search

MarlinFirmware / Marlin

Marlin is an optimized firmware for RepRap 3D printers based on the Arduino platform. Many commercial 3D printers come with Marlin installed. Check with your vendor if you need source code for your specific machine.
https://marlinfw.org
GNU General Public License v3.0
16.36k stars 19.26k forks source link

RE-ARM, RAMPS 1.6plus, and PT100 #13188

Closed khomouda closed 5 years ago

khomouda commented 5 years ago

I recently upgraded my CoreXY printer with Re-ARM and RAMPS 1.6 PLUS and installed TMC2130 drivers in it. the main changes were: 24v power supply with no extra changes except a voltage converter for the fans. BLTouch does not work off the servo 3.3v so I had to connect its power supply to 5v in J3 in the RE_ARM The PT100, using the e3d amplifier, works off A9 in AUX2 in the Ramps, which supply 3.3v so as not to burn the MCU, but I had to call it 3 in the pins file for it to be recognized, but, at room temperature, the bed is reading 22 and the PT100 is 24, when heating it goes quickly to 270 and halts the printer, although my DMM temp probe reads 95 at 200 of PTT, I think the temp table needs changes with voltage changes (0.8v at 24degree@3.3v vs. 1.2v at 24degree@5v)

landodragon141 commented 5 years ago

Yeah that seems logical. I just looked through the PT100 documentation and it indeed is intended for 5v. https://e3d-online.dozuki.com/Wiki/E3D_PT100_Amplifier_Documentation Based on the documentation for the INA826AIDGKR which is the amp chip it can accept a minimum voltage of 2.7. I guessing that we could build an alternative temperature table for 3.3v operation.

landodragon141 commented 5 years ago

So after looking at the temp table it looks like every ~0.04v equals a 10C delta. If you could do some measurements we can determine how we need to transform the existing table.

khomouda commented 5 years ago

I DON'T THINK MY DMM PROBE IS ACCURATE ENOUGH FOR MEASURING an aluminum block, how would you think measuring is accurate

landodragon141 commented 5 years ago

Well it just needs to be close enough to establish how it differs from readings at 5v. Anyways we don't care about exact measurements here. Even if your probe is off a degree or two we are just trying to establish a pattern to begin with. Heck I'd measure the temp with a meat thermometer taped to the block with aluminum tape. (I may or may not have done that before :D ). Honestly I think you'll have a more difficult time accurately measuring the voltage than the temp. Anyways just sharing some thoughts on how this could be accomplished in the easiest way.

khomouda commented 5 years ago

no, the voltage is easier from the terminal coming from the thermocouple and its ground, but my DMM probe is not accurate when measuring at high temperatures from the block, anyways I will try

landodragon141 commented 5 years ago

You don't need high temps. Measure at 40, 50, 60 etc

khomouda commented 5 years ago

up to which temp?

landodragon141 commented 5 years ago

The table at 5v is completely linear so high temps aren't necessary.

landodragon141 commented 5 years ago

As high as you want. As I said we are just trying to establish a pattern.

khomouda commented 5 years ago

ok

khomouda commented 5 years ago

here you are sir PT100 @3.3v.txt

AnHardt commented 5 years ago

Tried to recalculate table_20 from the theoretical values, PT100.zip For this table the PT100s resistance is assumed to be linear, instead of taking the exact values from literature. That makes the small difference to the original table_20. If amplifier board and ADC have the same voltage, regardless of if 3.3 or 5 or any other value, the table is the same. Voltage shorts out of the equation.

khomouda commented 5 years ago

Thanks for the reply, but I don't understand it, the reported temperatures by the printer were at least 50 to 100 degrees more than the measured by my DMM, so the table 20 is completely wrong with 3.3v. will DC 3.3 -5V MAX31865 SPI board work? and how do I connect it

AnHardt commented 5 years ago

the reported temperatures by the printer were at least 50 to 100 degrees more than the measured by my DMM, so the table 20 is completely wrong with 3.3v

No. It means the theory what table_20 is made with does not match your system, or, what is the same, your system does not match the theory.

An incomplete list of possible errors - ordered from PT100 to ADC:

khomouda commented 5 years ago

That same PT100 was working perfectly on the same printer on RAMPS14 (was on A11) before I upgraded to Re-ARM (now on A9 but mentioned in pins file to be 3) and taking voltage and ground from the adjacent pins in AUX2 (3.3v)

AnHardt commented 5 years ago

re-arm-analog The RE-ARM makes relatively high effort to keep the analog voltages separate from the digital ones. It could make a difference where you take the 3.3V and/or GND for the amplifier board. Analog GND is exposed on the RE-ARMS temperature pins. Don't know where to find analog 3.3V exposed.

