BIQU Microprobe

jens55

@omni, I do in fact run unshielded stepper wires. My issue is that the BLTouch works just fine but the accuracy/repeatability is crap. The accuracy of the BLTouch is between 0.001 and 0.05 and it is not unusual for my setup to take that full 0.05 range .... we are not talking 0.005 but 0.050 so when I see 0.001 to 0.003 I start to drool .....
I turn my bed heater off during probing and only the Z motors move (slowly) so I am reluctant to call 'interference' but I suppose I can see how the Z motor cabling is routed and see how that can be changed.

omni

@jens55 Well that may well be your main problem - you are probing a cold bed or a hot bed which is cooling during the probing since your bed heater is off. The aluminium plate can warp and fluctuate in height BY A LOT in just a matter of seconds depending on the isolation, much more than the problem of accuracy of the probe. Before you probe the bed you need to make it stabilize in tempreture and keep the printing temperature of the bed (preferably with PWM) at all times during the probing in order to get the most accurate results. Probing it cold or during cooling is pretty much useless if you are looking for very high accuracy during printing. 0.003 repeatability is plenty, those are not values that you are going to notice. But bed height fluctuating +/- 0.3mm because of the cooling you are definitely going to notice...

jens55

@omni I respectfully disagree - the bed and build plate I am referring to has a fairly high thermal mass and is very slow to change temperature. Further, RRF has specific features solely for turning bed heating off while actually probing. Before probing, the bed is thermally stabilized (substantial wait period after starting up the bed heat) and during probing the overall bed temperature is stable.
I used to run with bed heating on (PWM) and turning off the heating during actual probing improved the consistency of the readings.

omni

@jens55 Well yes, you might be right if it's a small by surface but very thick bed and well insulated it might not make such a big difference. I have a 525x525x6mm bed and it warps like crazy when heating / cooling. It can be +/- 0.5mm difference in corners. That is why I adjust and probe the bed only when fully stable at the printing temperature. But I don't think there is a need to turn off the bed while probing if you are using bltouch. For some other types of probes maybe...

jens55

I am having no end of issues with my BLTouch. A bed mesh of my glass plate looks like the Rocky Mountains (Western Canada). This is a Ver 1.0 so it's ancient.
After some contemplation, I was too intrigued by the 'Microprobe' promise of a higher accuracy and better repeatability over a BLTouch ver 3.1. I should get the sensor in about 3 weeks. I am quite skeptical about the accuracy claims but the only way to find out is to try it out.
I plan on reporting back after I have had a chance to do some testing.

dc42

@jens55 the BLTouch is affected by the magnetic field generated by some bed heaters. Have you tried using parameter B1 in the M558 command to turn heaters off during probing?

jens55

@dc42, the heater is turned off during the actual probing. I also have a BLTouch Ver 3.0 on another printer and the difference in repeatability is night and day. We are talking about 0.1 mm differences between repeated probings at the same spot!
Since I have 3 weeks to stare at a non-operating printer, I might try adjusting the ferrite slug in the BLTouch which some people have said improved their accuracy. Not sure why and how but I got nothing too loose at this point.

Phaedrux

@jens55 said in BIQU Microprobe:

I have 3 weeks to stare at a non-operating printer

There's always manual probing or even stall detection probing in a pinch.

jens55

@Phaedrux Ugh .... manual setup is ok for a rough levelling of the bed but I can't imagine running a printer without a bed mesh compensation map defined.
Just for the heck of it, here are probing results for two runs. The location of the probe point is the same for each run:

-0.036 -0.148 -0.171 -0.165 -0.178 -0.174 -0.175 -0.159 -0.163 -0.153 -0.163 -0.155 -0.175 -0.169 -0.163 -0.175 -0.163 -0.170 -0.165 -0.166 -0.180 -0.176 -0.164 -0.141 -0.165 -0.176, mean -0.161, deviation from mean 0.027

-0.009 -0.007 -0.004 -0.007 -0.015 -0.018 -0.014 -0.011 -0.018 -0.004 -0.011 -0.011 -0.014 -0.010 0.002 -0.005 -0.004 0.007 0.001 0.011 -0.002 0.007 0.012 0.001 0.012 0.005, mean -0.004, deviation from mean 0.009

