Hot end on Prusa_Mendel i2 not heating up any more. Thermistor OK AFAIK

Hello everybody

I have been getting acquainted with my pre-used (is it a Prusa-Mendel i2?) for a week or so now. I have not printed a single item yet though Cura drew a bad bed-calibration rectangle (should that have been a square?). I was prepared to have a lot of learning todo, wanted to build my Prusa or any printer myself and soldered a RAMPS 1.3 board but when that was done I got offered a competely-built printerwith a RAMPS 1.4.

Right now I have quite the show stopper waiting for a solution by the RepRap forum savants.

Since yesterday my hot end does not heat up.
Pronterface's temp. reading was first 357° and stable. Hot end stayed cold.
When i clipped the thermistor under the assumption it were the cause,
temp. fell to 0°C and hot end STILL did not heat up.
I exchanged the thermistor -> no change in behaviour. Thermistors seem to be working, resistance dropping with rise in temp.
I found a heater lead loose and reattached it but no change.
what may be at fault, what can i do to check or even repair? I hvenothing but a solder iron and a DMM to work on the elctronics though.

Some system info:

My setup as far as I see and know: (photo available below, if attachable)
Steel rod frame,
plastic printed joints (ABS or PLA?),
metal geared Wade extruder,
unknown hot end with brass nozzle 0.5 mm
carbon fiber hotbed
Arduino Mega 2560 RAMPS 1.4
plnning to use a Raspberry Pi B+ lying around, yet µSDcard does not boot it yet.

So I set up a few linux'es (Puppy precise and Debian several times each) to get my recommended Arduino-Slic3r-Pronterface suite up.
I had lots of trouble with compilation of both arduino (0023 or 1.05?, tried both) and Pronterface due to missing libraries, version inconsistencies and may be unsolved dependencies due to errors in Puppy's packet manager and my apt work-around. With Pronterface not extruding during print, Cura not heating the bed and all the compilation isues I feel quite like the pioneer I would not want to be.

But be all this as it may I am posting here because of the above stated maybe hardware issue because of course I am doing things to my printer already (calibration, modifications before print of first part).

Have a good day everybody
Geebee

---

Present hw setup: Prusa Mendel i2 with Arduino Mega 2560, RAMPS 1.4 , J -Head cloned HotEnd, .5mm nozzle, metal geared Wade extruder, CarbonEpoxy bed, Raspberry Pi B+ (intended, not yet fully functional), Debian & XP on desktop.
Eventual typos and omissions in above text(s) due to silicon rubber keyboard extremely hard 'n tricky to type on, dropp'n nd garbl'n nything.

The picture you posted is looking at the back of the printer - not that it matters, but when it homes, you want the hot-end nozzle in the front left corner (in case you wanna grab a bit of oozed plastic) rather than back right as it would be if you used it like that.
If you saw 357C in Pronterface, that would suggest a short circuit. If you saw 0.0C, that would suggest an open circuit.
I wouldn't be mucking about with Raspberry Pi's and linux flavors etc 'till you've got the thing working and have gained some experience - just use your windows PC.

---

_______________________________________
Waitaki 3D Printer

Yes, waitiki, thanks for the hint to ergonomy. Of course I know that it is seen from the back here and I hate using it thus, only when it came attached to the desk that was the way to cram it in my little space. Now I have lost my fear of discomforting it and
my belief in some of the seller's statements upon handover about "Never touch ..."(not applicable here obviously) and others I will change that as soon as I have a decent piece of multilayer wood for a mount.
So 0°C suggest a short circuit. Again i was not ignorant of that possibility but I have about 5V on the lead then! I had checked the replacement of course before soldering. So both are a hundred sth. KOhms. As soon as I make the temp meter show values
SENDING:M105
ok T:280 @:0 B:23
other than 0 and below the targetted temperature (I set it to 300 just for the test) P'face starts to heat the nozzle. What may be the cause then?

I hope you are not at your end's wit there. )

Thanks again and read ya
G.

Edited 3 time(s). Last edit at 01/12/2015 06:52AM by GeeBee.

---

Present hw setup: Prusa Mendel i2 with Arduino Mega 2560, RAMPS 1.4 , J -Head cloned HotEnd, .5mm nozzle, metal geared Wade extruder, CarbonEpoxy bed, Raspberry Pi B+ (intended, not yet fully functional), Debian & XP on desktop.
Eventual typos and omissions in above text(s) due to silicon rubber keyboard extremely hard 'n tricky to type on, dropp'n nd garbl'n nything.

