blazing acceleration improving cycle time during 3d contouring. My clear paths would error out after lots z retract drilling cycles and couldn't handle the loads at higher speeds.

[@thewoodologist8176]

Quick question, in your case with a mill I can see a tremendous benefit to running the dmm servos over clearpath sdsk, would you consider the DMM servos are worth the extra 1000$ premium over the SDSK for a cnc router, mainly for cutting sheet goods? I got quoted nearly 3k for the DMM setup with 750watt servos, plus all the additional, contactors, emi filters, fuses etc... I'd be looking at close to 3400$ vs the clearpaths I'm all in at 2300$. It would be going on an avid cnc so I won't run beyond a few hundred IPMs.

[@CNC4XR7]

I have been running ClearPath servos on my current mill for a couple of years they work fine and will work well for you on your Router.

[@Z-add]

Look at lexium mdrive by Schneider Electric. It is a stepper with features of clearpath sd servo. Closed loop with integrated drive, encoder and controller.

[@CNC4XR7]

Cost?

[@patrikj]

The linear drop in torque is a voltage limit (back-emf limiting motor current as the speed goes up.) The clearpath runs at 75V, do the DMM servos run at higher voltage since they manage to maintain peak torque up to such large speeds?

[@CNC4XR7]

Yes The DYN4 servo drive is powered by 110v-220v AC In my test video I was only using 110v. For my build it will be powered by 220v Single phase AC. Power will not be a Issue from what Ive seen and torque holds steady.

[@andy-maier]

Great video thanks! Question, I’m build a rig with 6 motors. What controller setup would you Reccomend? I’ve already written the code I’ll need to deliver all the angles to the motors, but I’m struggling to land on the controller setup

[@CNC4XR7]

Have you looked into Linux CNC

[@h2opower]

Thanks for posting this but I went and looked at Clearpath's site and the torque curves your showing don't match up with what they are showing so I will ask, "Where did you get the data from? Now for me personally I am looking to replace my AC motor with one of these so that I can have variable speed control and a lot of torque at the same time for the spindle. I'll look into replacing the stepper motors at a later time. I went to DMM's site and the 880 seems to be a good fit for my needs with a cost of $239 for the 880 and $218 for the servo drive it seems like something I might be able to do soon. Can these be run by Mach3 VFD? I think I might go with the one you are talking about as it seems to fit my needs also as with that much torque I'll be able to cut the hard stuff with ease as long as I follow the rules for feeds and speeds.

[@h2opower]

After taking a good look at the two setups I think I will go with clear path even though they are more expensive as they have a software that sets up the servo for you which takes out the guesswork. Now having the ability to go to 5k is nice and all but I only need 3k rpm. So for me I am looking at $388 for the servo and $248 for the power supply. I am also looking at weight as the AC motor it is replacing weighs 7lbs and this servo weighs the same so my z axis has the same load being applied to it. But trust me I will do my homework before I buy as there's nothing like getting something that you have no understanding in how it works to the point of not even being able to ask the right questions when you have some.

[@CNC4XR7]

Clearpath now offers the AC replacement servos I do not have experience with those I do have a friend who runs Clearpath on his PM727m and did try to use a Nema 34 the largest one for the spindle . The motor at high would not turn the gearbox with enough rpms. So he returned the servo. Good luck with your build.

[@CNC4XR7]

Mach 3 can run the DYN4 just like a vfd as this drive will except a 0-10v signal. DMM also released new software to Auto tune the motor and drive for easy setup.

[@JAMESHOPKINSIBXCNC]

Very good information as usual, hope these work out for you on your build.

[@CNC4XR7]

Thank You!

[@TheJR914]

Good presentation 👍 You have changed my mind on the next project of servos converting my Hardinge chucked lathe to CNC. My last project I used the clearpath servos 32 series on Mitsui Seiki Job Borer but couldn't use the auto setup had to do everything manual which was no fun. The auto setup i think is to violent on your drive systems. Happy Holliday 😀

[@CNC4XR7]

Yes the Auto tuning process is a violent one from the information I have gotten from friends who use them. Dmm's Drive software has recently added this Auto Tuning feature to there software, I have yet to try it but from what I have read it only travels + or _5 revolutions so I would not think it is as Violent as what you have described with ClearPath.

