Hot-end temperature: +328°C

Hello,

My hot-end temperature was, I think, working good but now it's reading 328°C.
I disconnected the thermistor and its value seems correct: 100k. (113k in fact).
I measured the resistance of the e-temp and I read 104 ohms (with the resistor disconnected - yes, one hundred)!!! When I measured the 4,7k resistor (4701 ref.), I get 1200 ohms (because there are other components in parallel I think).

So, I think I got a problem on my Melzi v2.0. and it seems that is not a new board (is it possible?), it seems, it has been repared before (two solders at least have been done by hand).

I contacted reprappro and I'm waiting for a response.

Could you help me, please?

Check for a short between the heater and the thermistor. That usually burns out the input pin on the ATMega and makes it a low resistance.

---

[www.hydraraptor.blogspot.com]

Oooh I hope no... What are my options?

Depends on what you mean by 'Melzi 2.0'.

I think reprappro use Melzi Ardentissimo, which has an expansion header to the bottom right of the board. Grab your soldering iron, construct a new thermistor circuit, add a pin header to the Melzi, reconfigure the firmware to suit. If you have Melzi Leonardo then you are, AFAIK, out of luck as theres no expansion header on that board. You could use the b-temp connector and configure the firmware to read it as the extruder thermistor, but you'd lose the ability to control the heated bed from the board.

In any event, the first thing to do is make sure this can't happen again. Have a good look at the hot end and make sure theres no possibility of shorting the heater to the thermistor.

Hello,

Thank you for your interresting answer. On my board, I can read: reprappro.com Melzi v2.0 designed by Joe Mosfet & Adrian
The microcontroller used is an Atmega1284P.
I'm looking for the schematic. I found that: [reprap.org]
but it's not the same microcontroller as mine.
I'm eletronic engineer. I know PIC microcontroller better but I think I will be able to program another Atmega and solder it on the board.
I have a "6 pins" expansion header which I think is the SPI AVR and a 10 pins header.

For explanations, does the atmega "breaks" because the thermistor is powered in 12V? If yes, it will be cool to add Zener diode to input entries...

Hum There are 8 ADC inputs on a 1284P so I think may be I could change the pin.
[www.atmel.com]

What do I need to download to compile the source code etc. Do you have a "wiki" for that?

Thank you!

You'll need the arduino IDE, the Gen7 arduino support package since arduino themselves don't support the 1284p, and a firmware. Reprappro provide a customized Sprinter which you can find in their github, Marlin supports Melzi, other firmwares should be usable but may take more work if they don't come with Melzi pin maps.

I'd go with reprappro's Sprinter, if only because should come with good default settings for a reprappro mendel. The only change you'd need to make would be the thermistor pins on line 450 and maybe the thermistor tables.

Hello,

For others people who where interrested to do the same, you can follow this link: [reprap.org]

I installed Arduino 1.03: [arduino.cc]
I installed the Gen7 2.1 in the hardware folder: [reprap.org]

Now in the menu, I can select for Atmega 1284P with 16Mhz Or 20MHz clock:
I exported the schematic from Eagle into a pdf: melzi.pdf

As you can see, a quartz of 16MHz is used. The hotend thermistance is connected to the PIN30/ADC7. There are 4 pins which are not used and there are connected to the external pins header (pins 33, 34, 35 and 36 - ADC4 to ADC1).

I found the schematic of the board on github here: [github.com]

Now I'm confused by the source code to use. If I follow you I need to take that: [github.com]
but it's marked as deprecated

The maintenance page says to use that instead: [github.com]


Before to do re-upload a new firmware, I'm going to unsolder the pin30 of the microcontroller. So I coud observe If the resistance is yet to low else I will look for a shortcut on the board.

Putting 12V on a 5V input definitely damages it. Sometimes it damages more of the chip as well.

The difficulty of protecting it with a zener is the leakage current will affect the reading at the bottom end. There is only a fraction of a volt headroom. You also need a series resistor to stop the zener exploding when you connect it to a 20A PSU. That will also introduce measurement errors due to the pin leakage.

---

[www.hydraraptor.blogspot.com]

First, I apologise for resurrecting such an old post, but I couldn't find a similar discussion in more recent threads.

I, too, short-circuited my microcontroller. I decided to follow Andrew Smith's suggestion but, since my knowledge on this subject isn't exactly outstanding, I have decided to ask a few questions before I proceed.

1. Resistor: Should I use something similar to the 1002 (10KΩ) instead of the 4K7Ω shown in the diagram? I find it strange that they are different.
2. Capacitor: The capacitance seems to match. Is V a voltage code?


3. Is it a problem to use any components as long as they meet the basic specs? For instance:


Thanks in advance!

---

Digital Fabrications Experiments

from [reprap.org]

Quote
http://reprap.org/wiki/Melzi
The BTEMP and ETEMP resistors (R4 and R5) are the wrong value. The original Melzi was also 10K but changed to 4.7K over two years ago (The Melzi 2.0's use the correct value), however Geeetech is still using the wrong value. This hurts the precision of the thermistor, therefore it is better for to rework and replace with 4.7K than use 10K and change the thermistor tables in the Marlin firmware.

so if your going to re do your firmware use a 4.7K then you can use standard marlin thermistor tables

I would use a %1 resistor
the cap is on 5v so anything 4.7uF over 5v should be fine

Thank you, Dust!

I suspect the short circuit may have damaged other parts of the board, though.

That's my suspicion, at least. Here's why:
The alternative circuit was buit (plan & circuit) and the firmware (Repetier) altered accordingly, it still doesn't work (temperature readings are constantly at 958°C / 1756°F)

Oddly, the voltage reading between any of the analogue pins and the GND pin (expansion header) are of less than 1V.
Shouldn't it be 5V?

Things I also checked but were all fine:

• continuity of the wires
• thermistor

Edited 1 time(s). Last edit at 10/19/2017 07:25PM by SteMega.

---

Digital Fabrications Experiments

Its an input... not an output, so no I would not expect 5v

It is possible you have zapped the ADC. Or it just died....