H-Calculation for Delta printers with FRS's

  • I have written a post on the SeeMeCNC forums about how to calculate the H-values for the individual G30 lines in the bed.g file. I you are interested in the process or wanting to try it out use this link.
    This uses the Duet 0.8.5 hardware and the Ultibots FRS Kit.
    It may work with other FRS systems and should work with the DuetWiFi as the mechanics of the process do not depend on the hardware outside of having the Duet to use the least-squares bed calibration methods from escher3d.com

  • Zep -
    I might be just having a blonde moment, but you think it would be possible for you to post a short video of the procedure you outlined in your post linked above? I did a dual major in EE and CS in college, but for some reason or other, (maybe I'd have been better served studying English, perhaps ☺️) I am unable to follow your seemingly simple process.

    Thanks. The reason I'm very interested in this topic is due to the fact that originally, I installed a duet board on my Rostock Max v2 (modded) printer but was never quite able to get decent prints from it until I literally became frustrated through and through and simply "accidentally" pushed the unit off relatively high desk I had it propped on to its satisfying death in a bunch of pieces. I took a few deep breaths and realized that I had in my closet a Metal Max game from Tricklaser along with CF arms and the white Metal Max mod kits that make up nearly an entirely brand new awesome printer that I had intended to build months ago but never got around to it due to medical complications (had to have my leg amputated so that took some time and focus at from my passionate hobby that is what we're talking about currently ?). I ordered the remaining couple of parts I needed to complete my most properly outfitted rig yet (heater for the platform which is a borosilicate glass plate from SeeMeCNC) and kept a positive optimistic outlook that this most righteous frame and all the trick components would lead me to an easy success story when paired with the also brand new Duet Wifi board to compliment the new build.

    Well, the build went together like butter,truly, sand major round of applause to Brian (tricklaser.com) for his brilliantly designed frame along with the folder of printable parts that enable one to truly complete the build with finesse, such as FSR holders that bolt into the verses of the game for a perfectly even spacing and other such niceties as belt belt adapters that allow for straightened belt runs when used in conjunction with his metal trucks). To my amazement and pleasure, when in operation,the machine (thanks, I believe,to the Duet Wifi boards new drivers and the motor dampeners I installed) is almost whisper quite even when in full motion. In fact, it seems too good to be true – obviously nothing ever is so let's get to the point, which is, to my total disappointment and added frustration, the goddamned printer (in the first attempt,which was really quite satisfying), didn't crash the year into the buikd plate, but rather performed the calibration routine as it ought to, but the result is still totally underwhelming and does not appear to be properly compensating for the calculated probes it takes and, ultimately, can't even successfully spit out a simple yet cube without either having an uneven first layer, or must recently observed, second layer rubbing on first layer, resulting in another version of crap just by another method.

    Thankfully (for my sanity sake), in the interim of dealing with the frustrations one give it and purchased a Lulzbot Taz 6, which at least it's rock solid reliable, but I can't help my mind from dragging me back on a daily basis during any /all free time to tinker with this delta -- certain that I ABSOLUTELY MUST BE within minutes of having the perfect delta setup. The only thing is yet to try is add in the H offsets to possibly remedy the annoyances I've been experiencing in trying to get the perfect printer printing perfectly. Sorry for the long post and venting herein, but I've been avoiding posting into forums with the confidence that "surely, I can get this machine to perform as it should with the best components available", and as such, keep hearing my head into a wall trying to find where I'm overlooking, but finally I fixed and an asking (begging) for any/all suggestions that might well save this ultimate set of components from suffering the same fiery death that my RM2 did.

    Once again, I apologize for the lengthy post/venting, and thanks in advance for any help I might be able to get from those who are Jedi's in this arena...I once thought I was,but this has set me back a few notches, for certain!



  • Jonathan,
    Many people with RM2s could relate to your story; well written by the way. However, there are many people who have figured it out. As an engineer you know that a printer has only a few systems, and in a few hours you could probably identify which one needs attention. Despite using metal max, 1) did you measure your frame and ensure that towers/arms/plate dimensions are good and that they are square/parallel? 2) Is your bed absolutely flat and square? 3) Is your homing accurate? 4) Is your probing accurate? If the answer to all 4 Qs is yes, you should not really need H compensation. I would think it might be easier to make necessary adjustments/replacements to your parts to get it dialed in vs messing with compensation. On the other hand, there are a lot of things that could produce crappy first or nn layer and which are not related to the bed leveling per se.

