Using the Taz6 without auto leveling?

I have a Taz 5 at home, and recently purchased a Taz 6 for work. I’ve had great luck with my 5 and assumed the 6 would be that great and better with the fancy bed leveling but like many others I find I get inconsistent first layers. Sometimes the nozzle is smashed on the bed, sometimes it is say .75mm above the bed, and sometimes it is just right. I don’t have time to baby the printer every time I send it a print and I am more than happy with the performance of my 5. I leveled it like, a year ago(?), and it cranks out parts like a champ.

Is there some way to simply level the 6 manually and run it like I run my 5?

Not without modification. The taz 5 bed isn’t a direct bolt up to the taz 6 mounts and the 5 leveling setup won’t fit a 6 bed. You would also need to add an endstop switch and adjust target. Doable but I don’t know of an existing retrofit.

Might be easier to troubleshoot and adjust the system you have as what you describe isn’t normal behavior and usually indicates a loose something and or a dirty nozzle

After printing nylon, leveling doesn’t work reliably on my 6 unless I scrape the carbonized goo off the nozzle with a knife and use a scotch-brite pad. It seems nylon is too tough to wipe off with the wiper pad (unless I heat the nozzle up to almost the printing temperature, but then it starts oozing nylon so then leveling doesn’t work because of that…)

I was thinking more of a software modification. I am assuming that the auto leveling process generates parameters that define the virtual bed plane? Couldn’t these values be captured and, if an appropriate G code existed, sent to the printer before printing? I could see a couple of different printer profiles. One that would contain the G code for auto leveling, and one that would use pre-entered values. You could then polish up your nozzle and get a solid level, copy the parameters into your other profile, and use that until you feel you need to re-level the bed. In my experience with my 5, you could go quite some time.

I mean, is it just me and a few others who have posted that are having these problems? Are the vast majority of users just hitting print over and over and getting fabulous prints every time? If so I’m feeling pretty bad about myself right about now. Or are Taz 6/Mini users just used to babysitting that first layer? Cancelling the print if something isn’t right, cleaning the bed, scrubbing the nozzle and trying again. Who has time for that?

Regardless, in order to better educate myself, what role does the Z button switch play in this? I know sometimes there will be a bit o plastic that oozes out that I can’t wipe before it hits the button. It then wipes the nozzle and touches the corners. Does it use that switch like my 5 does as a sort of Z stop and then “dials it in” touching the corners? In that case maybe that bit o plastic matters?

FWIW, the Marlin 1.1.1 I’m running does seem to save its bed leveling parameters, although they’re not active on powerup by default. I think the problem is that if you save them, you put all your faith in the z-home switch, which is mechanical and likely much less accurate than the nozzle probes.

As for the Z-button, my impression is that its only purpose is to set the z-position for the probe. If its trigger point varies a bit, it will only look to the probing sequence as if the bed has moved up or down a bit and once it calculates the bed plane from the probe points, that no longer has any function. I might be wrong about that, though.

There also seems to be an appreciable effect of temperature on the z-leveling, which makes sense. When the nozzle heats up, it will get longer, and the bed will expand upward. Both of these combine to put the nozzle closer to the bed as the temperatures go up. I don’t know how people who don’t re-probe for each print deal with these variations (unless they always print at the same temperature).

As for “hitting print over and over”, I’ve removed the probing sequence from the generated G-code, precisely because it’s unreliable, at least when printing nylon. Instead, I manually probe after powering up, make sure I get a good result, and then print multiple jobs without re-probing.

Hi lutorm,

I hadn’t thought about your approach. I’d be happy to have a “calibration” print that I run every morning when I turn it on that runs the probe sequence. I could spend 10 minutes and make sure the nozzle is clean and I get a good probe. Then run prints without probing throughout out the day.

But, how do you do it? I created a new profile and removed the wiping and probing business (but I still home X,Y,Z) and it is printing too low. It’s close, it isn’t pushing the bed down noticeably, but the nozzle is flat on the bed. Should I not home Z?


What I did was create a G-code file with just the home/probe sequence. That way I start up the printer, print “probe.gcode” which does the homing and bed leveling, and then I’ve removed that part of the start G-code from the generated prints. I basically replaced it with a M117 that says “You DID home the printer, right?” and waits for the user to ack it, and then just does the “wait for temp” part. I don’t think you can home between every print (at least not the Z-axis), because that resets the Z-height.

If you like, you could also add a move to the probe.gcode file to move to X0 Y0 Z0.5 or something so you can verify it’s got a good probe. (I’m running with the 1.1.4 Marlin now and there’s something funky about the autolevel code. Sometimes it’s off by ~0.5 mm even though it detects the corners perfectly. Most of the time just running G29 again will work. I haven’t tracked down what goes wrong because I’m in “production mode” now but at some point I want to do that. It’s unreliable enough to make me not trust it without verifying with a feeler gauge that it’s good. No idea whether the stock lulzbot marlin has the same problem.)

The only thing to be careful about is that you don’t bump the table so the steppers lose position when removing the prints. If you “move axis” right before removing them, the steppers will be powered on and you are much less likely to move them.

OK. Makes sense.

I’ll modify what I have to remove the Z home. I’ll try keeping the X and Y home for just the reason you state regarding moving the bed when removing parts. I generally don’t mean to, but it happens.


Removing the Z home was the ticket. I still need to tweak the g code, but the basics are there and my first non-probe print is in progress.

I will create a probe only g code file which I can run when firing up the printer for the day, and then run two printer profiles. One will omit the probe and just home X and Y, the other will include the probe sequence. Then I can choose between them as needed.

Again, thanks for the suggestions and help.

Cool, good to hear you got it going.