7X7 Mesh Bed Leveling much faster on new printer  

RE: 7X7 Mesh Bed Leveling much faster on new printer

Well they changed it in this text but there is still no information on the product description of the Upgrade Kit. Let's be clear. I had no issues with bed leveling with the PINDA V2. I cannot see any improvement with the Super PINDA. 

To be fair, they state however: "Otherwise, if your MK3/S runs without any issues, you will not gain any significant printing quality improvements." 

For MK3S users that have a perfectly running machine the upgrade is indeed more or less useless. On the contrary, in some cases the start of the print slows down significantly without any visible benefit. If PRUSA would not have replaced the SuperPINDA I would have use the old one again. 

SPINDA

I think the difference some may see is not in quality but in more consistent bed leveling.  I have on a single machine.  

Slow is relative (and subjective)

I think part of the thinking with SuperPINDA was that it would unify the part used on both the Mini and Mk3 and ease inventory issues as well as address a serious shortcoming on the original Mini. It works. It works well. It's just not as fast as some would like.

FWIW, the original 3x3 mesh bed leveling with the old PINDA probes wasn't much faster. Changes to the firmware made it much faster. The same may happen with SuperPINDA as the Prusa devs get familiar with it.

Hell, I still do a PINDA warmup before printing, so another 30-40 seconds on startup for a multi-hour job is nothing.

If it ain't broke, change the product description

Posted by: @alexander-3

[...] To be fair, they state however: "Otherwise, if your MK3/S runs without any issues, you will not gain any significant printing quality improvements." 

That disclaimer pretty much applies to any of the Mk3 upgrades. I still run my Mk3 in the original early 2018 configuration and it works wonderfully. I'm not about to make wholesale hardware modifications just because. Everything I have upgraded is not covered in any of the official upgrades. It's a great design that doesn't need a lot of fixes.

FWIW - Prusa did similar Ninja-edits on the textured bed product description. It went from "the one surface for all filaments that can be cleaned with a bit of acetone" to "a good surface for PETG that can only be cleaned with isopropyl alcohol" overnight.

acetone

Posted by: @bobstro

Posted by: @alexander-3

[...] To be fair, they state however: "Otherwise, if your MK3/S runs without any issues, you will not gain any significant printing quality improvements." 

FWIW - Prusa did similar Ninja-edits on the textured bed product description. It went from "the one surface for all filaments that can be cleaned with a bit of acetone" to "a good surface for PETG that can only be cleaned with isopropyl alcohol" overnight.

I am pretty sure this was a change in response to folks who used it as their primary cleaner and soaked their beds in the stuff.  

Almost, but not completely totally unlike perfect

Posted by: @cwbullet

 [...] I am pretty sure this was a change in response to folks who used it as their primary cleaner and soaked their beds in the stuff.  

Oh, no doubt, but it was still disappointing to see the description change between the time that I originally ordered the sheet and the 14 months later that it arrived. It went from being a universal print surface to a filament-specific surface that needed special care & feeding before I had a chance to try it out. Doubly disappointing when the damn thing didn't work with any filament, and triply so when the clone textured sheet I advertised performed exactly as was originally advertised. I understand it, but I just wanted to point out that I share that frustration. I love Prusa as a company, but they're not quite perfect.

Wow

Being the one that started this thread, I had no idea it would grow into this. 

Does the PINDA work even if it's slow?  Yes.   And most of the time, another 30-60 seconds on a hours-long print is no big deal.   

But when doing things like manual bed leveling (Nylock mod for example), or quick test prints for Live-Z adjustment, those extra minutes add up pretty fast.   

I never would have known I had an issue unless I had purchased a second printer and saw the change in speed, the newer printer being much faster.  

The physical appearance between the slow and fast SuperPINDA that came with the two printers is identical.   

The behavior is different, as the light stays on all the time on the fast one, and turns off during transits on the slow one.  

And the new SuperPINDA that Prusa sent also looks identical, but does have a different label on the plug end. 

One of two things happened here.   Either #1, the first batch of SuperPINDA probes have a defect, in which case you would think that Prusa could go after their supplier for replacement costs.    Or #2, there was a design flaw in the SuperPINDA which Prusa identified and changed / updated, in which case they would be on the hook for any costs themselves.  

