Pinda 2 dilemma

fcwilt

@fcwilt well it should be both bed and probe they both have compensation numbers. but if i have to only pick 1 would be the pinda probe. bed doesn't drift numbers as much.

Does the probe have a built-in temperature sensor of some sort?

Frederick

sinole

@fcwilt yes, its PINDA v2

fcwilt

@sinole said in Pinda 2 dilemma:

@fcwilt yes, its PINDA v2

So the sensor has a pair of wires for getting the signal from the distance sensor and a pair of wires for the temperature sensor.?

Are you using v3.3 or v3.4 of the Duet firmware?

Frederick

oliof

@sinole @sinole you need to do four things

a) connect and configure the temperature sensor as normal (from another thread M308 S2 P"exp.thermistor_7" A"PINDA" Y"thermistor" T120000 B4725, your pin will likely be another).
and set the thermistor in G31 using the H parameter.

b) determine the trigger height at normal ambient temp, and set Z and ambient temperature in G31 using Z and S parameters

c) determine the trigger height at elevated temp. Then calculate the difference between trigger height at normal temperature and trigger height at elevated temperature and divide by the difference between ambient and elevated temperature. Use that result in the T parameter of the G31.

d) probe with temp compensation

Example:

a) You defined the temp sensor as the third sensor after the bed sensor (H0) and the print head sensor (H1), so your G31 gets H2. If you have multiple print heads or a chamber heater sensor, it could be another sensor number. read the temperature of this sensor, let's assume it's 20C and roughly matches bed and print head sensor when they are inactive. Let's assume your ambient temp is 20C

G31 H2 S20

b) Following the normal method to determine trigger height and get 2.6mm trigger height, and your probe is 30mm to the left of the nozzle and 5m in front of it.

G31 H2 S20 X-30 Y-5 Z2.6

c) Set your printer operating temps to something you'd normally probe at, like bed temp to 90 and print head temp to 250 (assuming an all metal hotend and printing ASA). Read the temperature of the probe thermistor. Let's assume for the example that this gets the ambient temp of the probe to 70 (it'll likely be lower but this makes it easier to do the maths in the example). Then determine the trigger distance at these settings. Let's assume for the example that the trigger distance is now 2.9mm. Then you divide the difference between trigger heights (2.9-2.6=0.3) by the difference in temp (70-20=50) so 0.3/50= 0.006. That's your T value.

G31 H2 S20 T0.006 X-30 Y-5 Z2.6

fcwilt

@oliof

Is the firmware feature specifically for the Pinda v2?

Frederick

sinole

@oliof thats alot to take in but thanks, I will work on it. I connected the probe to marlin printer and measure all of the offsets. it should be easy to just calculate T values. I assume with G31 H2 S20 T0.006 X-30 Y-5 Z2.6 you set up a linear pattern and you don't and to define more parameters am I correct?

oliof

@fcwilt no, it works for any temperature compensating setup. What I don't know is whether you could re use the bed thermistor for poor man's compensation.

With the high frequency probes like SuperPinda, temperature compensation is not as much required. I prefer those setups. In my experience, they drift no more than 0.02mm in typical 3d printer setups without a heated chamber. I'm not sure that you get much better results with temperature compensation unless you introduce lots of measurement tools and effort.

@sinole yes, compensation is linear.

fcwilt

@oliof said in Pinda 2 dilemma:

@fcwilt no, it works for any temperature compensating setup. What I don't know is whether you could re use the bed thermistor for poor man's compensation.

Without knowing how a probe changes trigger height with temperature how can a fixed compensation scheme work?

Frederick

sinole

@oliof you can build your own pinda v2. when you set it up its basically same thermistor as the 100k thermistor we used on beds. its just inside of the probe. I think it will still work even if you kapton tape a bed thermistor to any regular probe. its just very unsightly.

oliof

@fcwilt said in Pinda 2 dilemma:

Without knowing how a probe changes trigger height with temperature how can a fixed compensation scheme work?

You measure that as I outlined in my process, but instead of having your own thermistor you use the bed thermistor. It will be less precise, but may be sufficient. It would be worth to do as an experiment.

oliof

@sinole Sure, but it's not required with a high resolution probe like SuperPINDA and the various probes that are the same or quite similar OEM product.

fcwilt

@oliof said in Pinda 2 dilemma:

@fcwilt said in Pinda 2 dilemma:

Without knowing how a probe changes trigger height with temperature how can a fixed compensation scheme work?

You measure that as I outlined in my process, but instead of having your own thermistor you use the bed thermistor. It will be less precise, but may be sufficient. It would be worth to do as an experiment.

I wasn't clear.

The firmware uses a formula that makes an assumption of how the probe trigger height changes with temperature.

Is it a given that all such inductive probe follow that formula?

Thanks.

Frederick

oliof

@fcwilt no idea (-: Marlin uses a table with discrete values and interpolates between them.

I'll leave the experimentation to others, since I'm happy with probes that are resistant to temperature changes.

sinole

@oliof I'd be open to any option that you can suggest. problem is that I live in US and only real heat resistant probe that I know is superpinda you have to buy directly from prusa. I have to pay more than probe price for shipping. Its really not budget friendly. printed solid was purchased by prusa, they are the official prusa distributor now, hopefully they will carry it soon.

fcwilt

@sinole said in Pinda 2 dilemma:

I'd be open to any option that you can suggest. problem is that I live in US and only real heat resistant probe that I know is superpinda you have to buy directly from prusa. I have to pay more than probe price for shipping. Its really not budget friendly. printed solid was purchased by prusa, they are the official prusa distributor now, hopefully they will carry it soon.

Why try your current probe using the steps outlines by oliof?

Frederick

Do you by any chance know the switching frequency of the Pinda probe?

Generally the higher the switching frequency the faster the response time of the probe and thus the faster you can probe.

I get my sensors from these folks - I have preselected the inductive sensor selection.

AutomationDirect Inductive Sensors

One nice thing about this site and their sensors is they provide actual specs covering all aspects of the device.

None of the ones I use have temp sensors but I have not found that to be necessary.

Frederick

sinole

@fcwilt its not really a perfect solution. the probe doesn't really behave linearly and it doesn't behave exactly the same every time, but marlin algorithm with 2 tables for bed and probe temperature makes a pretty good use of it.
PINDA is a name licensed by prusa i think. its justa regular inductive probe. pinda v2 has a thermistor in it. and super pinda is just a very expensive accurate inductive probe. as I said you can kapton tape a thermistor to your favorite probe. put it on the back or something make it look nice. dont open the probe though. they come with vacuumed seal to prevent humidity.

sinole

btw if it helps, I probe really fast with pinda and repeatability is pretty good. i basically probe as fast as tr8x8 can handle going. you dont need 80$ probe.

fcwilt

@sinole said in Pinda 2 dilemma:

the probe doesn't really behave linearly and it doesn't behave exactly the same every time, but marlin algorithm with 2 tables for bed and probe temperature makes a pretty good use of it.

Well it would be easy to implement a table or two with firmware 3.3.x or 3.4.x

Frederick

fcwilt

@sinole said in Pinda 2 dilemma:

repeatability is pretty good

What is "pretty good"?

Thanks.

Frederick

sinole

@fcwilt standard deviation of 8 micron