Threshold for auto-leveling?

I just replaced the nozzle on my Taz Workhorse with a brass nozzle and it seems to be failing autoleveling. I replaced the old nozzle since it seemed to become a magnet for filament and just gunk-up.

I’m able to measure continuity between the tip of the nozzle to the crimp on the print-head’s molex connector but the system seems to be struggling. My Fluke seems to be reading <25 ohms between these points.

I’m guessing I’m in a grey area for whatever threshold is being used. I’m guessing it’s being read by an ADC? Is there a way to adjust this value? Or is this a digital value and I may need to add some thermal paste to reduce the resistance on the threaded connections between the nozzle and the hotend and the hotend to the heat sink?

EDIT:// Correction on the resistance measurement. Moving from a handheld to my HP 34401A, the resistance measured is a consistent 6-ohm. The wire itself from the heat sink to the connector meaures to be the majority at about 4-ohm. This seems twice as high as it should be but with a two-wire setup and measuring in the ohms, it’s only going to be so accurate.

The red wire attached to one of the heat-sink screws is the zero-sense wire … that’s the wire used to detect when the nozzle touches each corner of the bed.

You should have less than 1Ω resistance between the nozzle tip and the zero-sense wire.

On a TAZ Workhorse, the zero-sense wire is attached to the heat sink (older TAZ printers attached the wire to the heat-block). This means the conductivity has to go from the heat-sink through the cold-end, to the heat-break tube, through the hot-end and into the nozzle.

Cold filament feeds down through the cold-end into the heat-brake tube … the threaded tube that connects the ‘cold’ and ‘hot’ sides of the extruder and provides some separation between them.
While the heat-break threads into the TOP of the heater-block, the nozzle threads up from the BOTTOM of the heater block … and they share the same threaded passage.

It is very important that the threads of the heat-brake and nozzle be solidly in contact with each other. If the heat-brake isn’t threaded in far enough, the nozzle will bottom-out (on the hex head) before the threads mate. This creates a gap and melted filament will press into that gap and start oozing through the threads. That will act as an insulator and result in unreliable bed leveling.

You should have < 1Ω from the nozzle to the zero-sense wire. If you have more, you may have a leak.

If you see filament oozing out through the heat-brake on the top of the extruder, then this is your issue. It’s difficult to see because the fan-duct is in the way. There are four screws that hold the heat-sink onto the rest of the AeroStruder assembly … the bottom two screws (they are very long screws) hold the fan-duct in place. If you remove just the bottom two screws (one of which is also holding the red zero-sense wire in place) then you can carefully lower the fan-duct (the fan will come with it – be careful that you don’t pull on the fan wires) you’ll be able to visually inspect the heat-brake gap.

I did have a leak on my Workhorse and had to adjust it.

The filament will seize the threads … so I had to heat it up to the melting point to unthread it. Once it was apart, I used a soldering iron to keep the threads hot while I used a brass brush to clear the threads of excess filament. I gave my nozzle one additional turn (so it’s one thread deeper into the heater block) and now my extruder nozzle makes solid contact with the heat-brake tube … no more leak and everything is great.

The wire itself should be basically a ‘dead short’, zero ohms.

“Dead short” is a good concept, but it always has a practical resistance attached to it. Measuring it externally, there seems to be a threshold between 4 and 6 ohms that differentiate a “pass” versus a “fail”. I have some lengthy prints over the next week, but I hope to be able to tear the system apart to see how it’s actually being used. Using the nozzle itself is the best method since it’s what’s actually dispensing the filament, but such a small difference in resistance causing a fault is a bit concerning. If the printer is still using an 8-bit Atmega 2560 to run the who thing then there’s a lot of opportunity for improvement to tighten the control loop to be able to make the leveling more precise and a lot faster.

The newer ST controllers, especially the H7 series, could really add a lot of horsepower to this to polish off the rough edges.

I agree but a dead short should not be measured in whole ohms. you said: ‘The wire itself from the heat sink to the connector meaures to be the majority at about 4-ohm.’ To me that seems very high if it’s copper wire. What gauge is the wire?