No, 0.0 degrees suggests an open circuit.

I'm not sure what you're saying though? If the machine is cold, i.e you have just powered it up, and you connect with Pronterface ok - is the temp reading in P'face around room temp? If so, what happens when you click "set" to switch on the heater - does the hot-end start getting hot physically and does the rise in temp show in Pronterface accordingly?

---

_______________________________________
Waitaki 3D Printer

You are fighting two battles.

First, let's sort out the hardware...

Disconnect the extruder and the thermistor from the RAMPS board and check the resistance through them. The thermistor should be below 100k and will change when the temperature changes.

The hot end heater should also show resistance, but a much lower value. If it shows zero ohms, it is shorted. If you get no reading, then it is an open circuit.

Let me know what you find and we will go from there...

---

-David

Find me online at:
Thingiverse
Instructables.com
LinkedIn
Facebook

Check out my FolgerTech Prusa i3 (plexi) at MindRealm.net

Hello back everybody,

when I did not get answers right away I restlessly decided to eliminate my personal causes not being electrically fit for Arduino and RAMPS and all that semiconduction. I measured a resistance of somewhere around 100 KOhms (fluctuating by +-5 KOhms, which may have been caused by a weak battery in my DMM, as I think today, when more peculiar measurements and loss of conductivity of the wires to the heater happened). So now I know that both the thermistors which I alternatively tested for the hot end make my Arduino circuit feel some measly 27°C when factual temperature is already well above 100°C (I brought the thermistor in contact with the front glass of a 35W halogen GU5.3 lamp). This threshold and little sensitivity probably had tricked me into thinking there was no measuring done at all by the Arduino. My resistor heater cartridge for the nozzle had 5.3 Ohms on my weak DMM today (yesterday I had shrunk back from further dismanteling the hot end of its capton tape and extracting the heater since I had neither a suspicion nor capton tape), thud i t might be a 30W type? The voltage across the resistor being several mV instead of the expected 12V (or should it be 24V or else?) I think that means that the MOS-FET is defect, am I correct?

How do I check the MOS-FET? I read about 5V and 12V to be found on the pins? Is there a way to do it on-board, in or off operation? Do I have to pay attention to polarity?

On Johnnyr's AMPS circuit plan there is a trio of STP55NF06L. I haven't identified mine yet because it is too crowded in that glued-together electronics box. Would it be OK in any case to order a MOSFET ST Microelectronics STP36NF06 (N-Channel, TO-220AB, I(D) 30 A, U(DS) 60 V as a replacement?

Thank you
GeeBee

Edited 1 time(s). Last edit at 01/13/2015 04:49PM by GeeBee.

Quote
GeeBee
The voltage across the resistor being several mV instead of the expected 12V (or should it be 24V or else?) I think that means that the MOS-FET is defect, am I correct?

Here's how to find out what's wrong...

Use this diagram as a reference:


With the RAMPS 1.4 connected to the power supply, and the wires removed from the D9 and D10 terminals, probe the following with a DC voltmeter, set to 20 volts if not auto-ranging.

1 - Black (negative) probe on 1 and Red (positive) probe on 2 This should read around 12 volts. If not, make sure the power supply is on.

2 - Black probe on 1 and Red probe on 3 This should read around 12 volts.

3 - Black probe on 1 and Red probe on 5 This should also read around 12 volts.

If step 2 AND step 3 resulted in readings of 0 volts, the 6A (smaller) fuse is blown. If only one of them read around 12 volts and the other one 0 volts, then there is a bad trace on the board, but the fuse is still okay.

If step 2 AND step 3 resulted in readings around 12 volts, this proves that the positive 12 volt path is good from the power supply, through the fuse, to the D9 and D10 positive outputs.

4 - Red (positive) probe on 2 and Black (negative) probe on 4 This should read 0 volts and the LED on the far left, over the D should be off.

5 - Red probe on 2 and Black probe on 6 This should also read 0 volts and the second LED from the left (over the 1) should be off.

If either of the readings result in a positive voltage, the MOSFET is probably damaged. If no voltage or LEDs, continue...

Activate the fan by sending M106 S255

6 - Red (positive) probe on 2 and Black (negative) probe on 4 This should read around 12 volts and the LED on the far left, over the D should be on.

Activate the extruder heater by sending M109 S40

7 - Red probe on 20 and Black probe on 6 This should also read around 12 volts and the second LED from the left (over the 1) should be on.