[@JAMESHOPKINSIBXCNC]

You will be surprised how much force can be generated in just a short distance by servos.

[@tjofab518]

Thanks again Sam for a very informative video. I was considering Clear Path, but when i called to get help with what to choose, they didn't seem to be too interested in helping me determine the best fit for my mill. I have a PM-940m. Did you find the folks at DMM more helpful if you talked to them?

[@Glamdring666]

I reached out to them regarding some very custom and unconventional use of their motors and they responded next day. They were very informative and helpful. They even offered to make customizations to their motors firmware as needed.

[@Glamdring666]

Just to play fair, I contacted clearpath firm also, and they too responded and were helpful and informative, but their motors could not do what I needed.

[@CNC4XR7]

To be Fair I found both ClearPath and DMM Tech to be equally helpful in my experience. All of my questions were answered. I'm sure you will get the same type of customer service That I received.

[@halfstep67]

What did you go with on 940? I have a 940 on order and am curious to what others are using. There isn't much conversion info on the 940.

[@russtuff]

That was very thorough. Great info!

[@CNC4XR7]

Thanks Rus , There is a lot to cover it was longer than I wanted but sometimes it's more important to present all the information.

[@thetooth]

can i get a link to the drugs you used to make this intro?

[@CNC4XR7]

LOL

[@thetooth]

@@CNC4XR7 good content tho, thanks for the overview, certainly makes getting into cnc a little bit easier

[@BMRStudio]

Hah! You clean out my mind! For CNC applications DMM is the winner ! I like the linear peak torque on almost all the range.

[@CNC4XR7]

yes the torque is very good

[@bluehandsvideo]

Holy cow Sam! Great video!!! I see a few of the guys I follow are already here....Kent V and russtuff. I need to do me some catchin' up!! lol Mike

[@CNC4XR7]

Thanks Mike! Yeah I have a few Celebrity Subs LOL!

[@louieatienza8762]

I think the big error here is comparing the Teknic SDSK models to the DMM servos. Since the SDSK "stepper killer" were intended to be "drop in" higher performance replacements for steppers, their torque curves more closely match the shape of steppers. Now if you compare the DMM servos to the SDHP "high power" servos, you can see peak torque curves that remain constant across a wider speed range, some up to the max speed of 6000rpm! Of course, you'll pay for these performance bumps, but you can compare the rated continuous power versus the max power output of the ClearPath SDHP to the DMMs and it's not too hard to see the ClearPath SDHP is vastly superior. The bigger issue to me is thinking those cheap-ass Chinese rolled ballscrews in a Chinese benchtop mill with dovetail ways will survive 3000rpm, let alone anything past that. It's really a moot point because the X axis travel is so short, you wouldn't see much of any benefit unless you had a job where you basically had to rapid traverse the entire length of the table every time. Also, the low-end torque benefit of the SDSK is that you can accelerate faster from a standstill, which in my opinion is far more important in terms of overall job speed and finish quality - and once you're accelerating, you'll need less torque to keep accelerating anyway. Related to any apparent resolution, you're still using a rolled ballscrew that inherently has more clearance (to accommodate the small irregularities in manufacturing) and a lead error, not to mention stiction in the ways of the mill, which is going to limit acceleration. Unless of course, perhaps, you're using precision ground screws. The main benefits of high speed servos comes from being able to gear them down; thus one can use a smaller servo with gear reduction as opposed to a large stepper direct-driven. In my opinion products like SDSK blur the line, provide that direct-drive torque with positional feedback in an all-in-one package. The SDHP has more of the power band one would expect from a servo. While Teknic's selection is overwhelming, their main business is as OEM; they just happen to sell components to the general public, which I find to be a great thing and I hope they continue. The DMM stuff is nice too, and great performance for the price.