Most of the measurements in one run are pretty much the same but there are extreme outliers like -0.036 in the first run with all the remaining runs centered around -0.160 or so. Note that the second run looks pretty good at first glance but even it ranges from -0.018 to + 0.011. While I wouldn't be too thrilled about a 0.029 mm in range, it is acceptable (barely) for even a 0.1 mm first layer. Now, when considering all the measurements from both runs which are for the exact same spot on the bed, we are going from +0.011 to -0.180 mm for a total range of 0.191 which is a long way from acceptable for the normal 0.2mm first layer thickness.

Doing a probing run on a printer with a BLTouch Ver 3.0:
-0.041 -0.039 -0.035 -0.037 -0.039 -0.035 -0.032 -0.039 -0.042 -0.044, mean -0.038, deviation from mean 0.003

I also had a chance to try tweaking the ferrite core on the BLTouch which apparently improved things with some people. I saw no discernible improvement.

dmandn

@jens55 I'd be tempted to see if there is any flex in the mount for the bltouch, as that seems to be where my current toolhead design is getting fluctuations from.

One printer is seeing 0.001 fluctuations, while the other with all same except toolhead, 0.02

jens55

@dmandn Both printers use the same bits for the carriage and all the parts thereof. The thing about flex that I don't understand - there is practically no load on the probe mount. It holds up the probe and the pin weighs what, a gram or two. The only thing that changes is in one case the mount carries the probe and the pin and in the second case the pin is supported by the build plate. How can there be flex at all? I have read somewhere to make sure the probe mount is solid but I don't understand how this could possibly be a factor. Maybe I am missing something ?

jens55

So I have finally installed and tested my BIQU Micro Probe. The overall impression - I am not impressed!
The details:
The pin has a very short movement range - significantly less than a BLTouch. The 'manual' says 4 mm (I measured slightly less but in the range). The BLTouch ver 3.1is 6.6 mm. It is only 2.6 mm difference but it makes a considerable difference. The retracted pin is only about 1.5 mm higher than the nozzle and when probing, the nozzle is only 1.5 mm or so above the trigger point. That doesn't add up to 4 mm of travel but it was as close as I could measure. In any case, these clearances are awfully tight.
Now the real issue - a couple of probing runs over the same spot:

G32 bed probe heights: 0.060 0.054 0.049 0.068 0.054 0.066 0.059 0.054 0.049 0.047 0.045 0.049 0.070 0.062 0.056 0.058 0.051 0.047 0.061 0.058 0.053 0.053 0.047 0.047 0.066 0.055, mean 0.055, deviation from mean 0.007
Total range of measurements: 0.021 mm

G32 bed probe heights: 0.055 0.052 0.066 0.061 0.055 0.052 0.059 0.060 0.058 0.056 0.061 0.068 0.060 0.057 0.061 0.054 0.050 0.056 0.051 0.057 0.052 0.074 0.064 0.066 0.065 0.076, mean 0.060, deviation from mean 0.007
Total range of measurements: 0.025 mm

Total range of both runs for the same testing spot: 0.029

These measurements are for the same spot and they are worse compared to the BLTouch. A lot of reports claim the Micro Probe to have an accuracy about ten times as good as the BLTouch but this is definitely not the case here. Now compared to the presumably defective BLTouch ver 1.0 there is a 10 times improvement but that is not the case against a BLTouch ver 3.0

G32 bed probe heights: -0.009 -0.009 -0.006 -0.003 -0.009 -0.005 -0.007 -0.009 -0.010 -0.009 -0.010 -0.009 -0.011 -0.005 -0.011 -0.010 -0.009 -0.010 -0.011 -0.010 -0.010 -0.010 -0.013 -0.014 -0.010 -0.011, mean -0.009, deviation from mean 0.002
Total range of measurements: 0.011

G32 bed probe heights: -0.006 -0.006 -0.006 -0.009 -0.010 -0.007 -0.009 -0.009 -0.005 -0.009 -0.009 -0.006 -0.007 -0.007 -0.007 -0.007 -0.007 -0.007 -0.007 -0.009 -0.009 -0.010 -0.011 -0.009 -0.011 -0.010, mean -0.008, deviation from mean 0.002
Total range of measurements: 0.006

Total range of both runs for the same testing spot: 0.011

Now these are two different printers .... but I can't see that to make a difference for these tests.