If either step (6 or 7) results in a reading of less than 10-12 volts, and the corresponding LED is lit, then the MOSFET is bad. The LEDs are lit (or flickering) according to the PWM output of the Arduino, which sends the same negative signal to the MOSFET, which in turn amplifies the potential to ground at D10 and D9, thus increasing the output voltage (as referenced from the constant 12 volts and the MOSFET ground-ish output).

Sounds confusing, sure. But follow this quick test and you will determine if the RAMPS components are good.

Edited 2 time(s). Last edit at 01/14/2015 11:05AM by MindRealm.

---

-David

Find me online at:
Thingiverse
Instructables.com
LinkedIn
Facebook

Check out my FolgerTech Prusa i3 (plexi) at MindRealm.net

Wow! Mindrealm, that is quite a manual you sent me there.
I appreciate that very much and will be doing all or most of that right away.

My terminal D9 (or 3,4 as your pic goes) is empty by the way.
The 3 MOS-FETs read thus:
P55NF06
7SABK VW
PHL 311
ST e3

Write you later.
GeeBee

Test results:
Step 1: +12 V OK (I had switched the leads on the DMM yesterday, fool that I am. No wonder the conductivity tests were inconclusive yesterday!)
Step 2: +12 V OK
Step 3: +12 V OK
Step 4: 0 V OK
Step 5: - 5,44 V KO, LED4 on (negative voltage, what does that tell you? I found no 20 in the pic above. But you may have meant the 20 of the I2C on the PCB upper right corner?)
No fan attched (to the Hot End?)
dit step 6 all the same:
Step 6: 0V
The rest I omitted because the HotEnd is disassembled. I'll rip the electronics and case abpart now an prepar for the soldering of my MOS-FET (which? Q1?).

Thanx again and read ya, D.!
GeeBee

Hmmh, if it's just the MOS-FET that's on strike could I then use the extruder2 logics (setting the thermistor leads and the firmware switches acordingly)?

GeeBee

Here's what your results reveal:

Step 1 at 12 volts proves your power supply is on and 12 volts is getting to the RAMPS. >> Good

Step 2 and 3 at 12 volts proves your 6A fuse on the RAMPS board is not blown. >> Good

Step 4 at 0 volts shows that the D9 MOSFET is not 'leaking' when it is off. >> Good

Step 5 > Please recheck. I meant 2, not 20 -sorry. The Red probe should be on 2 and the Black probe on 6. The hot end should be switched off (M109 S0) before testing.

MOSFET Q1 is for output D10, Q2 is for output D9 and Q3 is for output D8.

No. You cannot use the second extruder wiring for an output to heat the hot end. This requires a PWM output at 12 volts. This is not present in the stepper driver socket. Technically, you could use D9 and change the firmware to use that output for the hot end. The fan could be wired directly to the 12 volt inputs (1 and 2 in the image) to run 100%. My fans are connected elsewhere on the RAMPS to run 100% of the time that the power supply is on, and I use the D9 output for print cooling.

Please let me know what you find with rechecking step 5.

-David

---

-David

Find me online at:
Thingiverse
Instructables.com
LinkedIn
Facebook

Check out my FolgerTech Prusa i3 (plexi) at MindRealm.net

Hello again, masterMindsrealm

Continuing from before:
Step 5: red 2, blk 6, hot end switched off (M109 S0 entered, no echo unfortunately): 0 V OK
M106 S255: got no fan for the extruder, still entered the cmd, LED3 went ON
Step 6: red 2, bk 4: 0V, LED3 on
activated extruder heater by sending M109 S40
Step 7: red 20, blk 6: 12 V, LED4 off, LED 3 still on because of step 6 (just in case 20 should again have been 2: red 2, blk 6 gives 0V)
So what does this tell us? Is the MoS-FET Q1 not necessarily bad, but probably the Arduino?

Hope to read you soon!
G

Edited 1 time(s). Last edit at 01/15/2015 05:02AM by GeeBee.

Your results indicate that you do have a bad MOSFET. The LED indicates that the Arduino output is operating properly, and should have switched the MOSFET at the same time.

To be sure, now test as follows:

Power supply on, M106 S255 and M109 S40 sent to turn on both outputs.

D9 final test: Black probe on 4, Red probe on 3 should be 12 volts if the LED is on, should be 0 volts if the LED is off (by sending M106 S0). If no 12 volts when the LED is on, the Q2 MOSFET is bad.