[@CNC4XR7]

Thank you for your comment. the SDHP in terms of cost are far more costly than the DMM

[@louieatienza8762]

@@CNC4XR7 Yes, but a lot cheaper than similar options from Parker or Kollmorgen or Yaskawa or Panasonic... And when you add the cost of the DYN4 drive and servomotor together, it's really not that far out of reach. But as I mentioned, on an import mill with rolled ballscrews, it likely doesn't matter. The SDSK really were intended for manufactures that would use something like Parker or Kollmorgen high-voltage stepper drives and steppers. There, the SDSK has a big advantage in price and performance.

[@unionse7en]

R.P.M Revolution Per Minute (not MinuteS) , good video though!

[@CNC4XR7]

I stand corrected! well technically it's REVOLUTIONS PER MINUTE! I guess I got the "S" on the wrong word! LOL

[@dmitribovski1292]

Do you want high torque when you are rapiding into a crash ? With CNC you should only need high torque when you are cutting. If you don't have enough torque to rapid you need bigger motors.

[@CNC4XR7]

If you rapid into a crash you maybe need to rethink your Cam, Going at any speed in a direction you don't want to is not a good thing! LOL I want to have the power when ever I need it.

[@jarisipilainen3875]

1:17exactly you choosed show worse and wrong of them to compare lol. it depend on your aplication.if you want rated torgue compare at rated torgue not cost. yes you can buy chinese iphone copy. but is it good. if they come same product line then they are equal LOL

[@CNC4XR7]

It appears you did not watch the entire video! @ 3:40 I CLEARLY state that it was not fair to compare the 120W Clearpath servo with the 200 W DMM servo. I did However compare the 193W Clearpath servo with the Same 200W DMM servo @ 5:50 A closer match However the Clearpaths rated Torque is still Down by 28 oz. in. Again this was my perspective on the servo's. Thank you for the comment Be Safe, Sam

[@telelaci2]

But there is another question you forgot to ask. Who lies bigger. Those diagrams are not real, they are too straight, too horizontal, too artifical, obviously they are not the reality. What is the reality? Which firm is the honest one ? Can you trust those data ? Real life comparison and/or measurements could be much more useful I think. This is only speculation.

[@CNC4XR7]

Real Life test and Comparison would be more effective yes. However we dont have this option when we are making the decision to purchase. We can only go by the data listed on the Manufacturers website. The Data may very well be skewed in their favor. the fact that the DMM servos maintain the torque through a Higher RPM is a definite plus for me. And allows for Higher rapids.

[@telelaci2]

@@CNC4XR7 No, I disagree, if you don't have enough data, or if you know that exactly that your data is not trustable, then you don't have to make decisions. You don't need to come to conclusions, but you have to say the fact honestly - "I don't know. More data needed. Not enough information" - like a good robot AI. If you can't get better data from the manufacturer just only made up artifical diagrams with horizontal straight lines and similar nonsense, then you have to create the data with your own measurements, investigations, tests in real life. Its not so hard. There is no other way. I mean if you are amateur and hobbyist DIY guy. You know that any SERIOUS industrial manufacturer creates an exact and precise, repeatable torque - speed diagram, and many other data, technical specification. And they can not lie, they have the legal responsibility to keep those product parameters. Because that data is needed to an industrial CNC machine designer engineer. If a manufacturer does not provide that data, then the engineer can't use that product, its that simple. Its a must, its a mandatory. If you don't get precise data for a product, that manufacturer is not really serious, don't buy from it. Or if you are an amateur, not an industrial engineer, then you have to make the measurements. Here on youtube. This is a perfect place to do it.

[@CNC4XR7]

@@telelaci2 Yes I am a Home Hobbyist NOT a Engineer! You questioned the Data posted on the MFG website Not I. My intent is this video is not to sell you ClearPath or DMM servos but simply state the reasons why I choose Dmm over Clearpath. Obviously it's up to each individual as I state in the video to come to your own conclusion based on your needs. It sounds like you have a good handle on servo's and how to compare them properly. I suggest you purchase one of each conduct real world tests and Post the video! I would be happy to link your video here. Please forward the link when you have it posted. Thank you for your comments. Be Safe, Sam