Bob-the-Kuhn commented 5 years ago

I'm a bit confused. What pin is the PT100 amplifier tied to?

The PT100 amplifier doesn't have enough muscle to drive a 4.7k pullup resistor when powered from 5V. Dropping it to 3.3V makes it worse. It must be tied to an analog in that does not have a pullup on it.

To make matters worse the analog inputs are VERY noisy. The filter helps but it still tends to bounce. T0, T1 & T2 on the Re-Arm card are the best of the analog inputs.

I gave up trying to use the PT100 amplifier on my DUE based system. I'm currently running the PT100 thermistor/sensor directly into an analog input and have changed the pullup from 4.7K to 1K. The down side is that only 2 Ohms of wiring resistance will show up as a 5 degree C fixed positive offset. That really isn't a big deal since you normally need to do some test prints to find the temperature sweet spot every time you switch to a new spool.

AnHardt commented 5 years ago

There is no way to connect the amplifier to T0-T2. The 4K7 pullups do their job - the temps will be off. (except you remove the pullups.) But taking AGND from there could be a good idea.

Had relative good experience with PT1000 and 4K7 pullups. Best experience with PT100 i got with a SPI interfaced MAX31865-board with its own ADC (still not in marlin).

khomouda commented 5 years ago

Thanks for your input, I connected it to A9 in AUX2 together with a 100uF cap according to issue #7646 and called it 3 in pins file (never knew why). so what is your suggestion? the NTC thermistors did not work at all with PETG, may be are too high of temp difference to the low side, so that it does not melt the filament at 270 deg

khomouda commented 5 years ago

I will try using AGnd when I go home, I also ordered MAX31865-board just in case it is supported in Marlin

khomouda commented 5 years ago

I tried using AGROUND in J8 but same response rocket increase in temp reading till it halts the printer, although it never can be that high in this few seconds

Bob-the-Kuhn commented 5 years ago

My suggestion is:

  1. Dump the PT100 amplifier entirely
  2. Connect the PT100 thermistor to T0, T1 or T2 on the Re-Arm board
  3. Set J7, J8 or J10 as appropriate
  4. Highly recommended: replace the 4.7K pullup with a 1K pullup. This gives better resolution.

I have the advantage of several bad assemblies where I can salvage SMD parts.

The big advantage of the PT100 amplifier is resolution. At 210C the PT100 amplifier has a resolution of about 1.25 degrees. The 4.7K pullup would be 13 degrees. The 1K pullup would be 3.5 degrees.

called it 3 in pins file (never knew why).

Per this diagram, Analog Input 3 is on AUX 2 (pin 63 in the old system, P0_26 in the current system). It won't have a pullup resistor on it so it is a reasonable choice for the PT100 amplifier input IF you can get the PT100 amplifier to behave properly on 3.3V.

Pins that can be used as either analog inputs or digital I/O have two different names. In this case it is "3" if it is used as an analog input. If usesd as digital I/O then it is called "P0_26 ".

re-arm left rev 1

Another alternative (that I have not tried) - use a 3.0V 5% Zener across the PT100 amplifier to guarantee that it'll never exceed the 3.3V input spec on the analog input. The downside is that a new thermistor table would be needed.

khomouda commented 5 years ago

snap_001 I did what you suggested (as in picture) and connected the PT100 directly to T0, and changed #define TEMP_0_PIN 0 // A0 (T0) - (67) - TEMP_0_PIN but it is giving 0 degree at room temperature. note that the PT100 is a 3 wire and I am connecting one of the wires and ground, and the external pad is not soldered to the other 2, it is only glowing, (I looked at it)

landodragon141 commented 5 years ago

For a 3 wire pt100 you can treat it like a 2 wire. 2 of the wires should have almost no resistance between them. These two should be connected together.

boelle commented 5 years ago

Please post your question either on discord: https://discord.gg/n5NJ59y or on facebook: https://www.facebook.com/groups/2080308602206119/ The issue list is for bugs and feature requests only Please close this issue once you have posted it on one of the 2 links Thanks :-D

github-actions[bot] commented 4 years ago

This issue has been automatically locked since there has not been any recent activity after it was closed. Please open a new issue for related bugs.