D10 final test: Black probe on 6, Red probe on 5 should be 12 volts if the LED is on, should be 0 volts if the LED is off (by sending M109 S0). If no 12 volts when the LED is on, the Q1 MOSFET is bad.

I know... it seems like this test would have been enough. However, the earlier tests ruled out the power supply, the 6A fuse, the 6A 12 volt traces on the board and the Arduino.

---

-David

Find me online at:
Thingiverse
Instructables.com
LinkedIn
Facebook

Check out my FolgerTech Prusa i3 (plexi) at MindRealm.net

You are in error, David, if you think I'm getting angry or anything at the length of the test procedure. I am happy for any test to subject my device to and the longer it take the better to rule out issues. My first question was whether I had to reconnect the bed heating D8. This written, I proceed to the tests.

Thx
G

Oh, error is min. It is/was D10 which is not yet reconnected.

Both Q9 and Q10 seem dead because Arduino shows transmission, yet no LEDs alight and no Voltage whatsoever, DMM counter-checked, user error ruled out too.

So what MOs-FETs should I be ordering?

I have 3 of type
P55NF06
7SABK VW
PHL 311
ST e3
sitting on my board.
I could order ST's STP16NF06L N-channel case TO-220AB I(D) 16 A U(DS) 60 V
or STP36NF06 (N-channel, TO-220AB, I(D) 30 A, U(DS) 60 V, a bit more potent than the latter).

Thank you very much, David

The trend seems to lean toward replacing them with either FDP8870 or IRLB8743. It is important to make sure there are no shorted peripherals (fans, heaters, etc.) being plugged into the new mosfets, and that the exposed tab is at 12 volts (power supply output) when the board is powered, so also guard against grounding them on anything (tools, wiring, mounting hardware, etc.).

---

-David

Find me online at:
Thingiverse
Instructables.com
LinkedIn
Facebook

Check out my FolgerTech Prusa i3 (plexi) at MindRealm.net

Hmmh, so I will have to find a new casing soon.
And now I understand that tranparent acrylic is not a bad idea for Arduino and its backpack.

"Exposed tab" maning what exactly, David?

The metal part of the MOSFET that sticks up and has a hole in it. These are always at the save voltage as the power supply. Thus, they are 'live wires' sticking up, ready to cause trouble

---

-David

Find me online at:
Thingiverse
Instructables.com
LinkedIn
Facebook

Check out my FolgerTech Prusa i3 (plexi) at MindRealm.net

The FDP8870 is way more than double the price of the IRLB8743, which has Rds(on) of 3.2 mOhms compared to no data for the FDP8870 which will take a week to oreder. Will try to get 2 IRL's then. Thank you ver much Did and I hope I will do everything right this time I know my devices and the dangers. BTW: The end was covered in capton (thermal insulation probably), the wires in transparent plastic tubing.
I should probably try to find me some fiberglass hose for the wires near/in the Hot End. My thermistor has too little distance between the leads for that stout plastic or so I think.

G

Edited 1 time(s). Last edit at 01/15/2015 09:01AM by GeeBee.

Yes, I touched D9 ever so slightly while soldering/screwing some leads yesterday and made the LED light up... Maybe that's why D9 was gone today.

So fare thee well till next time and be that long off, MasterMindsrealm.

Tx for the testing manual. Archiving it.

GeeBee

Edited 1 time(s). Last edit at 01/15/2015 09:02AM by GeeBee.

Ah, what about the location 20 in step 7 of your/my testing manual? An error same as Step 5's 20?

Edited 1 time(s). Last edit at 01/15/2015 09:06AM by GeeBee.

Regarding the two 30 V MOS-FETs IRLB8743PBF / FDP8870
I see in their specs that they differ mainly in
Drain current of course: 19 / 150 A
typ. gate charge @ Vgs 36 nC V @ 4,5 / 106 nC V @ 10
typ. turn-Off delay time 25 ns / 70 ns
typ. turn-On delay time 23 ns / 11 ns
The typical entry capacity @ Vds of both is 5110 pF V @ 15 / 5200 pF V @ 15
which differ considerably from the 1700 pF V @ 25 Vds of the STP that I will be replacing.
Would you venture so far as to recommend the one over the other?
I do not know whether turn-on may be more important as turn-off lag.

That's it for the day then, I think.
I have just ordered some of both yet I would like to know what to solder preferably.

GeeBee

Edited 1 time(s). Last edit at 01/15/2015 11:07AM by GeeBee.

Honestly, the minutia is just that. They are just switching a few volts DC at less than 6 amps, and nano-second timing doesn't matter at all in this application. You will not be able to tell the difference.