  • That's very interesting advice, tera, so thanks. To be quite frank, no, I have not actually precisely checked the dimensions of the metal max frame from Tricklaser.com (A++++, BTW) for two reasons, first,for the premium the frame sells for one would automatically think that this was paramount during production to begin with,and second, given my experience with parts from TL inasmuch as all other parts is ordered from him in the past, I know their quality to be spot on. That being said, I suppose it couldn't hurt to spend a few minutes with a tape measure and micrometer just to be 101% sure, meanwhile eliminating sources of my frustration.

    Interestingly, (to me anyway, haha) since my rant posted above (thanks for taking the time to read the whole thing…I began to think maybe it was turning potential advisors away due to its length, but I'm happy it didn't) to my astonishment I I simply use medium grit sand paper to reduce the diameter of the plunger portion of my bed mounts via the. FSR mounts provided within the forums at SeeMeCNC, thus allowing them to slide more fiction freely than previously, and then allowed it to prove away and crossed my fingers. I believe there needs still to be some sort of offset value to be entered (but in which command in the config.g file to insert this value in??) due to the fact that whole I'm blown away just with the simple small stuff I had handy to print tests with as far as overall surface quality, and most of all how f'ing quiet this printer is while moving along at pretty high speeds so far (I have yet to explore the limits, at this point just thrilled to not be printing nothing over and over again. Believe it or not, I'm currently printing on the heart spreader aluminum cover sheet that normally I've got situated underneath the borosilicate glass, and even still, the results this far are amazing. Keep in mind in only one single successful print completed overall so far so there's still some experimenting and tuning to be performed and so I'll post updates of anyone's interested...dunno if that's going to generate any responses :). Oh, oh yeah, my point, which is that despite the print succeeding overall, the first layer still seems like it's too low to the plate, as I can't say for certain, but to the best of what I could tell was occurring whike splitting my attention between the likely to fail print and the TV, the very first layer didn't pay down anything (again, to the best of my recollection that is) but rather simply left indentations of where the filament ought to be. Fortunately, by the second layer the problem rectified itself and finished without issue!! Still, seeing as to how that's not par for the course, I'd love to know where to toss in that value to raise the z height a smidgen before commencing. In that way, it seems to me I'd be "in business" as well as "one happy camper"!

    Thanks again for reading this chapter of my forum posts 🙂 when I get the chance soon I'll make measurements to ensure geometric compatibility.

    Kind regards,


  • It may be worth your while looking at the Fisher section in the RepRap forums?

    As I understand it the issue with FSR type systems is the the force required to trigger them varies dependant on the distance from a sensor hence the reason for the H Parameter in the Bed.g file This compensates for that effect.



  • @ Zeplin, It seems that David had posted a similar, but a little more simple method for figuring out H, and it only requires four steps, found in here https://www.duet3d.com/forum/thread.php?pid=2681#p2681
    @ Johnathan, It's great to hear that you are making progress. I would suggest opening another thread where you can describe (and perhaps post a video/pics) of what you did, so that other could learn from your experience, and if you still need more help. I am sure someone will help. I would guess that in your setup FSR would be the weakest link, but a systematic approach always helps. Step back, examine the problem, examine possible solutions, and implement.

  • @all: I certainly appreciate the positive and supportive feedback I've received on this forum; enough so that I'm very inclined to give back a little, which is something which historically I've not really done much of. This is not a result of being selfish, but rather I often due to my feeling like I'm the last one to the party, if you know what I mean…?

    ....which brings me to....

    @terabyte: thanks for the no b.s., no attitude (as is overwhelmingly the case in so many other forms on the Internet ; an attitude of "know it all") response. I thought about what you wrote, above, regarding making a tutorial, build log, or video to help others out who might be facing similar issues. I only wish I had begun keeping video logs throughout my recent build as that would've been a great collection of issue / solutions for many to learn from, but back to the point, time permitting (as recently I've been asked to be the subject of a medical study on advancements in bionic prosthetic devices for above-knee amputees and so might become very busy in the coming months) I plan to try to keep good now going forwards as a result of the welcoming feeling I've received on this site.

    More to come, stay tuned... ?

  • Jonathan,
    It's great that you are staying busy. Forums here was one of the deciding factors for me to switch to DuetWifi, and I really love it. I have tried Replicape, and although I liked it too, I could not make it work without support.

    I wouldn't worry about "being last to the party," some actually never make it, and many, like me, still need help. Besides, most of us are here because we like it and not because we are in a race to get somewhere. Regardless of how my Rostock is going to turn out, I think I will build at least one more printer, maybe more.

    Good luck on your study, and if you have a chance, share at least some aspect of your build, or a print that you have made. It doesn't have to be educational either; there is a special forum here for showing off:)

Log in to reply