I have had nothing but excellent support from Prusa and I kind of almost feel bad that I started this S#!t storm here with this thread, but at the same time, there is definitely an issue here with the early SuperPINDA probes.   

Worst case, if you really need the faster probing, at least now you can feel pretty confident that purchasing an updated SuperPINDA should definitely fix it.   At the start of this thread, I (we) did not know for sure that was the fix, seems pretty apparent now that it is for sure. 

Glad you started this

@twoballscrewball, Glad you started this as I then knew there was a solution. And it was. I struggled to convince (or should I say force) PRUSA to send a replacement that fixed my issue. So thank you.

It's a good discussion

Posted by: @twoballscrewball

Being the one that started this thread, I had no idea it would grow into this. 

Heh. That tends to happen, but you end up with a lot of good information that can be searched for and found by others.

[...] One of two things happened here.   Either #1, the first batch of SuperPINDA probes have a defect, in which case you would think that Prusa could go after their supplier for replacement costs.    Or #2, there was a design flaw in the SuperPINDA which Prusa identified and changed / updated, in which case they would be on the hook for any costs themselves.  

I think there's a 3rd option: #3 - The part used for the SuperPINDA operates over a wide range of timings. There are no guarantees that any 2, regardless of production date, will perform identically, but all can be expected to provide consistent results regardless of timing variations. In other words, the part is "in spec" but that spec allows for a wide timing variation. There may be milspec versions out there with tighter tolerances, but they typically cost a lot more.

I have had nothing but excellent support from Prusa and I kind of almost feel bad that I started this S#!t storm here with this thread, but at the same time, there is definitely an issue here with the early SuperPINDA probes.   

No need to feel bad. Prusa's a great company, but they're not perfect. And sometimes, expectations simply don't meet reality. A few years back, we would happily report that "this print is so good that I only need to sand it again for a perfect print". Now the expectation has turned more towards "I want prints so good that I don't need to sand them for a perfect print". Consumer expectations evolve over time. It can be tough for the same hardware to meet the full spectrum. A discussion is good, and I certainly am not feeling heated in any of this.

Worst case, if you really need the faster probing, at least now you can feel pretty confident that purchasing an updated SuperPINDA should definitely fix it.   At the start of this thread, I (we) did not know for sure that was the fix, seems pretty apparent now that it is for sure. 

I don't think we have enough data to confidently say "updated are guaranteed to be faster", nor do I think Prusa is necessarily willing or obligated to ship replacements for functional units at cost. They seem to be willing to in extreme cases, but "works" covers a range of tolerances as long as the part is in spec.

TBH, when I read these threads, I'm just glad I stuck with my OG Mk3 setup with the v2 PINDA and happily let it sit for 5-10 minutes on the first print of each day warming up. It's consistent and reliable. Kids these days. 😀  

Disagree

I think there's a 3rd option: #3 - The part used for the SuperPINDA operates over a wide range of timings. There are no guarantees that any 2, regardless of production date, will perform identically, but all can be expected to provide consistent results regardless of timing variations. In other words, the part is "in spec" but that spec allows for a wide timing variation. There may be milspec versions out there with tighter tolerances, but they typically cost a lot more.

Honestly, if I were the manufacturer I would not spec parts that they behave differently. Frustration and anger on the customer side is predestined. Is it worth that saving a buck which they could add to the products. Remember: they raised the price of the MINI+ and stated this was because of the superior SuperPINDA. One EUR more would not have been a deal breaker.

So, tend to disagree on that 😊

Supply chain

It is superior. On the Mini. They did say it didn't do so much on the Mk3. Most people with one printer are probably satisfied. It may be that they have to live with what they can get given the supply chain issues these days. Subaru can't ship $30K+ cars and they've got deeper pockets than Prusa.

unrealistic expectations.

Posted by: @bobstro

It is superior. On the Mini. They did say it didn't do so much on the Mk3. Most people with one printer are probably satisfied. It may be that they have to live with what they can get given the supply chain issues these days. Subaru can't ship $30K+ cars and they've got deeper pockets than Prusa.

This makes sense.  I think a lot of buyers expect more than a small company can deliver.  

RE: 7X7 Mesh Bed Leveling much faster on new printer

After stumbling on this thread and having the same issue with the slow bed leveling, also issues with the print quality, I chatted with prusa support, and what they told me is that they don't replace the super pinda. It seems that I will have to go back and forth with them, but I don't see getting anywhere. I guess they prefer to loose a customer than replacing a faulty part.