---

-David

Find me online at:
Thingiverse
Instructables.com
LinkedIn
Facebook

Check out my FolgerTech Prusa i3 (plexi) at MindRealm.net

New woes

Edited 1 time(s). Last edit at 01/20/2015 02:55AM by GeeBee.

Hello Master David

Today I soldered IRLB8743 into places Q1 and Q2 (unlucky me! The defective devices unluckily had their legs bitten off as short as possible so that I had to desolder them with my wick and pump which both did not work well possibly due to bad sight, jittering hands and not lastly the age of my desoldering wick and the amounts of solder deposited through the hole. So I ended losing the middle pad of Q1 on the underside of the shield PCB which seems to be connected to the negative terminal of D10 but not connected on the upper side. Am I right?
As to the testing:
I can not see the switching which the Arduino does on Q1 upon the reception of my command or would the following sequence of entries in the Repetier-Log tell me as much?
22:09:17.187 : N704 M109 S40 *127
22:09:17.656 : ok

MOS-FET Q2 on D9 however works fine. The fan on/off signal is transmitted. LED3 shows the toggle alright, voltage is 12V if on, no fan attached though.

Extruder heating and thermistor are working yet the thermistor is about 200 below the factual temperature. Ist that a case to be solved by thermistortables / firmware. I ordered an Epcos NTC-Thermistor (type B57540G0104J, 100kΩ, 18mW 4,5%/°C, tol. ±5%, Ø 0.8 x 1.4mm ) along with the MOSFETs. Not knowing the performance of my old thermistors I may have to put it to use after hopefully solving my MOSFET issue first.

Having used 3 and possibly destroyed 2 of my new MOSFETs for Q1 I better wait for your eventual reply now.

Read you tomorrow?

Edited 1 time(s). Last edit at 01/19/2015 04:51PM by GeeBee.

Sorry to hear about the lost solder pad. There is a solution!

The MOSFETs are connected like this:

Q1
Pin 1 (farthest from the D10 terminals) is connected to the left of R13, just above the pin.
Pin 2 (center pin) is connected to D10 negative terminal.
Pin 3 (nearest to the D10 terminals) is connected to the GND bus (common with Q2 Pin 3)


This image shows the traces on the RAMPS 1.4 board so you can visualize the connections

It is perfectly fine to use wires to jumper the connections where traces or pads are missing or damaged. Just be sure to solder carefully to avoid connecting things accidentally.

Quote

I can not see the switching which the Arduino does on Q1 upon the reception of my command or would the following sequence of entries in the Repetier-Log tell me as much?
22:09:17.187 : N704 M109 S40 *127
22:09:17.656 : ok

Yes. This indicates that the D10 output was sent a command to set the temperatureto 40C.

---

-David

Find me online at:
Thingiverse
Instructables.com
LinkedIn
Facebook

Check out my FolgerTech Prusa i3 (plexi) at MindRealm.net

Thank you David for again helping me up and for the tracemap.

I took a day's break after yesterday's losses. Hopefully I'll find the courage to heat up the iron again tomorrow because the SMD devices are way too tiny for my liking especially since my eyesight etc. seems to have deteriorated considerably.
And then there is still the bad response of that HotEnd thermistor which has to be overcome because I couldn't possibly print with ABS of lukewarm 400°C, could I?

Read ya!

Ah yes - I forgot to mention some bad I had intended to hide:
One of my many bad yesterday's deeds was to measure on more than the points you wrote me and (probing 5 against the MOSFET's legs, or so I think) somehow I managed to repeatably contact the MOSFET Q1 in a way such that LED4? (second LED from the left - over the 1) glowed and the HotEnd heated up. Unfortunately Q¹ heated up too as I detected upon the last occasion so that I may have ruined that device twice yesterday and I did not want to try a third time then.

'til soon
Geebee

As a closing comment to this thread:
Thank you again to all those (waitaki and Mindsrealm - you are unforgettable) who helped me in my ignorance.
I hereby confess I bought a new set of Arduino/RAMPS and will be keeping the old one and the two remaining exchange MOS-FETs till I know better what may have gone wrong.
As it turned out, this was a good decision, but of course a bit more expensive too, because a stepper cable was defective too - I had spent several hours on those occasions disassembling my z-axis mechanics since I had erroneously attributed the steppers' blocking and jerking to some kind of overload.

GeeBee

Edited 1 time(s). Last edit at 01/31/2015 04:35AM by GeeBee.