The two supposed points in favour of the Micro Probe are lack of sensitivity to magnetic fields (ie from the heaters switching on/off) and possible larger range of operating temperature range (the BLTouch seems to top out at around 50C or so).
The Micro Probe is totally sealed and has a non-replaceable metal pin, the BLTouch has a replaceable plastic pin. I am not sure which is better.
I did not address size (the MicroProbe is about half the size of a BLTouch) nor speed (irrelevant to me)

I had anticipated replacing the BLTouch ver 3.0 with a Micro Probe for better accuracy but that will obviously not happen. There is some thought of replacing the just installed Micro Probe with a new BLTouch but I am not sure the slight gain in repeatability is worth the cost at the moment.

dc42

@jens55 said in BIQU Microprobe:

Now these are two different printers .... but I can't see that to make a difference for these tests.

It could make a large difference to the results, depending on how similar the Z axis construction is between the two machines. I wouldn't like to say that one probe is more reproducible than the other without testing them on the same machine.

jens55

@dc42, one printer is a CR10 (300-300) and the other is a CR10-S5 (500-500) so the z axis construction is basically identical. I am curious though how error can be introduced even if the z axis is constructed differently. Maybe the resolution (ie motor step angle and z screw pitch)? It's a slow speed process with no x/y action. It's not a question of accuracy but just repeatability with the Z axis going up and down.
What am I missing?

DIY-O-Sphere

@jens55
If you really want to know, you need a dial gauge.
The problem is that with every measurement the Z-axis is moved.
There is always some form of backlash. In addition, there are also statistical deviations.
Especially with a light granty, I would expect this to be more noticeable. On a machine where the bed moves in z, the weight of the print bed can be an advantage because it preloads the z-axis.
That is why a thousandth of a mm is never measurable on a hobby machine. You would need much more massive and precise mechanics for that.
Also, please don't look for one hundredth of a mm..... be satisfied if you get to about half of a tenth. You are trying to achieve an accuracy that our standard printers do not offer mechanically.
If your measurements with the BLT are correct, they are more than sufficient....
By the way, I also print my 1st layer with 0.1mm and a BLT V2. I print on FR4.
I have noticed that it degenerates after 3-4 prints at approx. 100°C.
The adhesion is then very spotty.
In my case, it does not help to clean the surface. Only sanding really solves the problem.
Until I realised this, I also spent ages optimising the 1st layer height!

fcwilt

@DIY-O-Sphere said in BIQU Microprobe:

The adhesion is then very spotty.

The biggest improvement for me came when I cut the first layer speed to something very slow - say 10mm/sec.

Frederick

jens55

@DIY-O-Sphere, thank you for your input. You are of course correct in that if the mechanics are not accurate or repeatable then the whole spiel about probe accuracy is irrelevant. No matter how accurate/repeatable the probe is, it can never show a repeatability that is better than the underlying mechanics. I completely missed this in my earlier contemplations.

I guess what I should do, if I was curious enough (which I am not), is to somehow install the probe on my Jubilee printer because the z probe of the jubilee is about an order more accurate/repeatable which would imply that the mechanics are more repeatable. If the printer mechanics are causing the issues then this would become apparent immediately.

I am impressed that you are able to do 0.1 mm first layers. It took me forever to get 0.2 mm layers dialed in and even after they are dialed in I still have sections where the gap is too much and other sections where the gap is too small. This is with 441 measurement points but of course the area in between the measurement points is interpolated so just an approximation.

I believe that the pebbly PEI surface of spring steel build plates helps a lot and possibly allows for a larger satisfactory adjustment range but I have a habit of dragging my nozzle across the build plate which would quickly destroy the PEI layer. I have stuck with glass because it is more forgiving to stupid. I have yet to warm up to specific adhesion enhancers.

jens55

@fcwilt, agreed. I go down to 50% of my normal speed and small features are reduced another 50% (molasses in the high artic moves faster)