RE: 7X7 Mesh Bed Leveling much faster on new printer

Also, when I mentioned this thread to the support person and asked how come they replaced the super pinda in this situation, they told me that, somebody (from prusa that is) might have made a mistake replacing it.

Sad thing, is that I was so determined to get a prusa instead of a cheaper alternative, just to avoid all the hassle of trying to make the printer to actual print, but I have started regretting it.

I will also send an email and see how it goes.

RE: 7X7 Mesh Bed Leveling much faster on new printer

Old post, same issue. I upgraded a MK3 to MK3S+, updated it to MK3S+ firmware v3.12.2-5713, and redid all calibrations.

MK3S+ is generally very slow at bed-leveling. (7x7, 5 samples per spot)

For every "touch-off," it acts like it did back when it was a MK3 and there was a spot in the bed that was raised a little bit by an errant bit of plastic under the steel. Just like the video in this comment:

For awhile after the upgrade, it was intermittent: Occasionally, it would go to a touch-off spot and do 100% of the samples at full speed, or do a few samples rapidly and the others slowly. But lately, 100% of the touch-offs are slow.

Printing is fine though, with a good first layer across the entire bed surface, so it's accepting the bed measurements and adjusting accordingly.

RE: 7X7 Mesh Bed Leveling much faster on new printer

I know this is an older thread, but I just noticed (or at least decided to research) why my printer was suddenly doing the bed leveling much more slowly.

I had a print failure a few months ago and one of the casualties was the PINDA support arm on the carriage.  After I replaced the carriage part, I found that the PINDA was non-functional (the sensor tip was broken off).  I debated ordering a replacement off Amazon and instead ordered direct from Prusa (along with a spool of Prusament).

The replacement was a SuperPINDA, of course, and once I replaced it and everything was working.  I thought the bed leveling was slow, but I just didn't care until recently.  I thought maybe a firmware update changed it, but then it dawned on me that the SuperPINDA was a variable that changed.

Based on this thread, it looks like that was probably the culprit. 

I don't plan to do anything about it, as I just see it as a minor inconvenience.  I just thought that I would chime in and say that even recently purchased SuperPINDAs may not solve the issue of the slow mesh leveling process.  I ordered mine in late October 2023 and installed it mid-November.

RE: 7X7 Mesh Bed Leveling much faster on new printer

My understanding is that there  are 2 super pindas. They both work fine  and the one is significantly slower than the other. I had to go back and forth for a considerable amount of time (a week or more I think of everyday communication) with prusa support and finally I got a replacement, but only after I found and shared a screen that was mentioning the required time, which was actually half that what my printer required.

Long story short, I don't think that it is worth going through all that with the customer support, to get the faster one, and I wouldn't if mine wasn't faulty as well, but the bottom line is, that after that I stopped being a customer and during black friday that I got a new printer, I chose another brand.

RE: 7X7 Mesh Bed Leveling much faster on new printer

Funny you should post now, spiff75. 🙂 Just recently, I finished upgrading another MK3 to a MK3S+.

It's got a SuperPinda, and it does the mesh-bed-leveling just as fast as it did when it was a MK3 with the original Pinda probe. The kit was bought several months ago, and I just finally was given time to do the upgrade.

The other two MK3S+ printers I have here do MBL at the usual SuperPinda-slow rate.

One of the PrusaSlicer programmers over at Github said that SuperPinda is slow due to, I think it was something about the thing having a wider hysteresis band. So if that's the case, then I don't entirely know what's going on with these printers I've got, with one consistently doing MBL at full speed, and the others are slow. All three printers are running the same firmware, v3.13.2.

Oddities I've found:

- While a slow-probing printer is in the process of z-sensing, if I put a small flat screwdriver under the probe as it's starting to sense a location, as expected, the printer pauses briefly and moves the head upward a little and prepares to z-sense again. But when it retries the sensing of that spot, it does it at full speed. The next z-probe location is slow again.

- I just thought of something else to try, and based on a few minutes of messing with it, there seems to be an extremely fine band where a slow SuperPinda will do z-sensing quickly: If I lift up on the print bed while it's doing MBL, occasionally I can get it just right that it'll start sensing quickly. 🤷‍♂️ It's very particular about it though.